(Dieter Mans, Gerhard Preyer (Eds.) ) Proto Sociolo (BookFi) PDF

Proto
Sociology
An International Journal of Interdisciplinary Research
VOL. 13, 1999
REASONING AND ARGUMENTATION

Dieter Mans, Gerhard Preyer (eds.)
Content
Introduction:
On Contemporary Developments in the Theory of Argumentation
Gerhard Preyer, Dieter Mans 3
Ralph H. Johnson
Reasoning, Argumentation and The Network Problem 14
Leo Groarke
The Fox and the Hedgehog:
On Logic, Argument, and Argumentation Theory 29
J. Anthony Blair
Presumptive Reasoning/Argument: An Overlooked Class 46
Robert C. Pinto
Argument Schemes and the Evaluation of Presumptive Reasoning:
some Reflections on Blair’s Account 61
Douglas Walton
The New Dialectic: A Method of Evaluating
an Argument Used for Some Purpose in a Given Case 70
Manfred Kienpointner
Comments on Douglas Walton’s Paper: The New Dialectic: A Method
of Evaluating an Argument Used for Some Purpose in a Given Case 92
Christopher W. Tindale
The Authority of Testimony 96
John Woods
Peirce’s Abductive Enthusiasms 117
Henry W. Johnstone, Jr.

“‘Any,’ ‘Every,’ and the Philosophical Argumentum ad Hominem” 126
ON CONTEMPORARY PHILOSOPHY
Hans Lenk
Interdisziplinarität und Interpretation 133
Ellery Eells
Causal Decision Theory 159
Authors 181
Imprint 182
On ProtoSociology 183
Volumes 184
Bookpublications 190
Subscription 204
Digital Publications – Special Offer 205

Introduction: On Contemporary Developments in the Theory of Argumentation 3
INTRODUCTION:
On Contemporary Developments in the Theory of Argumentation
Stephan Toulmin’s book – The Uses of Argument (1958) – was a bold attack
against the subsumption of theories of argumentation under formal logic.
Toulmin’s book received very critical reviews but even his harshest critics
cannot deny that many contemporary developments in the field are consistent
with his basic line of reasoning. Since the 70's there are broad researches on the
theory of argumentation called “informal logic” or “critical thinking”. Many
flowers have grown – some say too many – and our biotope shows a great
diversity.
Some points can be made to show the relevance of this biotope. Here is one
of them: The methods of scientific research is observation, description, and
theorizing. The medium of understanding scientific and other cultural
knowledge are texts with their arguments and explanations. Written language
is a medium which takes structural effects in the semantics of social structure.1
Content analysis is a tool for understanding texts and thereby society and it is
therefore a basic research method in the social sciences. The analysis of argu-
ments2 is a special form of content analysis and given the frequency of argu-
ments we should expect a lot of work in that area done by social scientists.
That is not the case. Protosociology intends to popularize the theory of argu-
mentation as a challenge for theoretical research and a possibility for new
applications. We publish some papers by leading researchers in the field to
outline the topology of a new and powerful interdisciplinary theory. We have
compiled these results and hope that this compilation will spur the interest in
the field. If so, we intend to publish an issue which shows the possibility of
applications for the social sciences.
In contemporary research we recognize elaborated results, expansions of
discovery, and new developments of conceptualization of the initial subject.
Some of the leading questions in particular in our collection are:
1 See in particular N. Luhmann, Die Gesellschaft der Gesellschaft I (2 Vol.), Frankfurt am Main
1997, pp. 249-290.
2 We prefer the terms “analysis of arguments” and “theories of argumentation” over “informal
logic” or “critical thinking” simply because the word “argumentation” points to the type of work
we are doing.
4 Dieter Mans, Gerhard Preyer
How shall we distinguish between argumentation and reasoning? What is

the meaning of crucial terms like inference and implication? What type of
inferences do we apply in everyday argumentation?
One of the significant problems in contemporary research is the analysis of
presumptive arguments because such arguments are a feature of common
argumentation.
How fruitful is the conception of New Dialectic to the study of fallacies?
What is an argument scheme and how can it be used to criticize an argument?1
When we write down a coherent sequence of sentences (may be a scientific
paper, a report on a sport event, a recipe for cooking, or a letter to a friend)
some of these sentences may be identified as an argument. Suppose some days
later you read your own paper and you say to yourself: “I tried to find an
argument, but I did not succeed.” But even if you were not successful in
finding the intended argument, your paper contains the trace of a mental
activity. Now if you try to describe, what was going on, you might say, “I
thought about x” or “I reasoned about x”. Most – if not all! – researchers
assent to the following statement: You cannot find an argument without
reasoning, but you can reason without finding an argument. It follows, that
reasoning and argument cannot be the same.
Ralph H. Johnson’s contribution gives a highly readable exposition which
clarifies fundamental terms in the theory of argumentation. His leading questi-
on is:
How shall we distinguish between argumentation and reasoning? What is
relationship between reasoning and rationality, intelligence, knowledge and
thinking?
He reviews some of the positions that other theorists have taken on the
relationship between argumentation and reasoning. Specifically, he addresses
the views expressed by Finochiarro (1984), Govier (1987), and Walton (1990)
and discusses their shortcomings. His answer is that reasoning can best be
understood as the generic concept of which argumentation is a species.
Besides that, his paper indicates that there is productive confluence of
cognitive science, philosophy and psychology in the theory of argumentation.
For the novice the theory of argumentation may seem more like a network of
different disciplines.2 That is not the use Johnson makes of the word
“net(work)” but it is a fine metaphor to characterize the multi-disciplinary
approach of our discipline. He uses the word to analyze the relationship
between the family of terms like reasoning, knowledge, rationality, intelligence
1 See on the topics, the new developments and problems of a theory of argumentation in particular
the contributions of R. H. Johnson, D. Walton, and J.A. Blair, in this vol.
2 But that view would disregard the genuine contributions of the theory of argumentation.
etc. Maybe that these philosophical family resemblances are one reason, why
the theory of argumentation is not too popular in the social sciences.
Suppose you are well acquainted with the family resemblances in the
argument/reasoning community and you have identified an argument. That
must be an easy task – at least under normal conditions – because taking part
in a discussion and replying to arguments is a standard social activity. Proba-
bly your next step would be to identify certain parts of the argument. You
might reason: That must be easy too, because if I can identify an argument, I
can identify the parts of an argument. But it is not. Arguments like many other
abstract entities are very difficult to break up into smaller pieces, compare the
difficulties we have in linguistics with sentences and their parts. Thereby the
subject of theorizing shows us a holism like in the semantics of natural langua-
ge (D. Davidson). It is irritating – not only for the novice – that a reconstruc-
ted argument – the real structure of the argument – can be very different from
the real argument in a text.
What are the premises and what exactly links the premises to the con-
clusion?
The cement of the argument1 which glues together premises and con-
clusion(s) is at the center of an intensive debate – see also Gilbert Harman
(1986), he suggests it is a mistake to “conflate” reasoned change of belief with
argument and proof, Don S. Levi (1995) rejects the assumption that arguments
are sets of premises and conclusions, Michael Gilbert (1997) argues for a
radically non-standard, non-linear, and non-logical conception of “coalescent
argument” which is needed to explain everyday argumentation.
In his contribution, Leo Groarke presents a wonderful parable about a
hedgehog and a fox. The fox uses 36 tricks to escape his enemies, the hedgehog
has only one, but this one and only trick works perfectly. The fox symbolizes
of course the type of argumentation theory, which uses many type of glues
while the hedgehog has only logic to tie premises and conclusion together. To
understand the hedgehog in his “one glue fits all” theory, lets look at an ex-
ample:
(P1) John is a mountain climber.
(P2) All mountain climbers are well trained.
Therefore
(C) John is well trained.
1 This is a variant of J.L. Mackie’s, Cement of the Universe, Oxford, 1974, which refers to causali-
ty.
This is a perfect argument in the following sense. Since the conclusion is a

logical consequence of the premises, you cannot find a better glue between
premises and conclusion. Yet suppose you try to find a better connection:
(P3) If (P1) and (P2), then (C) John is well trained.
Therefore
Someone who tries to find an implicit premise would have misunderstood the
nature of logical inference. There is nothing hidden and there is nothing to
make explicit. The same is true for the next trial, the change of a (deductive)
logical conclusion in an inductive one. Such an argument would be absurd:
(P3) In all known cases: If (P1) and (P2), we had (C) John is well
trained.
Therefore
The moral is: A deductive logical argument is complete since the conclusion
follows from the premises. Any non deductive argument can be turned into a
deductive logical argument, by adding an implicit premise. It is this asymmetry
which seem to give the hedgehog his superiority over the fox. Groarke’s point
is not, that there are no other types of argument. You cannot change an induc-
tive argument into a deductive one by any trick, if you consider the force of
the argument. But is the distinctive property of deductive arguments that
defines a challenge:
Are there arguments which cannot be reduced to deductive arguments in
the sense Groarke describes?
Groarke illustrates his discussion of persuasion with visual images which are
normally taken as paradigm examples of non-logical persuasion. And he
argues that the discussion of argumentation, reasoning and persuasion in a
variety of disciplines much too quickly dismisses the classical approach that he
proposes.
The argumentation fauna is inhabited by much more foxes than hedgehogs.
Probably Groarke is the only hedgehog in our collection. J. Anthony Blair
discusses the questions that a complete theory of inference/argument schemes
should answer. His aim is a metatheory of inference/argument schemes.
But what makes non deductive arguments so attractive? Why does the 36
– 1 tricks attract so much attention?
Blair gives an answer by looking at an very important class of arguments.

It is no exaggeration to say, that this class of arguments is by far the largest.
That an argument as presented in a text does not meet the standards of deduc-
tivism is of course no problem for Groarke. What Blair intends to show is, that
you cannot turn those arguments into deductive ones without changing their
very nature.1 Here is a realistic example:
(P1) In Kosovo people are murdered by paramilitary forces.
Therefore
(C) The use of military force for the protection of the people is
legitimate.
This argument seems to be a good one, many would agree. But you cannot
deduce (C) from (P1). Arguments of this type are called presumptive.2 Blair
gives a lot of examples for this type of argument. You can of course add a
premise, which makes the argument deductive and turns a presumptive in a
deductive argument:
(P2a)If in Kosovo people are murdered by paramilitary forces then
the use of military force for the protection of the people is legiti-
mate.
Therefore
legitimate.
Now the premises entail the conclusion but (P2) is too special. It is not the
added premise which gives the argument its force. If you ask someone “why
do you believe (C) given (P1), (P2a)” will not be the answer. Let’s try a more
general solution:
(P2b) If people are murdered by paramilitary forces – in Kosovo or
else-where – then the use of military force for the protection of
the people is legitimate.
Therefore
legitimate.
1 But Groarke might still insist, that his framework is the best way to analyze arguments.
2 This is a rather rough classification. Presumptive arguments are those where you jump from the
premises to the conclusion(s). The exact nature of presumptive arguments is the focus of an
intense debate.
(P2b) is a much better candidate for the answer you will get, when you ask
someone why he believes (C) given (P1). But (P2b) is too general; it is easy to
find counterexamples. The problem is to find a missing premise which belongs
to the argument – i.e. the added premise must be an arguers premise – and not
only a missing premise for the sake of deductive argument. You will find
presumptive arguments in virtually any text. The popularity of this argument-
type poses a serious challenge to the theory of argumentation:
What is the glue between premises and conclusion in a presumptive argu-
ment?
We can phrase the question in another way:
How can we differentiate between a good presumptive argument – where
the premises support the conclusion – and a bad one – where this is not the
case?
This leads us to Robert C. Pinto’s comment to Blair’s analysis of argument
schemes referring also to Walton’s account. Pinto evaluates Blair’s concept of
presumptive reasoning. He agrees with Blair’s that we do not evaluate all given
inferences with our traditional logical procedures, but he makes an important
difference concerning the use of argument schemes. Argument schemes in
presumptive reasoning in contrast to those used in deductive reasoning have
no normative force. To conform to an argument scheme is a necessary conditi-
on for the correctness of a presumptive argument, it shifts the burden of proof,
but it is not a sufficient condition for its correctness.
Douglas Walton’s paper raises many fundamental questions of argumenta-
tion theory and that is no surprise, since Walton has published extensively on
nearly any aspect of the subject. He outlines a new dialectical framework –
called New Dialectic – for the evaluation of arguments. Having its origins in
the old dialectic of the ancient world, the new dialectic is centrally concerned
with presumptive reasoning of a kind that is not well represented by either the
forms of deductive or inductive reasoning. To represent the forms of argument
of this third type of reasoning, argumentation schemes are used that model the
structures of common types of arguments like argument from sign, argument
from analogy and argument from commitment.
We will concentrate our discussion on this type of argument and include
some of the results published in his book on presumptive reasoning.1 For a
social scientist, the inference from correlation to causality defines an important
class of presumptive arguments. Look at the following example: In a certain
1 D. Walton, Argumentation Schemes for Presumptive Reasoning; Mahawh, N.J., Erlbaum 1996.
A book in German language on the subject is by M. Kienpointner, Alltagslogik: Struktur und
Funktion von Argumentationsmustern, Stuttgart 1992.
population 50% of all woman and 20% of all men get low payment. Using the
terminology of social science, we can say, that there is a correlation between
sex and payment. You want to infer that sex is the cause of the difference in
payment, but a social scientist knows that this inference is defeasible. We can
test the inference by asking critical questions.1 Walton specifies 7 critical
questions:
Q1: Is there a positive correlation between A and B?
Q2: Are there a significant number of instances of the positive correlation
between A and B?
Q3: Is there good evidence that the causal relationship goes from A to B,
and not just from B to A?
Q4: Can it be ruled out that the correlation between A and B is accounted
for by some third factor – a common cause – that causes both A and
B?
Q5: If there are intervening variables, can it be shown, that the causal
relationship between a and B is indirect – mediated through other
causes –?
Q6: If the correlation fails to hold outside a certain range of causes, then
can the limits of this range be clearly indicated?
Q7: Can it be shown that the increase or change in B is not solely due to
the way B is defined ... or classified?
Two of the seven answers are easy: Q1 “yes” and Q7 “yes”. The answer to Q2
is “yes” too, but with a qualification. If we compare the results of different
studies, we will find a considerable variance between different correlations. In
few cases, we may even find, that woman are better paid which implies a
change in the sign of correlation between sex and payment. This variance poses
a serious problem which might be cured by an answer to Q4. But any attempt
to find a third factor – or many third factors – will not lead to a stable correla-
tion. We simply do not have theories in the social sciences, which can precisely
give us the factors to control. But we have some knowledge how to influence
the correlation between sex and income – and do that all the time, at least we
try to change it. But since a causal relation cannot be changed by any means,
we probably do not have a causal relation between sex and income.
Why should the non existence of a causal relationship between sex and
income be of any methodological importance in the theory of argumentation?
The problem is that over and above the criteria given by Walton – or any other
1 These critical questions belong to an argumentation scheme. Walton discusses 25 of these

schemes in his book in the framework of his dialectical approach to argumentation.
criteria – you need some substantial field dependent knowledge1 to judge the
correctness of this type of presumptive argument. If this is correct, we will
never – in principal – be able, to have an “Alltagslogik” as a kind of generali-
zed logic to judge the correctness of an argument. Argumentation schemes are
of great value for a practical criticism of arguments. But one fundamental
difference between logic and Alltagslogik remains. Logic works in any context,
Alltagslogik cannot be used, without considering the context of an argument.2
We can differentiate between two hypotheses:
(H1) To judge the correctness of a presumptive argument you can use
a set of (sufficient) rules, which define a reasoning scheme. We
do not know at present if there are 25 or 250 schemes, but rese-
arch in the theory of argumentation will find these schemes.3
(H2) Schemes for presumptive reasoning are of great pragmatic value
for judging arguments, but you can never give a set of sufficient
rules. You will – always? / in a certain percentage of cases? –
need some domain specific knowledge which cannot be captured
in rules.
The difference between H1 and H2 may be residual importance from a practical
point of view. But if one looks at the possible shapes of a theory of argumenta-
tion, the difference is large and reflects another difference: Between those who
think, that human rationality consists of a set of rules – or algorithms, if you
like – and those who deny that. The slow development of Artificial Intel-
ligence seems to support the skeptical position.4
Manfred Kienpointner responses to Walton’s conception to outline an
alternative against the classical scientific philosophy and contemporary post-
modernism5, and evaluate his New Dialectic. He emphasizes the fruitful frame
1 This term has been coined by S. Toulmin (1958).

2 On the foundation of contexts in the theory of interpretation see G. Preyer, “Interpretation and
Rationality: Steps from Radical Interpretation to the Externalism of Triangulation, Protosociology
Vol. 11, 1998: Cognitive Semantics II - Externalism in Debate.
3 C. Lumer has defended an algorithmic/logical theory of argumentation in Praktische Argumen-
tationsrheorie, Brauschweig 1990. On Lumer’s account D. Mans, “Argumentation im Kontext,
Exkurs: Zu C. Lumers “Praktischer Argumentationstheorie”, in: G. Preyer, M. Ulkan, A. Ulfig,
Intention - Bedeutung - Kommunikation. Kognitive und handlungstheoretische Grundlagen der
Sprachtheorie, Frankfurt am Main 1997.
4 One might reply, that argument schemes can never give sufficicient conditions. They can only
indicate, when to shift the burdon of proof. But I doubt that you will get very far with that
substitution. We might still ask “how many rules do you need in a certain context to shift the
burdon?”. You can’t escape contextuality.
5 On modernism and postmodernism see also G. Preyer “Moderne and Postmoderne im Kontext
von Globalisierung”, http://www.rz.uni-frankfurt.de/protosociology
of Walton’s research but articulates also some reservations toward his criticism
of ancient dialectic, and the conceptualizing of presumptive and abductive
arguments.
Cartesian epistemology has rejected the “argument from authority” in
principle, and the refusal of this argument takes effect in modern thinking.
From this point of view the judgement of other people can not be the founda-
tion of knowledge. Yet central parts of our knowledge is not grounded in
personal experience but relies on expert knowledge. One of Walton’s 25 argu-
mentation schemes1 is the argument from expert opinion or appeal to authori-
ty. Chris Tindale takes up Walton’s argumentation scheme but his contributi-
on discusses the appeal to authority from a different and surprising perspecti-
ve. Suppose you are told by a medical expert, that taking a certain sort of pills
will cure your disease without serious side effects. You – as a layperson – can
believe the expert or you can begin to investigate his claims. In most cases –
since your no pharmacologist - the second route is not open for you. You
cannot think through the matter that obviates any appeal to authority. But
frequent appeals to authority reduce your autonomy. This cognitive threat to
our autonomy by our dependence on authority raises some questions:
What is the relation between my cognitive autonomy and the cognitive
autonomy of other people? How is our empirical knowledge formed by the
fact, that we are social beings? And what are the necessary cognitive capabili-
ties to create and maintain our autonomy?
Tindale reasons convincingly, that testimony is a key to answer these
questions. My knowledge about Patagonia for example, can go back to two
sources: Personal experience, and testimony by others.2 To criticize their
testimony, I have to develop a certain cognitive competence. Since testimony
seems to be a more fundamental source of knowledge than expert opinion, this
cognitive competence is the kind of competence we need to criticize expert
opinion. Tindale names basic skills in scientific methodology, reasoning and
ethics. His contribution is a further example for the strong connections the
theory of argumentation has with other disciplines especially with philosophy.
A trained philosopher might not be surprised, that the basic cognitive mecha-
nism of acquiring knowledge are of great importance for the theory of argu-
mentation. But all others will benefit as well from his intriguing and thorough
discussion of testimony.
1 D. Walton, op.cit. 1996.

2 To simplify the matter we disregard, pictures, films etc., since it is not easy to separate these
sources from testimony.
John Woods’ contribution is about abduction or abductive inferences.

Abduction is a sibling of induction and deduction and Peirce is its philosophi-
cal father. If a bag contains red marbles and you take out one marble, you may
infer that the marble is red. This is deduction. If you do not know the color of
the marbles in the bag and take out one marble and it is red, you may infer that
all marbles in the bag are red. This is induction. But if a bag of red marbles is
standing at some place and a red marble lies in the vicinity of the bag, you may
infer, that the marble is from the bag. And that is abduction. What you infer
from the premises – “A bag of red marbles is standing at some place.”, “A red
marble lies in the vicinity of the bag.” – is the conclusion – “The marble is
from the bag.” The hypothesis “the marble is from the bag” could serve as part
of an explanation for the fact, that a red marble lies on the floor. As you will
have noticed, abductive reasoning is a special kind of presumptive reasoning.1
Abductive reasoning is part of our everyday life – if we are looking for ex-
planations - and it is very essential part of any science. Finding explaining
hypotheses is one of the most challenging creative task and that may be the
reason why stories like those about Newton’s apple are so popular.
Abduction is an inference to a potential explanation. Those sociologists
who are engaged in hermeneutic text interpretation know the relevance of
abduction and its problems too. We will illustrate the point with one of Peir-
ce’s examples. You find a stone which looks like a petrified fish far in the
interior of the country. You can explain the finding by any of the following
hypothesis:
a The sea once washed over this land.
b John Doe has left the petrified fish at that place to fool me.
c The petrified fish was transported by some geological process to
the location.
How many potential explaining hypothesis exist?
There are good reason that one will always find an additional potential
explanation beyond a given set of potential explanations. Thus there is a
potentially infinite set of potential explanations.
How do you rule out a potential explanation as incorrect – or probably
incorrect –?
Just take same deductive consequences with an empirical content and test
them. In real life situations very few from a small set of hypothesis will survive
– in many cases there is only one hypothesis to test. That very abbreviated
1 If you take your first steps in the theory of argumentation, you are advised to differentiate
carefully between explanations and arguments. Abduction ties both together, which indicates that
you have left the initial stage.
description points to three ingredients in any theory of abduction. The first

ingredient is a hypothesis generator, the second ingredient is a generator of
testable deductive consequences and the third ingredient is theory of refutati-
on. Even the best understood ingredient – the second one – is very difficult to
formalize, because only a small subset of deductive consequences is “empiri-
cally interesting”. If you still ask yourself:
“what makes abduction so terribly difficult, after all, even the layper-
son does it all the time?”
you should carefully read the first sentence of Woods’ paper. Any research on
the computational aspects of abduction requires a high degree of formalizati-
on. So Woods’ contribution touches an ambitious aspect of argumentation
theory, namely the goal to clarify crucial concepts with formalized methods.
The explication of creativity within a computational framework would have
far reaching consequences over and above the theory of argumentation.
Henry W. Johnstone’s contribution reminds us, that reasoning about reaso-
ning has a long tradition. Probably men begun to use deductive and non
deductive arguments not too long after they begun to use language.1 But the
logic of Aristotle has been for a long time the only tool for the analysis of
arguments. Therefore we are pleased to include a paper with a strong reference
to one of the most influential of all philosophers. Nonetheless our edition may
have convinced some readers that modern theory of argumentation is more
than just a footnote to Aristotle.
Dieter Mans, Gerhard Preyer

J. W. Goethe-University, Frankfurt am Main, Germany
1 We are not aware if the search for “the oldest published argument” is a topic in the community.
Hints are welcome.
14 Ralph H. Johnson
RALPH H. JOHNSON
Reasoning, Argumentation and The Network Problem
I. Introduction
The question is: What is reasoning? Why is this question important generally
and to the cognitive sciences specifically? One strong incentive to ask this
question has come from the increased attention to the importance of teaching
reasoning, which naturally leads to the question of just what reasoning is.
(For discussion of this entry point into the question, see my paper “Reaso-
ning, Critical Thinking and The Network Problem” (1996).) This question is
also of great interest to the cognitive sciences, because one important tributary
that flows into cognitive science is research on AI which historically has been
interested in developing computer models to simulate reasoning. But what is
it that they are simulating? It is one thing to develop programs that model
deduction, quite another thing to write programs which will model argumen-
tation.
It is clear that how we answer this question hinges on careful delineation of
the term reasoning. However, here we encounter the problem that there is no
uniform understanding of this term. As we shall see, many have adopted the
view that reasoning can more or less be equated with inferring; others see
reasoning as a change in view. I will argue here for a third view. This same
kind of divergence attends the term argument. To illustrate, the following
specimen is classified by some logicians as an argument:
(S1) If the moon is not made of green cheese, then arithmetic is com-
plete.
The moon is not made of green cheese.
So arithmetic is complete.
So also:
(S2) The sky is blue.
Grass is green.
Therefore, it is not the case that tigers are carnivorous.1
Other logicians, myself included, demur.
Probably some would include as arguments the following:
1 See Lambert and Ulrich (1980) and Govier (1987:182).

(S3) 99% of ravens are black.

That bird is a raven.
Probably,
that bird is black.
(S4) Tweety is a bird
Probably,
Tweety can fly.
Now I although I do not challenge the view that (S1) through (S4) are speci-
mens of reasoning (though the credentials of (S1) and (S2) are somewhat
suspect), my sense is that it is mistaken and unhelpful to classify them as
arguments.
The issues I want to address in this paper is the nature of reasoning, and the
relationship between reasoning and argument. If we are going to get a clearer
read on these issues, we are not going to be able to do without some awareness
of the multiple meanings associated with these words. For, on the one hand,
there are legitimate senses of both reasoning and argument in which they are
virtually synonymous. Thus, one speaks of a line of reasoning and more or less
means an argument. On the other hand, there are uses of both terms according
to which they have quite different meanings. Such divergent usage means that
any position on this issue should be taken as stipulating how these terms can
best be used and should provide appropriate rationale for such stipulations.
But there is more, for these terms not only have a close association with
one another; they have close associations with other terms, so close that it does
not seem feasible to give any significant account of one of them in isolation
from the remainder. I refer to this situation as The Network Problem. Let me
now turn to it.
II. The Network Problem
Once we begin to focus our attention on reasoning, we open up another a host

of questions such as: What is the connection between reasoning and rationali-
ty? What is the relationship between reasoning and intelligence? reasoning and
knowledge? reasoning and thinking? These terms seem to form a family or
network in the sense that achieving clarity about any one of them appears to
be dependent, to some degree, on seeing it in relationship to the others.1
1 The situation here is not unlike the way that Grice and Strawson (1976) construe the situation of
terms like “analytic” and “synthetic” which they see as part of a larger family (or network) which
includes “a priori,” “contingent” etc.
16 Ralph H. Johnson
reasoning
knowledge rationality
intelligence thinking
argument
THE NETWORK PROBLEM-1.1
Much work has been done on the theory of knowledge, the theory of ra-
tionality, the theory of intelligence, but very little work has been done on the
theory of reasoning, though there are indications of its emergence beginning
with Finocchiaro (1984). The term argument itself is closely connected to a
family of terms:
argument
inference implication
reasoning
NETWORK PROBLEM-2
In my view, inquiry into both forms of The Network Problem is the task of
the theory of reasoning. I shall not attempt to solve any of the three forms of
that problem in this paper. I will however be setting forth my position regar-
ding third form of the problem. Hence this paper is intended as a contribution
to the theory of reasoning. In the next section, I review the positions of some
prominent logicians on the nature of reasoning.2
III. The Nature of Reasoning: Some Positions
A. Introduction
The key issue for the theory of reasoning is to get clear on the nature of reaso-
ning itself. In this section I examine the views of Finocchiaro, Walton and
Harman indicate why I find them inadequate. That will pave the way for my
attempt in Section IV to provide a more adequate account.
1 See my (1996) where I identify two forms of The Network Problem, the first of which has to do
with critical thinking and problem solving. It is the second form of that problem that we are
concerned with here, but to avoid confusion, I shall refer to it here as The Network Problem-1, to
distinguish it from yet another form of the problem, to be introduced shortly.
2 I shall sometimes phrase the problem in the material mode – as the nature of reasoning, and
sometimes in the formal mode – the meaning of reasoning (“reasoning”). (For the distinction
between the formal and material modes, see Carnap (1951).)
B. Finocchiaro on reasoning
In “Informal Logic and the Theory of Reasoning” (1984), Finocchiaro called
for the development of the theory of reasoning and in the course of that
provides his own conception of reasoning. The focus of Finocchiaro’s inquiry
is reasoning rather than argumentation. This is a deliberate move, made so as
to include “besides the study of arguments, such activities as problem-solving,
decision-making, persuasion and explaining which cannot be equated with
argumentation but which may involve reasoning in an essential way” (3). Here
Finocchiaro endorses the view that reasoning and argument are different, and
seems to be advocating for a broad conception of reasoning – one which makes
evident the connection between the theory of reasoning and The Network
Problem-2 above. Finocchiaro now adds a series of clarifying comments about
his definition of the theory of reasoning, after which he anticipates and re-
sponds to four objections, each of which challenges the philosophical legitima-
cy of the theory of reasoning. It is the first objection – the one he calls “the
most fundamental”– that is crucial for our purposes. It is that no such subject
matter (as reasoning) really exists. This objection, says Finocchiaro, is based on
the view that reasoning is “an epiphenomenal illusion deriving from using a
general label to refer to a number of disparate activities” (4). From the vantage
point of such a conception, it would seem to follow that “a theory of reaso-
ning per se, as distinct from theorizing about particular instances or types or
fields of reasoning, makes no more sense that a theory of success in general”
(4).1
Finocchiaro’s rejects this view of reasoning and proposes, as a counter-
suggestion, that “the essential feature of all reasoning is the interrelating of
individual thoughts in such a way that some follow from others and that the
normal linguistic expression of such interrelated thinking involves the use of
particles like ‘because,’ ‘therefore’” (4). Finocchiaro refers to this as a “minimal
conception,” by which I think he means that he is aiming just to get at the
basics. Even so, my chief reservation about this way of conceiving reasoning is
that makes inference to be the core of reasoning. For “the interrelation of
thoughts so that some follow from others” could serve equally well as a de-
finition of inference, and the particles he mentions typically attend inferences
(and perhaps arguments). I refer to this view as inferentialism and discuss it
later. That then is the first problem with his minimal conception.
I see two other problems. First, what happens when the reasoner thinks
that one thought follows from another but it does not follow, as happens
1 Finocchiaro is here rejecting an approach to conceptualizing reasoning that approximates the

one I shall later present.
18 Ralph H. Johnson
when the reasoner produces what is called a non-sequitur? Is such a specimen

to be categorized as reasoning? I think it should be, but it would not qualify
under Finocchiaro’s definition because the one thought does not follow from
the other, though the reasoner thinks it does. As stated, then, the definition is
too narrow because it would rule out what seem like clear cases of (bad)
reasoning. Finocchiaro can modify his concept so that it read s”the interrelati-
on of thoughts some of which are believed to follow from others.” Still he
faces the question of what it is for one thought to follow from another? That
is the second problem. If one starts down this road, one is immediately going
to think of deduction, where one proposition is deduced from another. But
this would be much too limited as a definition of inference, never mind reaso-
ning. Thus, he would have to broaden the notion of inference to allow for a
“weaker” kind of “following from.”1 Even if he makes this adjustment, he
remains within the grips of the idea that reasoning is essentially inference. Not
only are we then saddled with the problem of defining inference, which is task
that must be embraced at some point, but the conception of reasoning has
been narrowed unduly. I conclude that Finocchiaro’s minimal conception
won’t satisfy our demands and turn to another account, that of Walton.
C. Walton on reasoning and argument (1990)

In “What is Reasoning? What is an Argument?” (1990), Walton raises a num-
ber of important questions pertinent to our inquiry.
He writes:
It is problematic to see how reasoning is related to argument. Are reasoning and
argument essentially the same thing? Or is one a proper subpart of the other? Or
can you have reasoning that is not in argument? It seems hard to know where to
begin replying to these questions. (400)
Walton begins by citing Govier’s way of demarcating between argument and
reasoning:
An argument is a publicly expressed tool of persuasion. Typically it takes thin-
king to construct an argument. Reasoning is distinguished from arguing along
these lines: reasoning is what you may do before you can argue, and your argu-
ment expresses some of your best reasoning. But much reasoning is done before
and outside the context of argument. (401)
I have some difficulties with this view. First, Govier seems to equate thinking
and reasoning, whereas I will later argue that it is worthwhile to distinguish
1 In private conversation, Finocchiaro has said he would now change his definition and use the
notion of “depend upon” rather than “follow from” in which case my objections here will lose
their force.
them. Second, in Govier’s view, reasoning generally takes place before and
outside of the context of argument, which suggests that argument is not for
her a kind of reasoning. This raises the question of exactly what Govier means
by argument. If argument is not itself reasoning, then what is it, and how does
it stand related to argument? Once again we are dealing with the second form
of The Network Problem, and we see how much hinges on how we construe
the fundamental terms here – argument and reasoning.
Walton begins his own account by citing the definition given by Angeles in
his Dictionary of Philosophy (1981), according to whom there are three senses
of reasoning, the first of which is “the process of inferring conclusions from
statements.” Walton says: “The first definition is positive and does seem to
capture the basic idea behind reasoning” (401). (The second and third senses
do not enter into discussion and so I omit any discussion of them here.) Thus
Walton also seems to endorse an inferential approach to reasoning. He adds a
way of distinguishing the two: inferring leads from premises to conclusion, but
we may reason not only to propositions but also to actions. Walton likewise
agrees with Angeles that reasoning can be used for a variety of purposes. (Note
that argument and reasoning are here conceived as non-equivalent terms.)
Following further reflection on Angeles’s definition, Walton arrives at the
view that “Reasoning is the making or granting of assumptions called premises
(starting points) and the process of moving toward conclusions (end points)
from these assumptions by means of warrants” (403). Walton perhaps thinks
that this statement lines up with his earlier statement, but I detect some non-
negligible differences. Specifically, while this second version remains inferen-
tialist, it has begun to move the definition of reasoning in the direction of
argument. Indeed the terminology Walton employs – premise, warrant, con-
clusion – is the language of argument, especially reminiscent of Toulmin’s
approach (1958, 1979) to the structure of argument (or reasoning).
Walton offers us two somewhat different definitions of reasoning, one of
which seems clearly inferentialist (i.e., construes reasoning as inferring), while
the second leads in the direction of construing reasoning as argument. Walt-
on’s account thereby reflects the tension and the ambiguity I am confronting
in this paper. What is needed is an account of reasoning that will at least in-
itially set it apart from inferring and arguing. But before I turn to that matter,
I want to pause to say something about the position I have labeled inferentia-
lism.
20 Ralph H. Johnson
D. About Inferentialism
I have claimed that both Finocchiaro and Walton adopts an inferentialist view
of reasoning. It is time to say more about just what that means. By “inferen-
tialism,” I mean the view that reasoning is essentially inference (or else reduci-
ble to it). This the bias in favor of inference (particularly deductive inference)
as the paradigm is evident in much of the empirical-psychological research. A
classic study, Human Reasoning (1978) by Revlin and Mayer, contains no
explicit definition of reasoning, but appears to operate on the assumption that
reasoning is essentially inference. In the Introduction we read:
The present volume contains a series of manuscripts concerning current research
methods, results, and theories on human reasoning. While this collection does not
represent an exhaustive survey of research on human inference ... (7)
Note how readily and naturally Revlin and Mayer move from discussing
reasoning to inference, as though they were pretty much the same thing. The
papers in Human Reasoning almost all take syllogistic reasoning to be the
most representative form of reasoning. So although the title of the book refers
to reasoning, all of the articles deal with inference and the vast majority with
deductive inference. This is also true Wason’s landmark study The Psychology
of Reasoning (1972) which with no less justice might have been titled The
Psychology of Deductive Inference. Not just inference, then, but deductive
inference has enjoyed pride of place when it comes to theorizing about reaso-
ning.
There are many reasons for this focus. Aristotle’s syllogistic laid the foun-
dations for logic – for what we might call the systematic normative study of
reasoning. The term “syllogismos” in Greek could equally be rendered by the
term “reasoning.” Even though Aristotle’s own work on reasoning far out-
strips his work on syllogism in Prior and Posterior Analytics, still many retain
the belief that deductive inference is either synonymous with reasoning or else
certainly among the most central forms. In psychological research on reaso-
ning, the tendency to focus on inference is perhaps explained both by histori-
cal influence,1 as well as by the fact that clear cut answers exist to the question
what follows from what, so that empirical studies can be undertaken – so-
mething less easily accomplished with argumentation and explanation. Furt-
her, there is no denying that inference is prominent in many of our cognitive
endeavors.
The last 20 years have witnessed tremendous advances in cognitive psycho-
logy and the study of reasoning. There has been a marked movement to broa-
den the scope of psychological research to include other forms of inference.
1 Piaget believed that “reasoning is nothing more than the propositional calculus itself” (12).
Here one thinks of Human Inference (1980) by Nisbett and Ross and also
Judgement under Uncertainty (1982) by Kahneman, Slovic and Tversky. Yet
the fixation on inference abides. In his book, Mental Models (1984), John-
son-Laird sets forth “seven goals for evaluating any theory of reasoning.” A
careful read will show that in his formulation of these goals, Johnson-Laird is
thinking not so much of reasoning as inference.1
The thought here is that we can benefit from a wider and different view.
One author who has offered a different approach is Harman. I turn next to a
somewhat brief consideration of his views.
E. Harman on Reasoning
In Change in View (1986), Harman develops the view that reasoning can be
characterized by the idea of “change in view” (belief, intent). However, I
believe that this way of characterizing reasoning is both too broad and too
narrow. It is too broad because unless one specifies that the change in view
must be the result of reasoning (in which case the definition becomes circular),
there are many instances where people have changed their view but have not
reasoned. For instance, over time, someone’s view may change with any
explicit attempt at modification, simply by the press of circumstances. Someo-
ne might say: “I used to think that .... but now I don’t.” And when asked
“What changed your mind?” he or she might reply: “I’m not sure.” In short
sometimes our views change because we have reasoned and thought about
them; and sometimes they change without direction attention or focus. Har-
man’s conception is too narrow because although reasoning often does produ-
ce a change in view, it can just as often produce a reinforcement of the view.
That is, I may review the arguments offered for a position and decide that
those arguments do indeed support that position. I have engaged in reasoning
but my view has not changed. Harman’s approach, though a welcome break
from the tradition of inferentialism, seems open to serious criticism.
1 The criteria given by Johnson-Laird are as follows (65-66):

1. A descriptively adequate theory must account for the evaluation of conclusions, the relative
difficulty of different inferences, and the systematic errors and biases that occur in drawing
spontaneous conclusions.
2. The theory should explain he differences in inferential ability from one individual to another.
3. The theory should be extensible in a natural way to related varieties of inference rather than
apply solely to a narrow class of deductions.
4. The theory should explain how children acquire the ability to make valid inferences.
5. The theory must allow that people are capable of making valid inferences, that is, they are
potentially rational.
6. The theory should shed some light on why formal logic was invented and how it was developed.
7. The theory should ideally have application to the teaching of reasoning skills.
22 Ralph H. Johnson
Without wishing to claim completeness for this inventory of how phi-

losophers and logicians have construed reasoning, we have seen that there are
problems with the inferentialist accounts offered by Finocchiaro and Walton,
and with Harman’s alternative. In the next section, I present my own account.
IV. Reasoning and Argument
A. Reasoning
The position I advocate attempts to adhere to the sense of the word common
in everyday discourse, where reasoning would likely mean something like
“figuring it out.” Taken this way, reasoning would be almost the same as
thinking. However, I believe there is something to be gained by distinguishing
the two.1 I propose that we take thinking to be the having and/or processing of
mental representations. Thus, daydreaming would be a case of thinking but
not reasoning, because it lacks the direction associated with reasoning. Similar-
ly, remembering a joke would be thinking, but not reasoning. Thinking thus
would be the genus of which reasoning is a species.
To understand reasoning and set it apart from other modes of thinking, I
would begin with the nominal conception already stated: reasoning is the
seeking and/or the giving of reasons. One who seeks reasons is reasoning, as is
one who gives reasons. However, we humans do not simply seek and have and
give reasons. We seek reasons for a reason, and we give reasons for a reason.
Thus, what distinguishes thinking (the having of representations – like day-
dreaming) from reasoning (e.g. investigating the role of daydreaming in our
emotional lives) is that in the latter case the representations are controlled by
purpose. Reasoning then is thinking directed by purpose.2 In line with this
admittedly broad specification, it seems to me that the following specimens all
qualify as instances of reasoning.
1. When I find a puddle of water in my basement and seek an explanation
for it, I am looking for a reason. I am reasoning. When I then give that
explanation to my wife (“it’s just water from the pipes because they are
sweating”), I am giving her my reasons. I am reasoning; so explaining is
one type of reasoning.
2. When I predict the outcome of the Super Bowl game by figuring out
which team has the best chance of winning, I am finding reasons. I am
reasoning; so predicting is another type of reasoning.
1 Max Black (1945) also makes this distinction, though not in quite the same way.
2 I am reminded here of L. Susan Stebbing’s phrase – “thinking to some purpose.”
3. When I make a controversial assertion and back it up with reasons, I

am giving reasons; I am reasoning; so asserting is another type of reaso-
ning.
4. When I develop an argument to support my view that Canada should
not commit forces to an invasion of Iraq, by considering possible objec-
tions and replying to them, I am presenting reasons to support my positi-
on. I am reasoning; so arguing is another type of reasoning.
5. When I offer a definition or clarification of the term “pornography”
(against the background of the current debate), I am finding reasons to
think of “pornography” in a specific way. I am reasoning; so defining
and clarifying are types of reasoning.
6. When I infer that we must do our shopping on Friday because it has to
be either Friday or Saturday but Saturday is out, I am reasoning; so
deducing is another type of reasoning. And so on.
All of these are instances of purposive thought in which the finding and giving
of reasons plays a crucial role. All of these are instances of reasoning. Now
either all of these operations are essentially inferential in character; or, there is
some operation – reasoning – that is individuated in each of them (this seems
to be Angeles view); or else there is some third possibility, which is what I
hold to be the case. Even if one wishes to claim that inferring occupies pride of
place in one’s account of reasoning; i.e., that defining, persuading, predicting,
arguing are essentially inferential in character, that will have to be shown by
some form of argumentation. I am doubtful that such an argument can be
made successfully, but the possibility must be admitted. The second alternative
– to hold that reasoning is something over and above these instances in which
it is embodied – also needs to be shown. Anyone minded to give such an
account owes us an explanation of what reasoning is such that it is different
from all of these, and yet instantiated in each of them.
I here defend the third alternative, which is close to what Finocchiaro
termed the “epiphenomenal” position. Even though all reasoning may be
described as thinking under the governance of purpose, there is no one specific
mental activity that can properly be identified with reasoning. Semantically,
then, “reasoning” is a generic term. By that I mean that there is no one activity
which just is reasoning, full-stop. Rather reasoning is the “name” for a host of
different activities.1 If this recommendation makes sense, then we can see why
it is a mistake to identify reasoning with inferring, or indeed with any of its
1 This is not unlike Wittgenstein’s position on language which, in Zettel, he talks about “langua-
ge” as “a name for a collection” (#322).
24 Ralph H. Johnson
species. That would be a mistake comparable to identifying fruits with apples.

It may be that apples are the most common fruit and the one that immediately
comes to mind whenever we speak of fruits. But it remains a mistake to identi-
fy the genus with an of its species. Just as there is nothing which is fruit, over
and above its various species, so too there is nothing that is reasoning, over and
above its various instantiations.
From my account it follows that while all inferring is reasoning, the reverse
is not true. There are forms of reasoning which it does not make sense to
identify as inferring.
Pre-eminent among them is argument. I cannot here present a complete
account of argument, but I can at least attempt to show how I would characte-
rize so as to distinguish it from inference and reasoning.1
B. Argument
It is clear that argument, or better arguing, is one type of reasoning. My imme-
diate concern now is to show that argument is not the same as inference. This
needs to be shown because in the 20th century there is a strong tendency to
merge the two. As evidence of this let me cite the definition provided by Copi
in his Introduction to Logic – one of the oldest and best selling logic textbooks
in North America.
An argument, in the logician’s sense, is any group of propositions of which one
is claimed to follow from the others which are regarded as providing evidence for
the truth of that one. (7, 2nd ed., 1961)
(In later editions, “evidence” is replaced by “support or grounds” (1986, 7th
ed., p. 6).) A few lines later, Copi adds: “An argument is not a mere collection
of propositions, but has a structure” p. 7). Here we see the same essential
structure and relationship as has been used to characterize inference, and so
the impression is created that argument and inference mean pretty much the
same thing. To further complicate matters, the notion of “follows from” is
ambiguous as between implication and inference, and so we are right smack in
the middle of the third form of The Network Problem.
So the three terms – argument, inference and implication – have become
enmeshed. I cannot here undertake a full attempt to disentangle but let me at
least indicate the line of reasoning that I would pursue.
To begin, I cite an example of what I take to be clear cut instance of each
Implication:
Boston is a city and Boston is in the United States.
1 For a better treatment of both argument and the Network Problem, see my Manifest Rationality
(forthcoming from Lawrence Erlbaum).
Therefore Boston is in the United States.

The first statement implies (entails) the second, and this is so quite apart, from
whether any human reasoner ever reasoned in just this way. (S1) then is an
instance of an implication.
Inference:
From the fact that the dog did not bark, it follows that the intruder was
known to the dog.
(Based on Sherlock Holmes)
Here the second thought (proposition) is in some sense derived from the first
(whether it implies it or not). The reasoner has inferred the second from the
first. Argument:
The President should be impeached; for he lied to the grand jury; he lied
in a deposition in the Paula Jones case; and he has obstructed justice.
These are certainly impeachable offenses. The objection that he should
not be impeached because the American people don’t support impeach-
ment makes two unwise assumptions: first, that the polls are accurate;
second, that representatives are obliged to vote as their constituents say.
This is an argument. The reasoner is attempting to persuade someone else of
the truth of a particular claim and the doing of this requires evidence or rea-
sons. Clearly the reasoner thinks that the reasons cited are good reasons for
the claim and he wants others to think so also. This may be true even when (a)
the reasoner has not himself made any inferences, and (b) the statements do
not take the form of implicatures.
Notice as well that the reasoner has produced not only a line of reasoning
that leads to the conclusion but has as well attempted to defuse a possible
objection.
Let me now return to the examples with which I began.
(S1) If the moon is not made of green cheese, then arithmetic is complete.
The moon is not made of green cheese.
So arithmetic is complete.
No one who is serious about arguing that arithmetic is complete would produ-
ce such a specimen of reasoning to make his point or to attempt to persuade
others. I suspect that no one has ever drawn such an inference. In fact, such
specimens as (S1) and do not originate in the real life practice of argumentation
but rather are artificial specimens constructed to illustrate a point about the
propositional calculus – here that the validity of an implication does not
establish the truth of the conclusion. (It also illustrates the relationship that
26 Ralph H. Johnson
some have called material implication.1) (S3) and (S4) might well be taken as
examples of inference. However, they do not seem to be example of argu-
ments, because their conclusions are not the sorts of controversial matters
which spark arguments.
What about (S2)? It is a strange specimen, no doubt. It does not appear to
be an implication, nor an inference, nor yet an argument (though it was been
labeled “argument” by the authors who produced it (Lambert & Ulrich, p. 24).
Is it then reasoning?
I don’t know what to say about (S2). On the one hand, I want to say that
it is just atrocious reasoning, which means, of course, that it is reasoning. On
the other hand, (S2) does not much resemble any actual reasoning, how any
human reasoner would think. It seems like just a series of statements randomly
linked by “therefore,” as if someone generated the following sequence of
words “windows I furiously spring door at” and asked if these constitute a
sentence. Whatever we say about its status, (S2) is a queer specimen. Though
we want our theory of reasoning to deliver a verdict on its status, it is not the
sort of example we wish our theory to illuminate. It may help if we look at
how (S2) came into being. It was invented by Lambert and Ulrich to make a
series of points about the difficulties of providing a formal account of informal
fallacy. I would say, then, that while it is not itself an example of reasoning
(since it does not satisfy the definition produced earlier), it was produced in
order to make a point about one particular kind of reasoning. Perhaps then it
is best relegated to the category of “thinking” rather than “reasoning.”
To summarize this section: By the term argument then, I understand an
intellectual product, essentially public and social in character, which seeks to
persuade rationally. By inference I understand a movement (of the mind) from
one item (usually a thought represented in a proposition) to another, and
where the former serves as the basis for and leads to the latter. An inference
may come to be the conclusion of an argument, but it need not be. By im-
plication I understand a relationship between two proposition, such that one
of them follows from the other(s). An implication may well be nested in an
argument but it need not be. An inference can give rise to an argument but is
itself a quite different type of reasoning. If I am right, argument cannot be
reduced to a concatenation of implicatures or inferences.
1 I here bypass the whole debate about whether material implication is in fact a kind of im-
plication. For discussion of this, see Anderson and Belnap (1975).
V. Conclusion
In this paper, I have attempted to set forth a position on the nature of reaso-
ning that differs from the dominant view in the 20th century – inferentialism.
I have also attempted to provide an account of how reasoning and argument
are related. That attempt has involved me in an effort to distinguish argument
from inference and implication.
For too long the intellectual culture of the 20th century (influenced as it
was by the Positivist Ideal) has tended to merge the identity of these three. I
hope to have given some basis for distinguishing argument from both im-
plication and inference. All three are important; but they have different condi-
tions and confer different benefits and play importantly different and indeed
complementary in the intellectual life of both the individual and the culture.
Argument, inference and implication are all types of reasoning. Their full and
proper definition and differentiation is the subject of what I call the theory of
reasoning. This paper has been an effort to make a contribution to that inqui-
ry. I hope that my attempt to answer these questions, limited and controversial
though it is, will suggest to other just how important this inquiry is.
References
Anderson, A.R. and Nuel D. Belnap Jr. 1975. Entailment: The Logic of Relevance. Princeton:
Princeton University Press.
Black, Max. 1945. Critical Thinking: An Introduction to Logic and Scientific Method. New York:
Prentice-Hall. (2nd ed, 1952).
Carnap, Rudolf. 1956. “Empiricism, Semantics and Ontology”. In: Meaning and Necessity.
Chicago: University of Chocago Press.
Copi, Irving. 1961. An Introduction to Logic. New York: Macmillan. (7th ed., 1986.)
Finocchiaro, Maurice. 1984. “Informal Logic and the Theory of Reasoning”. Informal Logic 6, 3-
8.
Govier, Trudy. 1987. Problems in Argument Analysis and Evaluation. Dordrecht: Foris.
Grice, H. P. and P. F. Strawson. (1956). “In Defense of a Dogma”. In: Rosenberg, J.F. and Travis,
C. (Eds.), Readings in the Philosophy of Language, pp. 81-94, Englewood Cliffs, NJ: Prentice-
Hall.
Harman, Gilbert. 1986. Change in View. Cambridge: The MIT Press.
Inhelder, Barbel. 1958. The Growth of Logical Thinking from Childhood to Adolescence. New
York: Basic Books.
Johnson, Ralph H. 1996. “Reasoning, Critical Thinking and The Network Problem.” Chapter 14
of The Rise of Informal Logic.
Johnson, Ralph H. 1996. The Rise of Informal Logic. Newport News, VA: Vale Press.
Johnson, Ralph H. Manifest Rationality: “A Pragmatic Theory of Argument”. Mahwah, NJ:
Lawrence Erlbaum (forthcoming).
Johnson-Laird, P. N. 1983. Mental Models: Towards a Cognitive Science of Language, Inference
and Consciousness, Cambridge,: Harvard University Press.
28 Ralph H. Johnson
Kahneman, Daniel, Paul Slovic and Amos Tversky. 1982. Judgment Under Uncertainty: Heuristics
and Biases, Cambridge, Cambridge University Press.
Lambert, Karl and William Ulrich. 1980. The Nature of Argument. New York: Macmillan.
Nisbett, Richard and Lee Ross, Human Inference: Strategies and Shortcomings of Social Judgment.
1980. Englewood Cliffs, NJ: Prentice-Hall.
Revlin, Russell and Richard E. Mayer, Eds. 1978. Human Reasoning, Washington, D.C.: V. H.
Winston, 1978, pp. 4-5.
Stebbing, L. Susan. 1939. Thinking to Some Purpose. Harmondsworth (Middlesex): Penguin
Books.
Toulmin, Stephen E. 1958. The Uses of Argument. Cambridge: Cambridge University Press.
Toulmin, Stephen E., Richard Reike and Allan Janik. 1984. An Introduction to Reasoning (2nd
ed.). New York: Macmillan. (1st ed, 1979).
Walton, Douglas. 1990. “What is Reasoning? What is an Argument?” The Journal of Philosophy
87, 399-419.
Wason, Peter C. 1972. The Psychology of Reasoning: Structure and Content, Cambridge: The
Harvard University Press.
Wittgenstein, Ludwig. 1967. Zettel. Trans. G.E.M.Anscombe.Oxford: Basil Blackwell.
The Fox and the Hedgehog: On Logic, Argument, and Argumentation Theory 29
LEO GROARKE
The Fox and the Hedgehog: On Logic, Argument, and Argumentation

Theory
Introduction
According to a traditional Aesop’s fable, the fox once asked the hedgehog how
many tricks he knew for outwitting the hunters and their dogs. “Only one”
the hedgehog answered. “I know how to play dead.”
The fox was aghast at this response and proceeded to boast of the thirty-six
tricks he knew. Just as he was finishing, a pack of hounds came over the hill.
Immediately, the hedgehog fell to the ground and played dead. The hounds
came yelping up to him, poked him with their noses, and ran after the fox. The
hunters ignored the hedgehog as they went rushing by.
Thirty-six tricks later, an exhausted fox finally managed to escape the
hounds. As it was walking back to its lair, it met the hedgehog.
“How,” it said, “did you ever manage to elude the hounds and hunters
with just one trick?”
“I have just one trick,” the hedgehog replied, “but it’s a good one.”
It may seem a long jump from this fable to contemporary argumentation
theory, but we can make the jump if we properly understand the latter’s aims
and its origins.
The goal of argumentation theory might be described as a comprehensive
understanding of argument and persuasion. This encompasses an understan-
ding of argument in scientific and (less so) mathematical contexts but argu-
mentation theorists have been primarily concerned with the study of argumen-
tation in non-formal contexts – in advertising, political commentary, legal
argument, public discourse, advocacy and other forms of everyday discourse.
It is tempting to describe such argumentation as “natural language” argument,
but this characterization is too narrow, for argumentation theorists also study
instances of argument and persuasion which employ non-verbal forms of
communication like visual images, physical movement, and other forms of
“body language” (see, e.g., Birdsell & Groarke 1996 and Gilbert 1997). Within
the last fifteen years, the field has expanded – one might say exploded – to
become an interdisciplinary field of inquiry which is influenced by philoso-
phy, formal and informal logic, traditional and contemporary rhetoric, com-
munications studies, linguistics and cognitive psychology.
30 Leo Groarke
In many ways, argumentation theory is a reaction against the concept of

argument which characterizes traditional formal logic – a concept which takes
as its paradigm argument and proof as they occur in mathematical and scienti-
fic contexts. In reacting against this narrow focus, argumentation theorists
have maintained that logic’s account of argument is too narrow to explain the
broad range of arguments that they study. Toulmin (1958), Mortenson &
Anderson (1970), Perelman (1982), Johnson (1996) and innumerable other
commentators have, therefore, roundly criticized the account of argument
which characterizes traditional logic and formal analyses of argument more
broadly. Though they concede the usefulness of this account in a narrow range
of contexts, they argue that the attempt to understand argumentation more
broadly requires a fundamentally different account of argument.
In the present paper I want to question argumentation theory’s attempt to
formulate conceptions of argument and persuasion which are fundamentally
distinct from the account that is assumed by traditional formal logic. A full
defence of the latter is beyond the scope of the present paper, but I will argue
that there is no clear evidence which shows that the notion of argument em-
bedded in traditional logic cannot account for the complexities of ordinary
argument. For though there are ways in which this notion must be broadened
and relaxed if we are to make sense of ordinary discourse, this can be accom-
plished without countenancing the radical alternatives which have been propo-
sed. In criticizing these alternative accounts of argument I will, in particular,
criticize a tendency to promote a view of argument and persuasion which
distinguishes a great many categories of argument which require different
kinds of analysis. Instead of adopting the diverse notions of argument analysis
this implies, I argue that we will do better to retain the one which is implicit in
traditional logic. Like the hedgehog in the fable, this approach leaves us with
only one trick, but it is a trick which is a good one, and one which promises a
more effective analysis of ordinary argument than the many tricks which tend
to characterize argumentation theory in its present state of development.
Argument Traditionally Conceived
We might describe the “traditional” concept of argument simply. It defines an

argument as a set of reasons (premises) in support of a conclusion. Sound
arguments are arguments with true premises which are deductively valid. A
deductively valid argument is in argument in which it is impossible for the
premises to be true and the conclusion false. It is in this sense that the con-
clusion follows from the premises.
Though they are often misconstrued in argumentation texts (see Groarke

1999), these are familiar notions. In the context of argumentation theory, the
important point is that the notion of argument which logic has assumed is
easily detached from formal logic, and is not necessarily tied to the attempt to
provide a formal account of argument. Many commentators assume the latter,
but this is an elementary mistake. Arguments which are sound and/or valid
need not be arguments which are proved so in some standard formal logic. In
ordinary contexts, it is often redundant to back such arguments with formal
proofs. In other contexts, proofs may be difficult or impossible because there
is no formal system which adequately captures the features of ordinary infe-
rence an argument depends on. The attempt to create formal systems which
better capture such complexities is one of the (many) driving forces which
motivates the development of new formal logics.
In the present paper, my concern is the concept of argument traditionally
assumed by formal logic rather than the attempt to formalize it. In detaching
this notion from formal logic and applying it to the kinds of contexts studied
by argumentation theory, it must be granted that ordinary argument and
persuasion are not governed by the strict standards that characterize argument
and proof in formal contexts like science, logic and mathematics. Even in cases
of good reasoning, arguments in ordinary discourse usually rely on premises
which are acceptable rather than true or certain. This is an inevitable conse-
quence of the fact that the claims (and values) which are the basis of public
discussion, political debate, advertising, conflict resolution, etc. are characteri-
zed by contentiousness rather than consensus. At least from a practical point
of view, a definitive treatment of relevant facts and values is frequently impos-
sible. It follows that argumentation in ordinary discourse must often depend
on premises which are acceptable, probable, plausible or likely rather than true
or certain. As rhetoricians emphasize, this may also mean that premises may be
acceptable or not depending on the audience one addresses. Or as pragma-
dialectics puts it, attempts to resolve differences of opinion often must be
conducted in a way that reflects the beliefs and the attitudes of the specific
parties to the dispute.
The role that premise acceptability must play in argumentation contexts
suggests that soundness sets too crude a standard when we judge ordinary
arguments. In the frequent contexts in which we cannot clearly demarcate true
and false premises, we are in no position to judge soundness or unsoundness.
While this means that we must look for some other paradigm of good argu-
ment, it does not follow that the basic idea behind soundness is mistaken, but
only that the assessment of ordinary argumentation requires an analogue
which understands a good argument as a valid argument with acceptable rather
32 Leo Groarke
than true or certain premises. This is an important change to the traditional

account of good argument – most significantly because it allows the same
argument to be good in one context and not another – but there is an obvious
sense in which it modifies soundness rather than rejects it.
Another way in which ordinary argument and persuasion departs from the
ideals which motivate traditional formal logic is in their reliance on implicit
premises and assumptions. Outside of mathematics, logic, science, and legal
and scholarly endeavors, attempts to fully enunciate the structure and content
of an argument are rare. Instead, arguers rely on a host of implicit under-
standings that play a central role in the argumentative exchange. Arguments
are forwarded without explicit premise and conclusion indicators. Enthyme-
matic constructions leave crucial premises unstated and rhetorical questions
are used to obliquely state conclusions. Methodological understandings are
assumed rather than elaborated, loaded terms are used to make implicit
judgments, stylistic goals (not a concern for content or explicitness) determine
the form of many claims, and sarcasm, insinuation, allusion and irony abound.
In some cases, arguments may be entirely implicit because they are not forwar-
ded verbally, but visually or through other non-verbal forms of communicati-
on. In many ways, pragma-dialectics sums up these aspects of argumentation
when it points out that argumentative speech acts are often implicit and indi-
rect (Van Eemeren & Grootendorst 1992, 44-72).
One might easily contrast the implicitness in ordinary argumentation with
the extreme explicitness which tends to characterize the arguments tradi-
tionally emphasized by formal logic. But it is a mistake to think that this
implicitness requires a rejection of the conception of argument which logic
traditionally assumes. It implies only that one must analyze ordinary argument
and persuasion in a way that recognizes their frequently implicit nature and
renders it explicit. The goals that this implies are already assumed in virtually
all variants of argumentation theory, which maintain that argument analysis
must incorporate some attempt to recognize implicit arguments, premises and
assumptions.
Even non-verbal arguments can be accommodated in this way. It is in this
regard worth noting that traditional logic recognizes that propositions may be
expressed in a variety of ways. This recognition is easily expanded to include
non-verbal forms of communication. This expansion requires only that we
recognize that a proposition may be expressed in ways that do not use langua-
ge in the ordinary sense.1 To take a pertinent example, the proposition that
someone is lying about some crime that they committed can be expressed with
1 I am indebted to Lorne Falkenstein for this point.

a variety of sentences (“He’s lying.” “Don’t believe him.” “That’s a tall tale if
I ever heard one.” etc.) but also in non-verbal ways which may include facial
expressions (the raising of one’s eyebrows accompanied by a look of exaspera-
tion) or a caricature which portrays the person accused of lying as a Pinnochio
look alike with an absurdly extended nose. So long as we recognize that non-
verbal claims of this sort have verbal counterparts we can treat them as propo-
sitions in a very ordinary way, and can understand arguments which contain
them as arguments in the traditional sense.
It follows that the implicit character of much ordinary argumentation
does not show that it is impossible to formulate an account of ordinary argu-
ment and persuasion which is based on the account of argument formal logic
traditionally assumes. If this account is to be challenged by argumentation
theory, it must do more than point out the role that implicitness and accepta-
bility play in ordinary reasoning. To see that the challenges which it has pro-
posed are problematic or at least open to debate, we need to consider some of
the competing conceptions of argument which argumentation theorists have
suggested and the questions which they raise.
Argument in Argumentation Theory
Though many argumentation theorists have criticized the notion of argument

which is embedded in traditional formal logic, most have conceded that it is a
powerful tool in the analysis of some kinds of argument. Argumentation
theorists who argue that it is too narrow to account for ordinary argumentati-
on therefore grant that the traditional model is useful in the context of logic
and mathematics, but maintain that it cannot be usefully extended to other
contexts. As a result, argumentation theory frequently endorses an account of
argument which distinguishes between argumentation which does and does
not conform to argument as it is traditionally understood, and proposes a new
model of argument in the latter contexts. When Johnson (1996) criticizes the
notion of argument which has been assumed by formal deductive logic
(“FDL”), he therefore writes “that FDL had in mind one important subset of
arguments, but the realm of argumentation was much broader” (90).
Judged from this perspective, the development of argumentation theory has
been the development of a typology of argument which distinguishes different
kinds of argumentation which are assessed by different kinds of standards. A
small subset of the arguments which are studied by argumentation theorists
are understood in terms of logic’s traditional model of argument. Other kinds
of argument are understood and assessed in ways that employ other models of
good reasoning. Deductive arguments are, for example, understood in the
34 Leo Groarke
traditional way, while inductive arguments are understood and assessed in

other ways. There is much that might be said about this particular distinction
(see below), but it this is only the beginning of the story, for most argumenta-
tion theorists have argued that other distinctions are needed if we are to ade-
quately classify the different kinds of argument and persuasion that characteri-
ze ordinary discourse. This trend is already implicit in Govier (1987), who
criticizes the inductive/deductive distinction on the grounds that it fails to
recognize fundamental distinctions between different kinds of non-deductive
arguments. As she puts it, “The great divide between deductive and inductive
arguments is spurious and theoretically dangerous, because it makes it too easy
to ignore the many nondeductive arguments which are not classically inducti-
ve.” (1987, 53)
The different kinds of argument which have been distinguished by theorists
pursuing this line of reasoning are too numerous to be catalogued here, but
some examples can be illustrative. In her criticism of the deductive/inductive
distinction, Govier distinguishes statistical generalizations, consistency argu-
ments by analogy, “conductive” arguments, and non-conclusive philosophical
arguments (1987, 50-51). Authors like Perelman (1982) distinguish between
arguments which are dialectical and those which are not, maintaining that the
traditional account of argument cannot account for the former. Johnson (1996)
distinguishes between inference, which is properly studied by formal logic,
and “an argument in the complete sense,” which “can only develop against the
background of heterogeneity of point of view and of other arguments” (1996,
94). And Walton distinguishes between different kinds of dialogues which are
characterized by different sets of assumptions and procedures (see, e.g., Wla-
ton & Krabbe 1995; Walton 1998).
In its most radical form, argumentation theory proposes kinds of argument
which are notable precisely because they constitute a radical rejection of the
model which logic has traditionally assumed. Levi (1995) has, for example,
argued that it is a mistake to think that arguments in ordinary discourse must
contain premises and conclusions. Gilbert (1997) maintains that “the logical
mode” is viable in restricted contexts, but argues that we must recognize three
alternate modes if we are to understand the full range of ordinary argumentati-
on and persuasion. The “emotional” mode legitimates appeals to emotions, the
“visceral” mode is “primarily physical and can range from a touch to classical
nonverbal communication, i.e., body language, to force” (84), and the “kisce-
ral” mode relies on “the intuitive, the imaginative, the religious, the spiritual,
and the mystical” (86). In attempting to understand and assess argumentation
in terms of these three “alternate” modes, Gilbert argues that we must trans-
cend the linear approach to argument which logic has traditionally assumed.
Another important instance of argumentation theory’s willingness to

countenance forms of argumentation which are radically distinct from
argument as it is traditionally conceived is reflected in a distinction Gilbert has
labeled the “convince/ persuade dichotomy.” As he puts it:
...to convince is to use reason, dialectic and logic, while to persuade is to rely on
emotion, prejudice, and language. This distinction has moral as well as logical
implications insofar as “convincing” has been considered to be a superior
method. “Persuading” appeals to the “baser” components of the human psyche,
namely, the emotions, while ‘convincing’ speaks to the “higher” aspects, namely,
reason. On this view, one who is persuaded may be so for reasons that have little
to do with the value of the arguments or the truth of the premises put forward.
(Gilbert 1997: 4)
Gilbert criticizes the preference for convincing implied in this account, but still
assumes a clear distinction between convincing and persuading which holds
that persuasion must be understood and assessed in a different way than
arguments in the traditional sense.
Johnson and Blair (1994) employ a variant of the convince/persuade
dichotomy when they explain the dynamics of argumentation within
advertising. According to their account, advertising “have the appearance of
argumentation” but this in reality “a facade,” for “most advertising works not
at the rational level but at a deeper level.” It follows that “advertising has a
‘logic’ of its own” which must be distinguished from “the logic of real argu-
ments” (221) on the basis of a proposed distinction between “rational” and
“psychological” persuasion. As Johnson and Blair put it: “[M]any ad-
vertisements have the facade of arguments. They look like premises leading to
a conclusion and like exercises in rational persuasion. In fact, we are persuaded
by the school that holds that advertising is best viewed as psychological
persuasion – an attempt to use psychological strategies to implant the name of
a product in our unconscious minds. Hence criticism of advertising as a form
of argumentation is misconceived. Learning how to decode ads and making
ourselves aware of the strategies that advertisers use is more useful than
looking for fallacies in the arguments.” (225).
One might easily expand this account of argumentation theory by in-
troducing other kinds of argument which are said to contradict the traditional
account of argument. In the present context, it is enough to say that it is
increasingly characterized by such distinctions, and the different methods of
analysis and assessment they imply. If we compare this approach to argumen-
tation to the attempt to analyze all arguments in terms of some form of the
traditional conception of argument, then it will be clearer how the fable of the
fox and the hedgehog might be applied to argumentation theory. For there is
36 Leo Groarke
a sense in which its development has been characterized by the fox’s approach
to problem solving – i.e. by the development of a great variety of tricks for
analyzing ordinary argumentation (I believe that Dutch pragma-dialectics is in
some ways an exception to this rule, but I cannot discuss it in detail here). In
contrast, an approach which attempts to preserve and extend the traditional
conception of argument is hedgehog-like insofar as it is built upon a single
way of understanding arguments, which is proposed for use in every circums-
tance. A definitive comparison of the distinct approaches to argumentation
this implies is impossible in a single article, but in the remainder of the present
one I will sketch some reasons why we should be skeptical of the fox-like
approach which now characterizes argumentation theory – an approach which
is much more problematic than argumentation theorists have assumed.
Kinds of Argument and the Deductive/Inductive Distinction
Argumentation theory’s attempt to distinguish different kinds of argument is

in part problematic because the kinds of argument it distinguishes are not
“natural kinds” in the sense that they are not clearly distinguished within
ordinary reasoning and persuasion. It is argumentation theorists – not ordina-
ry reasoners – who distinguish between deductive, inductive and conductive
arguments; between attempts at convincing and persuading; between “persua-
sion,” “negotiation,” “pedagogical,” “information-conveying,” “cooperative
informing,” and “eristic” dialgoues; and between logical, emotional, visceral
and kisceral arguments. The extent to which such distinctions are removed
from ordinary contexts is itself a matter of concern. Ordinary reasoners do not
distinguish the kinds of argument and persuasion which argumentation theory
proposes as fundamentally different kinds of argument.
Consider the most entrenched distinction within argumentation theory –
the distinction between inductive and deductive arguments. It is said to show
that there are some everyday arguments which cannot be understood in tradi-
tional terms. But accounts of the distinction are in many ways problematic.
Textbook treatments regularly misconstrue the meaning of deductive validity
and suggest that inductive arguments are arguments which are intended induc-
tively rather than deductively (see Groarke 1999a). Even if we ignore problems
that this raises about intentions (which are in an important sense unobserva-
ble) it makes little sense to think that ordinary reasoners intend their argu-
ments deductively or inductively, for most ordinary reasoners are unaware of
the distinction and are, in view of this, in no position to intend their arguments
one way or the other. As Govier puts it “People who argue do (at least im-
plicitly) distinguish conclusions from premises.... But they often do not, even
implicitly, make claims about what sort of connection is supposed to hold

between these premises and their conclusion” (1987, 46).
It follows that the inductive/deductive distinction is a theoretical distincti-
on which is imposed on ordinary argumentation, rather than one which guides
it in practice. And it is a distinction which is very difficult to apply in practice,
for those arguments which are normally classified as inductive arguments can
usually be construed as deductive enthymemes which include implicit premi-
ses which establish a deductive link between the premises and conclusion. The
argument “I know the apples on the table are good because I have had two of
them” is a paradigm example of the kinds of inductive inference that characte-
rize ordinary reasoning. It can be described as a statistical generalization based
on a sampling of two. But it can easily be construed as a deductive enthymeme
which conveys the argument:
Premise: I had two of the apples on the table.
(Unexpressed) Premise: They were good.
(Unexpressed) Premise: They are like the other apples.
Conclusion: The apples on the table are good.
As Gerritsen (1994) and Groarke (1992, 1999) have pointed out, it is in princi-
ple possible to treat all inductive arguments in a similar way. This does not
eliminate the uncertainty which is characteristic of inductive arguments, but it
locates this uncertainty in implicit premises which allow one to treat inductive
arguments as attempts at deductive reasoning. The implicit premises which are
in the process recognized are, as pragma-dialectics has noted, a natural extensi-
on of the treatment of enthymemes and implicit premises which is a necessary
part of any account of ordinary argument. In the present context, the impor-
tant point is that such a strategy shows that one can extend the traditional
conception of argument to the kinds of contexts studied by argumentation
theory without relinquishing its deductive nature.
But even if one grants some version of the deductive/inductive distinction,
it is difficult to justify the plethora of distinctions which now characterize
many variants of argumentation theory. For the distinctions between different
kinds of dialogue; between emotional, visceral, kisceral and logical arguments;
and between other forms of argument are much less clear and well established
than the distinction between deductive and inductive arguments. Though a
discussion of all of these distinctions is impossible in the present context, the
problems that they raise can be illustrated in terms of the distinction between
convincing and persuading. It is of special significance in the present context
given that persuading as it is standardly understood encompasses the kinds of
38 Leo Groarke
argumentation which are standardly assumed to be most removed from argu-

ment as it is traditionally understood.
Convincing and Persuading Reconsidered
One might describe the proposed distinction between convincing and persua-
ding as a distinction between two ways of changing (or reinforcing) belief. The
former is an appeal to reason, which attempts to change or strengthen belief by
weighing evidence in the literal, rational, deliberate way that characterizes
science, mathematics or detective novels. Persuasion is, in contrast, an appeal
to emotion which relies on emotional associations, hidden meanings, humour,
and other psychological ploys. One might question the tenability of the con-
vince/persuade distinction is by questioning its assumption that emotions are
in some important sense not rational. One might do so by developing an
account of emotion that understands it as some kind of cognitive phenomenon
(as does De Sousa 1987). But it is more important, in the context of argumen-
tation theory, to consider whether the persuade/convince distinction clearly
separates two different kinds of ordinary reasoning.
We can see that there are problems with the convince/persuade distinction
even if we restrict ourselves to clear examples of arguments in the traditional
sense, for it is a mistake to think that the emotional dynamics that characterize
persuasion are absent from such arguments when they are used in day to day
discourse. The argument “Homosexuality should be illegal because the Bible
condemns it and whatever it condemns should be illegal.” is a deductive argu-
ment which is a paradigm example of an argument in the traditional sense – its
premises and conclusion are easily identifiable and it is deductively valid. Yet
the use of such an argument in everyday contexts is inevitably affected by
complex, pervasive and powerful emotional resonances (both positive and
negative) about conservative religion and homosexuality. It follows that argu-
ments in the traditional sense cannot be easily separated from emotional
considerations in the way suggested by the convince/persuade distinction.
One might reply that there is a sense in which the emotion which sur-
rounds this example is irrelevant – does not provide legitimate evidence for or
against the conclusion and that it should not play a role in the argument’s
assessment. Two things can be said in response to this suggestion. First, that
much the same could be said about paradigm instances of persuasion, in which
the emotional ploys which are used typically do not provide legitimate eviden-
ce for or against the point (or the behavior) which the persuasion attempts to
establish. Second, it cannot be assumed that emotion has no legitimate role to
play in judging the acceptability or unacceptability of arguments in ordinary
reasoning. An impassioned plea for an end to genocide in Kosovo will inevita-

bly appeal to emotion but it would be strange to say that it does so inap-
propriately. In the contexts which charaterize traditional formal logic – ma-
thematics. logic, and science – emotional distance is a goal, but the same can-
not be said when we deal with contentious practical issues. On the contrary,
most social and political issues have an emotional element which legitimately
comes into play when we judge whether the premises of an ordinary argument
are acceptable or unacceptable, even when the argument is a paradigm instance
of argument in the traditional sense. Much more needs to be said about the
way in which this emotional element should be judged and assessed is in
principle compatible with the account of argument which has already been
proposed.
As examples of traditional argument may legitimately involve emotion,
examples of persuasion are often best understood as appeals to reason and
rationality (and as convincing in this sense). Consider, to take a hypothetical
example, an instance in which fear is the basis of persuasion. Imagine that you
are kidnapped, taken to an apartment, and held captive. When you think the
kidnapper isn’t looking, you reach for the phone intending to call the police.
But your abductor hears you and spins around. He gives you a threatening
look and thrusts his gun in your direction. You stop reaching for the phone.
What has happened in this scenario? In some sense, the kidnapper made
you change your mind about your phone call. But did he rationally convince
you, or did he just persuade you? He has not said anything, so there is a clear
sense in which he does not explicitly forward premises and a conclusion.
Moreover, the situation is a thoroughly emotional one. The kidnapper plays
on your fear of being hurt and it is this fear which directs your decision to
accede to his wishes. One might conclude that this is a paradigm instance of
persuasion and yet there is a very clear sense in which the kidnapper forwar-
ded an argument in the traditional sense which has already been elaborated.
For even though he has said nothing, his actions function as a way of elabora-
ting an argument which might be summarized as follows:
Premise: Stop or I’ll shoot.
Premise: You don’t want me to shoot (more than you want to conti-
nue).
Conclusion: You should stop what you’re doing.
In principle, this argument could have been made verbally, in a variety of
ways. Instead of pointing his gun, the kidnapper might have uttered the rheto-
rical question “Do you want me to shoot you or do you want to continue?”
40 Leo Groarke
Or he might have said “If you don’t stop, I’ll shoot. You don’t want me to
shoot. So you should stop what you’re doing.”
In real life, a kidnapper is unlikely to make such an argument fully explicit
just because this takes time and because there is no need for him to do so.
According to argumentation theory, the argument is conveyed more directly
when it is verbalized, but there is a sense in which this is wrongheaded: an
aggressive pointing of the kidnapper’s gun gets the point across more directly
than a verbal claim. Certainly one is likely to appreciate the force of his argu-
ment more strongly when a gun is thrust in one’s direction. Whatever needs to
be said in this regard, the important point is that the kidnappers thoroughly
emotional attempt at persuasion can easily be understood as an attempt to
forward an argument which conforms to the traditional conception of an
argument. Like our previous examples, this suggests that there is no easy way
to separate convincing and persuasion in ordinary discourse.
Still more importantly, the decision to treat this example of persuasion as
a case of convincing and thus argument has significant theoretical advantages.
For it allows us to assess it in the same way we assess its verbal counterpart.
Considered from this point of view, we can recognize the argument as a va-
riant of disjunctive syllogism and the traditional ad bacculum. It follows that
the example can be understood, not merely as an argument in the traditional
sense, but also as a widely recognized form of argument which can be assessed
accordingly. Though a detailed evaluation is beyond the scope of the present
article, it is worth noting that the argument in question is arguably a good one
which has the characteristics which convince Wreen (1988) and Woods (1995)
that there can be good instances of ad bacculum. In their way a decision to
treat this instance of persuasion as a case of argument in the traditional sense
pays dividends, for it permits us to understand it and evaluate it in standard
ways, using well understood criteria. In contrast, an adherence to the persua-
de/convince dichotomy requires that we understand it in terms of criteria
which have not been clearly and convincingly established.
A full discussion of convincing and persuasion lies beyond the scope of the
present paper, but it can still be said that such examples suggest that the dis-
tinction is not clear in ordinary argumentation and that ordinary instances of
argument and persuasion frequently contradict its attempt to distinguish
appeals to reason and emotion. In view of such considerations, we may do
better to recognize instances of persuasion as attempts to convey arguments in
the traditional sense, albeit attempts that rely on more subtle and more implicit
means. Such an account intuitively corresponds to our ordinary use of terms
like “convince” and “persuade” and still allows an analysis of both argument
and persuasion which recognizes their affinities to each other, and to the
traditional account of argument. It will take time to see how far such an ap-
proach can be taken, though it must at least be said that it has a potential
which most argumentation theorists have failed to appreciate.
Expanding the Realm of Argument
So far, I have criticized argumentation theory on the grounds that it has been
too quick to embrace conceptions of argument which radically reject the
conception of argument embedded in traditional logic. Historically, its en-
thusiasm for such conceptions is a reaction against logic’s preoccupation with
a very narrow range of argument in science, logic and mathematics. As Perel-
man (1982) emphasizes, argumentation in real life is a much broader enterprise
that involves many other contexts. As Johnson (1996) points out, traditional
attempts to teach logic in a deductive, formal way have often emphasized
peculiar examples of argument which have little relevance to ordinary argu-
ment and persuasion. This much must be granted. But these historical facts do
not show that the traditional conception of argument is necessarily tied to the
kinds of contexts which have characterized formal logic. In assuming that it is,
argumentation theorists have overlooked the possibility that the notion of
argument assumed in such contexts can be expanded and extended so that it
encompasses a much broader range of argumentation. We have already seen
that we can move in this direction by detaching this conception of argument
from formal concerns and by modifying it in ways that make room for the
features of ordinary argumentation which most clearly distinguish it from
formal arguments (i.e. implicitness and acceptability).
The extent to which the traditional concept of argument can be extended
beyond mathematical, scientific and logical contexts can be illustrated in the
context of non-verbal arguments. Though some argumentation theorists have
recognized that they play a prominent role in ordinary argument, the examples
of argument which are most prominent in discussions in argumentation theory
are still predominantly verbal. Such examples present a typical argument as a
set of sentences which includes premises that support some conclusion. This is
the model which has been traditionally assumed in logical and mathematical
proofs, though Barwise and Etchmendy (1992) have recently argued for the
possibility of visual deductions. In everyday contexts, it can similarly be said
that visual elements play a largely unappreciated role in the argumentation that
characterizes political debate, advertising, and so on.
The kidnapping example we considered earlier already shows that non-
verbal argumentation can sometimes be accounted for in a way that uses rather
than rejects traditional notions of argument, for this is a case in which the
42 Leo Groarke
kidnapper presents his argument in a non-verbal way. A famous example

which also illustrates this basic point is found in the traditional Life of Aesop
(Daly 1961). According to its biography of the famous fabulist, he was born an
ugly mute slave and received the gift of speech (logos) in return for a pious
favor to a messenger of Isis. The incident which illustrates the possibility of
non-verbal arguments occurs before this transpires, when Aesop is still unable
to express himself verbally – a situation which forces him to argue in a non-
verbal way. In the incident in question, Aesop’s fellow slaves try to take
advantage of him by stealing the master’s figs, eating them, and blaming it on
Aesop. Though he cannot speak, he manages to outwit them by swallowing
warm water, putting his finger down his throat and forcing himself to throw
up, the results proving that he has not eaten the figs (when he points to the
other slaves and they are forced to do the same, the evidence literally falls from
their mouths and they are punished both for stealing the figs and for trying to
frame a man who cannot defend himself).
How should we understand this incident? Clearly, Aesop has not forwar-
ded a verbal argument. But he has still managed to convince his master that he
is innocent by providing evidence to this effect. We might summarize the
implicit argument as:
Premise: If I was guilty of eating the figs, they would be in my vomit.
Premise: They are not in my vomit.
Conclusion: I am not guilty of eating the figs.
This is a straightforward instance of modus tollens which can be represented
as:
Premise: g Ú v
Premise: v
Conclusion: g
where g = I am guilty of eating the figs; v = They are in my vomit. In demon-
strating his innocence with this argument, Aesop provides an example of a
simple propositional argument which is conveyed non-verbally. In the process
he provides an example of a physical argument which is “visceral” in Gilbert’s
sense (1997, 75-86), but one which shows that it is a mistake to think that such
arguments need to be understood and assessed in a way that makes it useful to
differentiate them from “logical” modes of argument.
Other cases of non-verbal argument can be dealt with similarly. Consider
a pharmaceutical company which tries to persuade us that we should buy their
brand of cough syrup by airing television advertisements in which men and
women in white lab coats enunciate its superior effects. If one restricts one’s
attention to their verbal claims, one might easily recognize a verbal argument
that maintains that we should buy their brand because it has the qualities they
enumerate. But this is a small part of what is going on. Visually, we can easily
discern another argument. Considered from a visual point of view, the lab
coats that the actors wear are not accidental, but an attempt to present them as
experts on the medical properties of cough syrup. If they have stethoscopes as
well as lab coats, we are being visually told that they are doctors. We can thus
recognize another argument which might be summarized as follows.
Premise: These doctors recommend this brand of cough syrup.
Premise: You should accept their recommendation given their status.
Conclusion: You should buy this cough syrup.
Though this argument is never explicitly stated, it can still be usefully treated
as a case of argument in the traditional sense – more specifically, as an attempt
to construct an appeal to authority which should be assessed accordingly.
It goes without saying that it may, in the process of assessing an argument
of this sort, be useful to say something about our psychological reactions to
authority (or white lab coats or stethoscope). This is an important aspect of
the argument about which many useful things might be said. The problem is
the prevalent assumption that the importance of this aspect of the argument
implies that we cannot treat it as an argument in a very traditional way. Rat-
her, it should be said that what needs to be said in this regard is entirely com-
patible with the attempt to see this visual message as an argument in the tradi-
tional sense. It is in keeping with this that much recent work on “visual argu-
ments” suggests that a great many non-verbal visual images are used to for-
ward implicit premises and conclusions, implying that the standard notion of
argument can be use fully employed in the attempt to explain and assess
persuasion in such contexts (see, e.g., Birdsell & Groarke 1996, Blair 1996, and
Groarke 1999).
Concluding Remarks
In the present paper I have argued that argumentation theory has been too
willing to embrace alternatives to the notion of argument which logic has
traditionally assumed. Instead of adopting these alternatives, I have argued
that it would do better to extend this traditional notion and understand ordi-
nary argument and persuasion in terms of it. There is no way to conclusively
establish this approach in a single article, but it can still be said that this ap-
proach to ordinary argumentation has not been given the serious consideration
it deserves, and that the problems with the more usual approach have been
44 Leo Groarke
ignored rather than appreciated. If nothing else, the approach I have suggested
challenges argumentation theory to more convincingly demonstrate (1) that
there are ordinary which show that some forms of argumentation cannot be
understood as arguments in the traditional sense; (2) that non-logical forms of
argumentation are in some way irreducible to argument in the traditional
sense; and (3) that the distinctions which are standardly made between diffe-
rent kinds of argumentation can withstand careful scrutiny. I think the ex-
amples I have already provided suggest that this is much more difficult than
usually imagined, and that the standard approaches to argumentation theory
tend to needlessly complicate our understanding of ordinary argument and
persuasion. Until argumentation theorists show otherwise, we will do better to
develop an approach to argumentation which has more affinities to the tradi-
tional account of argument which argumentation theory has rejected. It is
better to be a hedgehog with one good trick than a fox with a bag of tricks that
do not work as well.
References
Birdsell, David & Leo Groarke (1996). “Toward A Theory of Visual Argument.” Argumentation
and Advocacy, Vol. 33, No. 2.
Blair, J. Anthony (1996). “The Possibility and the Actuality of Visual Arguments.” Argumentation
and Advocacy, Vol. 33, No. 2.
Daly, Lloyd W. 1961. Aesop Without Morals. New York: Thomas Yoseloff.
de Sousa, Ronald (1987). The Rationality of Emotion. Cambridge: The MIT Press.
Barwise, John & John Etchmendy. (1992). “Visual Information and Valid Reasoning.” In Leslie
Burkholder, ed. Philosophy and the Computer. Boulder: Westview Press.
Gerritsen, S. (1994). “A Defense of Deductivism in Reconstructing Unexpressed Premises,” in van
Eemeren & Grootendorst, 1994.
Gilbert, Michael (1997). Coalescent Argumentation. Mahwah: Lawrence Erlbaum Associates.
Govier, Trudy. (1987). Problems in Argument Analysis and Evaluation. Amsterdam: Foris
Publications.
Groarke, Leo (1992). “In Defense of Deductivism: Replying to Govier,” in Van Eemeren et. al.,
eds. (1992)
—— (1996). “Logic, Art and Argument.” Informal Logic. 18, Nos. 2&3, 105-129.
—— (1999). "Deductivism Within Pragma-Dialectics," Argumentation 13, 1-16.
Johnson, Ralph H. (1996). The Rise of Informal Logic. Newport News: Vale Press.
—— & J. Anthony Blair (1994). Logical Self-Defense. (United States Edition) New York:
McGraw-Hill.
Levi, Don S. “The Case of the Missing Premise,” Informal Logic Vol 17, No. 1, Winter 1995.
Mortenson, G.D. & R.L. Anderson (1970). “The Limits of Logic.” Journal of the American
Forensic Association, Vol. 7.
Perelman, Chaim 1982. The Realm of Rhetoric. Notre Dame: University of Notre Dame Press.
Toulmin, Stephen 1958. The Uses of Argument. Cambridge: Cambridge University Press.
Walton, Douglas (1992). The Place of Emotion in Argument. University Park: The Pennsylvania
State University Press.
—— (1998). Argument Structure: A Pragmatic Theory. Toronto: University of Toronto Press.

—— and Erick C. Krabbe (1995). Commitment in Dialogue. Albany: State University of New
York Press.
Woods, John (1995). “Appeal to Force.” In Hans V. Hansen and Robert C. Pinto, Fallacies:
Classical and Contemporary Readings. University Park: The Pennsylvania State University Press.
Wreen, Michael J. (1988). “May the force be with you,” Argumentation 2.
Van Eemeren, Frans H. & Rob Grootendorst 1992. Argumentation, Communication, and Falla-
cies: A Pragma-Dialectical Perspective. London: Lawrence Erlbaum Associates.
——, eds. 1994. Studies in Pragma-Dialectics. Amsterdam: International Centre for the Study of
Argumentation.
——, J.A. Blair, and C.A. Willard, eds. (1992). Argumentation Illuminated. Amsterdam: Interna-
tional Society for the Study of Argumentation (ISSA).
46 J. Anthony Blair
J. ANTHONY BLAIR
Presumptive Reasoning/Argument: An Overlooked Class1
(a) Mother to child: Jay, you mustn’t say that your grandmother’s trifle looks
like the dog’s throw-up. That hurts her feelings, and it’s wrong to hurt
someone’s feelings.
Jay: But Mum, it’s true!
Mother: Never mind.
(b) Matthias: Professor Pinto, I would like you to reconsider my grade, please.
You gave Wichert’s term paper an A-. I read it, and I can’t see how my
paper is any worse, but your teaching Assistant gave mine a C-.
Professor Pinto: Certainly. If you’re right either your grade should go up or
Wichert’s should come down.
(c) Sally (calling from another room): Scott, what time is it?
Scott: Nine-thirty.
Sally: Thanks. I’d better get going: I promised Frans I’d meet him at ten.
(d) Town mayor to petitioning merchants: Ladies and gentlemen, we can’t
possible declare the first week in June the town’s official “Ice-Cream
Week.” If we did that, then the vendors of every product sold in town
would want their own official “Week,” and we’d have to give it to them.
But there are more than fifty-two products sold in town, so we’d have to
either turn some down, or have double or even triple official “Weeks.”
And there’d be competition for weeks deemed “better” than others. The
result would be a disastrous legal nightmare or an equally disastrous confu-
sing tangle of product promotions, or both. I’m sorry, but it’s out of the
question.
(e) Coach Scotty Bowman (addressing the Detroit Red Wings hockey team,
April 1999): Men, we all want to win the Stanley Cup for a third year in a
row. You know as well as I do what that means. We’ve got to fore-check,
we’ve got to back-check, we’ve got to move the puck up-ice fast when we
get a turn-over, we’ve got to get somebody in front of the net when we get
1 Many of the ideas in this paper owe their origins to what I took away from conversations with
my colleague, R.C. Pinto, who may wish me to take complete responsibility for them. My debt to
Douglas Walton’s work is, I hope, clearly indicated by the numerous citations of it. I also thank
Professor Dieter Mans, whose comments on an earlier version of the paper have led to corrections
of fact and improvements of detail and exposition.
the puck in their end, we’ve got to clog between the blue lines when
they’re trying to bring the puck up ice, we’ve got to use the body aggressi-
vely, we’ve got to protect the goalie, we can’t take retaliation penalties,
we’ve got to work, work, work! You know all this. You’ve done it before.
Let’s do it again!
(f) Father:Son, I don’t want to catch you smoking again. Smoking is bad for
you.
Son:But Dad, you smoke.
(g) Consumer’s Digest magazine: The Zero refridgerator rates “excellent” for
quick-freezing, temperature stability, and storage space; the shelf-configu-
ration and flexibility score “very good,” as does the service record; its noise
level and ease of cleaning the air-intake grill are only “fair”; and price and
resale value rate “poor.” On balance, and in comparison to what else is on
the market, Consumer’s Digest rates the Zero a fair-to-good buy in a re-
fridgerator, about in the middle of the pack.
1. Introduction
The aim of this paper is to draw attention to an important, distinctive, relative-

ly neglected class of reasoning or argument. Theorists in Artificial Intelligence
and Expert Systems who have wanted their modelling or simulations of human
reasoning to be empirically informed (and not based on a prioristic assump-
tions about its nature) have been giving the large class of reasoning and argu-
ment discussed here their study for some time. (See the discussions of non-
monotonic reasoning reported in Gabbay, Hogger and Robinson, 1994, for
example.) Philosophers and others, too, at least to the extent that they are
interested in accounting for human reasoning and argument, should give it
their close attention.
The paper begins with some terminological clarifications, because words
like ‘reasoning’ and ‘argument’ are slippery, yet are often, even in scholarly
discourse, used in ill-define ways. The concept of presumptive reaso-
ning/argument is then introduced via a discussion of one way of assessing
reasoning or arguments. There follows a list of some of the types of presump-
tive reasoning/argument and a sketch of the main generic properties of the
class. The paper ends with a rejoinder to a challenge to the claim that presump-
tive reasoning/argument is a distinct class of reasoning and argument.
48 J. Anthony Blair
2. Reasoning and argument, and their norms
Reasoning, whatever else it may entail, is uncontroversially at least the mental

act of inferring. One plausible analysis of inferring is: changing or reinforcing
a belief or other attitude in response to exposure to a body of information (see
Harman 1986).
How good the reasoning is depends on two things: whether the reasoner
takes the appropriate epistemic stance toward the information, and whether
the relation of the body of information to the propositional attitude inferred
from it satisfies the appropriate norms. Thus, first, reasoning is not good if the
reasoner drew an inference from information he took to have a given value (for
example, to be true) when he should have known it did not have that value (for
example, it was clearly false or was certainly dubious). Garbage in, garbage
out. And, second, reasoning is not good if the reasoner took the information
to support his conclusion to a certain degree when in fact it provided a diffe-
rent degree of support. He might have mistaken strong support for weak
support, moderate support for strong support, and so on. (We also say that
reasoning is bad if the agent failed to draw any inference at all from informati-
on he was aware of when in fact that information supported some attitude of
importance to him. Example: “When you noticed that you passed that hotel a
third time, did it not occur to you that the taxi driver was taking you for a
ride?” This point suggests that the norms of reasoning have a pragmatic ele-
ment.)
Some might prefer to restrict reasoning, or its assessment, to the second of
the above factors. On that conceptualization, if the person’s conclusion is well
supported by his premises, assuming them to be true, then his reasoning may
be said to be good, even if it should turn out that his premises were false and
he should have known that. On the other hand, to accept a premise that is
patently implausible seems also to be, or to imply, a failure of reasoning, for
the person who accepts a wildly implausible claim and then reasons from it
fails to recognize the inconsistency between that premise of his reasoning and
other, well-established beliefs that he holds. For example, the person who
begins making large financial commitments on the basis of having received in
the mail an announcement from Publisher’s Clearinghouse that he has, out-of-
the-blue, “won” $10 million, has not “thought it through,” as we might say.
This consideration argues for the broader conception of reasoning assessment,
which in turn implies a broader conception of reasoning than just the act of
inferring from information received.
The word ‘argument’ is a veritable Hydra: besides its seven meanings
described in the Oxford English Dictionary (1971), the more it is analyzed, the
more senses are distinguished (or stipulated). In one sense, common in phi-
losophy, the information from which a reasoner inferred a conclusion, toget-
her with the conclusion expressed as having been inferred from that informati-
on, is termed an “argument,” or that person’s “argument.” We say of such a
person that he reasoned, not that he argued; but we, or philosophers, at least,
nonetheless speak of the “argument” of, or implicit in, or corresponding to, his
reasoning, or indeed, of “his argument”: “Corresponding to every possible
inference is an argument, . . .” (Copi and Cohen 1990: 6).
The merits of such arguments are (unsurprisingly) determined in exactly
the same way that the merits of the product of the reasoning represented by
such arguments are determined, noted above, for the two labels denote the
same thing.
That first sense of ‘argument’ is frequently confused with a somewhat
different one, namely, the proposition/s communicated by one person (the
arguer, the person arguing or making the argument) to another or others (the
audience or the interlocutor) as being true or belief-worthy and as supporting
some further proposition, such that, in the attempt to persuade or convince the
audience to accept that further proposition, the arguer invites the audience to
infer the latter from the former (Pinto 1995: 276).
In addition, sets of propositions that might represent reasoning, although
no one has reasoned in that way, or that might be used to try to convince an
audience, although they have never been so used, are also in many quarters
termed “arguments.” These might more accurately be called possible argu-
ments.
Notice that someone can use an argument to try to convince an audience
(‘argument’ in the second sense) although it does not represent his own reaso-
ning (‘argument’ in the first sense) – as, for example, when an arguer presses an
audience with the implications of its previously-expressed belief commitments,
although he does not share those commitments.
The merits of the products of arguments of the second sort are much more
difficult to determine than those of the first sort, because arguments of this
second sort are much more complicated than the former. They are not just sets
of propositions, but sets of propositions put to particular uses in a complex
social practice. As van Eemeren and Grootendorst (1983) have pointed out,
such arguments are multi-dimensional: functional, social, expressed in langua-
ge and dialectical (or interactional). As a result, their evaluation must be multi-
dimensional as well. These various aspects of any such argument are in additi-
on to the argument’s “logical” (Walton 1996b) or “illative” (Blair 1995) core
(the grounds adduced in support of claims). Thus such an argument can go
wrong by failing to persuade, by failing to communicate, by violating speech-
50 J. Anthony Blair
act conventions, or by failing to respond to the interlocutor’s challenges as

well as to by appealing to false premises or by inviting a strength of com-
mitment to a conclusion inappropriate to the force of the evidence.
For whatever reasons, philosophers interested in the norms that govern
reasoning or argument have traditionally focused on the illative core, in ab-
straction from the social practice in which it is embedded in arguments of the
second sort. There can be no objection to that predilection, provided that all
the factors germane to the assessment of the argument’s illative core can be
identified in abstraction from the surrounding context of argumentation, and
as long as the verdict on the illative core is not presumed to apply equally to
the entire argument (in the second sense). Classically, philosophers have
conceived the normative principles governing the two aspects of the illative
core – the premises’ adequacy in themselves, and the adequacy of the premises’
support for the conclusion – to belong to two distinct fields: epistemology and
logic, respectively. On the basis of that traditional conception (with no com-
mitment to its correctness), the subject matter of this paper can be identified as
belonging to logic, for the proposal of the paper has to do with the criteria for
the adequacy of the support that stipulated premises provide for an argument’s
conclusion in the illative core of argumentation.
Consider one type of illative core support structure, the type in which the
grounds adduced in support of the claim entail it. In such an argument or
reasoning, the conjunction of sentences expressing those grounds and the
negation of that claim is an inconsistent set. Entailment in this sense is almost
always uncontroversially a virtue of arguments and reasoning. When an illative
core structure has the property that the premises entail the conclusion, it is,
with rare exceptions, a good argument or good reasoning in that respect – that
is, with respect to the support the premises, taken as given, provide for the
conclusion. (The qualification is needed for the rare case in which the premises
are synonymous with, or presuppose, the conclusion, that is, for the case of
circular or question-begging reasoning or arguments.)
Is entailment a necessary as well as a (normally) sufficient condition of
good argument or reasoning support? Are there other ways an argument’s or
reasoning’s support can be good besides by entailing the conclusion (using
‘entailment’ as stipulated above)? Many philosophers think that entailment is
not the only virtue of argument or reasoning support. The following is an
argument for that position.
(1) The general criterion for any type of premise-conclusion support is that
it be rational, on the basis of accepting its premises, to accept the conclusion of
any argument or reasoning with that type of support. (2) There are various
types of support to be found in arguments or reasoning such that, given the
premises and their relation to their conclusion, it is rational to accept that

conclusion on the basis of those premises, although the premises do not entail
the conclusion. Hence, (3) premise-conclusion entailment is not a necessary
condition of good premise-conclusion support in reasoning or argument.
Since the conclusion of this argument is entailed by its premises, the argu-
ment does not beg the question by assuming to be acceptable the possibility of
the rational, non-entailment premise-support that it is used to argue for.
Examples that support premise (2): “All observed chimpanzees have had a
vegetarian diet, so chimpanzees do not eat meat.”; “60% of a large representa-
tive sample of voters have indicated their intention to vote Tory in the next
election, so the Tories will win the next election.”; “Cats tend to dislike water,
so probably this cat will dislike water.”; and so on. These examples suggest
that various types of reasoning/argument with inductive support are rational.
But there are other types of support as well that are rational yet non-deductive
and to these we now turn.
3. Presumptive reasoning/argument
So far, the discussion has reviewed old ground. Consider next the following
classes of reasoning or argument exemplified by the invented, but realistic,
examples provided at the beginning of the paper. (The list of types is meant to
be illustrative, not exhaustive.)
(a) Inferences or arguments from general moral principles, plus descriptions of
particular circumstances, to judgements about what ought to be done in
particular cases.
When Jay’s mother appeals to the principle that one ought not to hurt
another’s feelings, she need not take that principle to be exceptionless. She
merely insists that it applies in this case. In fact she takes it in this case to
override the principle, invoked by Jay, that one should tell the truth, which
thus itself cannot be a moral principle that has no exceptions. Presumably,
both principles can be overridden. To this date moral theorists have agreed to
no meta-rules that strictly settle conflicts of moral principles. (Sir David Ross’s
classic discussion of “duties” remains illuminating in this respect (1930, Chap-
ter 2).) The result is that such inferences as the one Jay’s mother makes and
urges on him are not entailments. Nor, obviously, are they inductive in any
scientific sense.
(b) Case-by-case analogical reasoning (of the following form): “A token of
action/judgement-type X was correct in case A; case B appears to be similar to
52 J. Anthony Blair
case A in the salient respects; so a token of action/judgement-type X is correct

in case B.”
In Matthias’s argument to persuade Dr. Pinto to reconsider the grade his
essay received, Matthias is claiming that his essay was analogous Wichert’s
essay in relevant respects (presumably, with respect to those of its properties
that satisfied the criteria for an A-range grade), and that therefore his grade
should be similar to Wichert’s. Dr. Pinto must judge not only whether the two
essays are similar in the respects noted by Matthias, but also whether they are
similar in all respects relevant to the grade. For example, were Matthias’ argu-
ments as thorough, was his exposition as lucid, was his paper’s organization as
perspicuous as those of Wichert? Such assessments are, as we say, “judgement
calls.” Furthermore, even if the papers were similar, unless they were identical
to a degree that suggests plagiarism, Dr. Pinto must judge whether the degree
of similarity justifies a similar evaluation. So far, we have no calculus for the
determination of such judgements.
(c) Appeals to testimony: reasoning of the form, “According to S, P is true; so
P is probably true.” – and the many variants of it.
Some argue that basing a belief on another’s testimony in simple cases, such
as when you ask them the time of day or for simple directions (“Which way to
downtown, please?”), entails no inference, but rather is epistemically akin to
basing a belief on what one sees (for example, believing there is a tree ahead on
the ski-run because one sees a tree ahead on the ski-run). If they are right, then
the example of Sally’s accepting that it is 9:30 on the basis of Scott’s sayso is
not a good example of a presumptive inference because it is not an example of
an inference.
Without taking the time to dispute that position here, it can be said that
there are other kinds of case in which beliefs based on appeals to testimony are
uncontroversially inferential. A major class consists of appeals to (expert)
authority – reasoning that a proposition is (presumptively) true on the
grounds that one or more other people who have a privileged epistemic access
in such cases have declared it to be true. Thus, when several independent
oncologists consulted by a cancer patient produce the same diagnosis and
recommend the same treatment, it is reasonable for the patient to conclude
that he has the condition they name and would be well-advised to take the
treatment they prescribe. Yet, famously, medical diagnoses can be wrong and
recommended treatments can turn out to have been ill-advised. So the patient’s
inference is no entailment. And, again, it is not an induction in any standard
sense.
(d) All “balance of considerations” or “pro and con” reasoning or arguments.

This is reasoning or argumentation in which there are both reasons that sup-
port the conclusion and reasons that support the contradictory of the con-
clusion. Typically, there is more than one line of support on each side. The
reasoner must weigh the force of each line of support, sum the weight of
considerations on each side, and decide which side’s reasons or arguments are,
on balance, more compelling.
In the example given, the town mayor lists only the disadvantages of the
proposal that the ice-cream sellers have made. In arguments like this, which are
designed to persuade an audience, it is not uncommon to find reference to the
reasons on one side only. The mayor can safely assume that the ice-cream
vendors know the reasons they advanced in favor of declaring the first week of
June “Ice-Cream Week.” Still, the mayor’s reasoning tacitly includes the
judgement that the “con” arguments he mentions outweigh the “pro” argu-
ments the vendors had advanced or others that might exist. While heuristics
would doubtless be helpful, there obviously can be no mechanical decision-
procedure for such reasoning.
(e) All means-ends reasoning or arguments: for or against an action or policy
on the basis of the value of the predicted consequences of performing or
implementing it. There are many variants of such reasoning.
The Detroit Red Wings’ coach, Scotty Bowman, listed what seemed to him
to be the salient means to the end of his team’s winning hockey games in the
National Hockey League playoffs in the spring of 1999. Given the premise
that the team wanted the end (to win the Stanley Cup), if Bowman is right that
these are means to that end, then he has given the team a good argument for
executing those means. But there is no entailment, for a couple of reasons. For
one thing, the members of the team may want that end, but want other ends
more (for instance, to avoid career-threatening injuries). For another thing,
they might agree that the means Bowen cites are good ones, yet also believe
that other, conflicting, means are in some circumstances superior. [Historical
note: the Detroit Red Wings were eliminated in the second round of the 1999
Stanley Cup playoffs by the Colorado Avalanche, 4 games to 2 in the best of
7 games series.]
(f) All reasoning or arguments for or against someone’s credibility on the
grounds of his/her past conduct or character.
The son’s reply to his father’s admonition not to smoke is a classic example
of a circumstantial ad hominem argument. Walton (1998) discusses this ex-
ample, and argues persuasively that such reasoning can be non-fallacious, even
though the son’s retort identifying the inconsistency between his father’s
54 J. Anthony Blair
advice and his behaviour does not succeed in discrediting his father’s position.
In general, we usually have nothing else to rely on than (the record of) a per-
son’s past conduct when we assess their credibility or their suitability for a job
or a role. This reasoning is not statistical (the candidate conducted herself well
in S1, S2, S3, . . . Sn, so she will conduct herself will in Sn+1), for we allow for
balance-of-considerations reasoning in coming to a conclusion about the net
indication of a record that is a mix of favourable and unfavourable performan-
ces, and we select certain past circumstances to be more salient as predictors of
the present case than others.
(g) The reasoning involved in evaluations of any kind.
This class is huge. The field of evaluation is a professional and academic
discipline (see Scriven 1991). Product evaluation of the sort exemplified by the
Consumer’s Digest magazine assessment of the Zero refridgerator, is but one
large subclass. There are also, among others, personnel evaluation, program
evaluation, policy evaluation, technology assessment, and environmental
impact assessment. Beyond its “professional” exercise, “evaluation is what
distinguishes food from garbage, lies from truth, and science from superstition.
In short, it is the sine qua non of intelligent thought and action . . .” (ibid.:
139-140). A high percentage of the inferences involved in evaluative reasoning
are presumptive. The identification of the relevant criteria of evaluation, the
determination of the appropriate standards to use in applying those criteria to
the particular case, the grading of the entity being assessed using those criteria
and standards, all involve complex prima facie inferences. Heuristics are useful,
but this is another type of reasoning in which mechanical decision-procedures
seem to have limited application, if any.
In vast numbers of the arguments in such reasoning the premises do not
entail the conclusion, or so I will argue. In just this sense, these are not types
of “deductive” reasoning or argument. Neither is any of these a type of “in-
ductive” reasoning or argument in any but the vacuous sense that they are
non-deductive. None exhibits the standard kinds of inductive reasoning or
argument typically found in science.
(a) – (g) are all types of what Walton (1996a) has called “presumptive”
reasoning or arguments. Presumptive reasoning has the following properties.
(1) The reasoning-types have identifiable patterns. These are termed reaso-
ning (or argument) “schemata” or “schemes” (singular: “schema” or “scheme,”
respectively) (see Schellens 1987, van Eemeren and Kruiger 1987, Kienpointner
1987, 1992a, 1992b, Walton 1996a). It is useful to distinguish between schemes
that are patterns of good reasoning (here termed “normative” schemes) and
schemes that simply describe how people tend to reason or have reasoned in
the past (called here “descriptive” schemes).
These normative schemes can be expressed with greater or less generality,
and may be classified in various ways. The scheme of relying on a spouse’s
sayso for the time of day is a particular type of the more general scheme of
relying on a person’s sayso for information. And clearly, relying on a physici-
an’s judgement for a medical diagnosis belongs to the same genus of schemes
as the first two, yet is a particular type of the more general scheme of relying
on expert judgement for an assessment or classification of a phenomenon –
which in turn is related to, but different from, relying on expert judgement for
a recommendation of a treatment or a solution to a problem. To mention
another example, involving a different type of reasoning/argument scheme,
Walton (1998: 261) identifies altogether 18 species, subspecies, sub-subspecies,
and sub-sub-subspecies of the generic ad hominem argument, for instance, the
circumstantial ad hominem (broad sense), the direct ethotic ad hominem
(including the variants: from veracity, from prudence, from perception, from
cognitive skills, from morals), the bias ad hominem, poisoning the well, guilt
by association, two wrongs, tu quoque, and so on.
(2) (a) Each normative scheme describes a pattern of reasoning or argument
the premises of which, if true, provide support for the conclusion, but only if
other things are equal. This point may be expressed in a variety of other ways.
(b) A normative scheme is a default form of good reasoning, and any such
scheme is subject to override in a variety of circumstances. (c) Reasoning or
arguing in the pattern of a normative schemes is prima facie good reasoning,
but whether it is good reasoning all things considered depends on the circums-
tances in that case. (d) The grounds that instantiate the premise-types of a
given normative scheme create a presumption in favor of the proposition
instantiating the conclusion of that scheme. Whether that presumption is
borne out depends on the particulars of the case. (e) In an dialogue in which
arguments are traded by the interlocutors, the party with grounds instantiating
a normative scheme thereby shifts the burden of proof to the other party. This
variety of idioms attests to the ubiquity of this kind of reasoning.
An example will illustrate the point concretely. Having promised to do
something is a reason to do it, but (pace Kant) it is a decisive reason only in the
absence of countervailing considerations. Perhaps in this case keeping the
promise conflicts with a stronger obligation, or would cause enormous harm
and produce little benefit. In either of these, or similar, circumstances, the
promise is overridden. However, in their absence, and that of other similar
disqualifiers, the promise is a decisive reason to perform the promised action.
56 J. Anthony Blair
Similar examples could be given for the other types of reasoning or argument
listed.
(3) For any given reasoning/argument scheme, the types of circumstances
under which other things are not equal (in which the default does not apply /
the reasoning is not good all things considered / the presumption is overrid-
den) cannot be exhaustively specified, because (a) the variations on each sche-
me are innumerable, and (b) the circumstances of the instantiations of each
scheme are too varied for the exempting conditions of any one instantiation to
be generalizable to all of them.
(4) However, in spite of their capacity for infinite variation, human moti-
vation and conduct tend to follow a limited number of well-defined paths.
Hence, it is possible to outline in a general way for any scheme the principal
classes of exceptions, and so to provide useful general, if not universal, guideli-
nes for the employment of that scheme. In the literature on schemes some
theorists (Schellens 1987, van Eemeren and Kruiger 1987, Walton 1996a) have
formulated what they call “critical questions” for each scheme, the role of
which is to remind its user of the types of circumstances that typically derail
reasoning of the pattern represented by that scheme. The critical questions
function as a check-list to help determine whether any of the standard types of
excepting conditions that should cancel the default represented by the scheme
are presented in that particular instances of its employment.
An example will illustrate. The following are Walton’s (1996a: 77-79)
scheme and critical questions for “argument from analogy”:
Scheme: Generally, case C1 is similar to case C2.
A is true (false) in case C1.
Therefore, A is true (false) in case C2.
Critical Questions: 1. Are C1 and C2 similar, in the respect cited?
2. Is A true (false) in C1?
3. Are there differences between C1 and C2 that would
tend to undermine the force of the similarity cited?
4. Is there some other case C3 that is also similar to C1, but
in which A is false (true)?
The thesis of this paper is that presumptive reasoning/argument represents a
sui generis class of reasoning/argument. As noted, computer scientists have
long been aware of this type of reasoning, and a few philosophers have also
(see Govier 1987, 1999, who has drawn attention to Wellman 1971 and Wis-
dom). But it has not, at least in philosophy, received much attention as a
distinctive class with a wide variety of sub-types that is on a par with, and
deserving as careful attention as, “deductive” and “inductive” reasoning or
argument. Scriven (1987) was the first philosopher I am aware of who discus-
sed it as a set of types of reasoning or argument exhibiting a distinctive logic
(or set of logics), which he called probative logic, but his highly suggestive
paper is rarely cited in this connection. Argument schemes lurk in the back-
ground of discussions of informal fallacies, as Walton (1996a) noted. He
deserves credit for having made explicit the role of presumptive reaso-
ning/argument in informal fallacies. However, although he explicitly recogni-
zes presumptive reasoning/arguments as a distinctive type, Walton limits his
discussion of it to its application in illuminating informal fallacies.
Walton (1996a, Ch. 2) identifies the rationality of this type of reasoning
with the rationality of practical reasoning – reasoning about what to do. That
seems true for means-ends reasoning, reasoning from moral principles and
case-by-case reasoning applied to action decision or judgement. It does not,
however, neatly fit the rationality of case-by-case reasoning applied to classifi-
cations, to testimony assessment, to balance-of-considerations reasoning, to
credibility judgements, or to evaluations. A case could be made that the latter
sorts of reasoning either always, or for the most part, bear eventually on
reasoning about what to do. But that risks stretching “practical reasoning”
beyond any useful denotation. Better, at least at the outset, to remain open-
minded about the possibility that each type within this broad class obtains its
rationality in a distinctive way.
4. The challenge of deductivism
Whether these variants of presumptive reasoning/arguments represent a dis-

tinct class presupposes that their premise-conclusion relation cannot adequate-
ly be modeled as deductive or by some well-established inductive model. To
be sure, deductive reconstruction is always possible for any reasoning or argu-
ment. That is, with the addition of “missing”/ “tacit”/ “unexpressed” premi-
ses, the reasoning or argument can be interpreted so that the premises entail
the conclusion. But are reasoning and arguments of these types correctly
understood “as attempts at deductive arguments” or reasoning (see Groarke
1999: 1)? Are they, for instance, best interpreted as follows?
1. P.
2. If P, then C, other things being equal.
3. In this case, other things are equal.
j 4. In this case, C.
The trouble with this interpretation is that 3 is too strong. It is not possible to
be justified in asserting categorically that other things are equal. Thus, a re-
58 J. Anthony Blair
sponsible reasoner or arguer would not assert 3, but rather something like 3’:
“In this case, other things appear to be equal.” But on that formulation, 1-3’
do not entail 4.
Perhaps there is another interpretation of the reasoning that does permit
responsible deductive reconstruction:
5. P.
6. If P, then, if other things are probably equal, probably C.
7. In this case, probably other things are equal.
j 8. In this case, probably C.
The trouble with this interpretation is that premise 6 is too tentative. In pre-
sumptive reasoning or argument the contention is that the default definitely
applies, in the absence of reasons or evidence to the contrary. By replacing 2
with 6, deductive reconstruction is obtained only by distorting the original
reasoning or argument.
Groarke (1999: 5) correctly points out that deductive reconstruction does
not imply that the conclusion is certain, but only that the premises necessarily
imply the conclusion. Now, this claim is standardly taken to mean that the set
of the premises and the denial of the conclusion is inconsistent – that it is a
contradiction to affirm the premises and deny the conclusion; that there is no
possible world in which the premises are true and the conclusion false. Yet in
the case of presumptive reasoning schemes, it is impossible be sure that other
things are equal. A world in which the evidence fitting a scheme all points to
the conclusion, and a thorough consideration of the circumstances uncovers
no excepting condition, yet some excepting condition has been overlooked, is
always possible, however unlikely. (In this respect, presumptive reasoning/ar-
guments are akin to inductive reasoning/arguments.)
In addition, it is no defect in such reasoning or arguments that their premi-
ses do not entail their conclusions. These conclusions are always, in principle,
subject to reconsideration in the light of new information which reveals a
fuller understanding of the circumstances.
It would appear, then, that deductive reconstruction without distortion is
not possible, and that it is not plausible to suppose those who reason or argue
presumptively are trying to reason or argue in a way that can be modeled as a
deduction. Assuming that inductive reasoning and argument are also obvious-
ly different, it then follows that presumptive schemes represent a distinct sort
of reasoning or argument.
5. Conclusion
This paper has been programmatic. The goal has been join my voice to others’
(Govier’s 1987, 1999, Scriven’s 1987, Walton’s 1996a) in drawing attention to
what appears to be a distinctive class of arguments or reasoning that cannot be
reduced to the standard categories of “deductive” or “inductive.” As the
examples listed illustrate, presumptive reasoning and arguments represent a
significant subset of all human reasoning and argument. They deserve tho-
rough and systematic study.
References
Blair, J. Anthony. 1995. Premise adequacy. In F.H. van Eemeren, R. Grootendorst, J.A. Blair and
C.A. Willard, (eds.), Proceedings of the Third ISSA Conference on Argumentation.Vol. III,
Reconstruction and Application, 191-202. Amsterdam: SicSat.
Copi, Irving M. and Cohen, Carl. 1990. Introduction to Logic. 8th edition. New York: Macmillan
Publishing Company.
Eemeren, Frans H. and Grootendorst, Rob. 1983. Speech Acts in Argumentative Discussions.
Dordrecht: Foris.
van Eemeren, Frans H. and Kruiger, Tjark. 1987. Identifying argumentation schemes. In Frans H.
Eemeren, Rob Grootendorst, J. Anthony Blair, and Charles A.Willard, (eds.), Proceedings of the
Conference on Argumentation 1986, Vol. 2, Argumentation: Perspectives and Approaches, 70-81.
Dordrecht-Holland/Providence-U.S.A.: Foris Publications.
Gabbay, Dov M., Hogger, C. J. and Robinson, J.A. (Eds.). 1994. The Handbook of Logic in
Artificial Intelligence and Logic Programming, Vol. 3: Nonmonotonic Reasoning and Uncertain
Reasoning. Oxford: The Clarendon Press.
Govier, Trudy. 1987. Problems in Argument Analysis and Evaluation. Dordrecht-
Holland/Providence-U.S.A.: Foris.
——. 1999. The Philosophy of Argument. Newport News, VA: Vale Press.
Groarke, Leo. 1999. Deductivism within pragma-dialectics. Argumentation, 13.1: 1-26.
Harman, Gilbert. 1986. Change in View, Principles of Reasoning. Cambridge, MA: MIT Press
(Bradford).
Kienpointner, Manfred 1987. Towards a typology of argumentative schemes. In Frans H. Ee-
meren, Rob Grootendorst, J. Anthony Blair, and Charles A.Willard, (eds.), Proceedings of the
Conference on Argumentation 1986, Vol. 1 Argumentation: Across the Lines of Discipline, 275-
287. Dordrecht-Holland/Providence-U.S.A.: Foris Publications.
——. 1992a. How to classify arguments. In F.H. van Eemeren, R. Grootendorst, J.A. Blair and
C.A. Willard, (Eds.), Argumentation Illuminated, 178-188. Amsterdam: SicSat.
——. 1992b. Alltagslogik, Struktur und Funktion von Argumentationsmustern, Stuttgart-Bad
Cannstatt: frommann-holzboog.
Oxford English Dictionary. 1971. Compact Edition. Oxford: Oxford University Press.
Pinto, Robert C. 1995. The relation of argument to inference. In F.H. van Eemeren, R. Groo-
tendorst, J.A. Blair and C.A. Willard, (eds.), Proceedings of the Third ISSA Conference on
Argumentation.Vol. I, Perspectives and Approaches, 271-286. Amsterdam: SicSat.
Ross, W.D. 1930. The Right and the Good. Oxford: The Clarendon Press.
Schellens, Peter Jan. 1987. Types of argument and the critical reader. In Frans H. Eemeren, Rob
Grootendorst, J. Anthony Blair, and Charles A.Willard, (eds.), Proceedings of the Conference on
60 J. Anthony Blair
Argumentation 1986, Vol. 3, Argumentation: Analyses and Practices, 34-41. Dordrecht-

Holland/Providence-U.S.A.: Foris Publications.
Scriven, Michael. 1987. Probative logic: review and preview. In Frans H. van Eemeren, Rob
Grootendorst, J. Anthony Blair, and Charles A.Willard, (eds.), Proceedings of the Conference on
Argumentation 1986, Vol. 1 Argumentation: Across the Lines of Discipline 7-32. Dordrecht: Foris.
——, 1991. Evaluation Thesaurus. Fourth edition. Newberry Park, CA: Sage Publications.
Walton, Douglas. 1996a. Argument Schemes for Presumptive Reasoning. Mahwah, NJ: Lawrence
Erlbaum Associates.
——. 1996b. Argument Structure, A Pragmatic Theory. Toronto: University of Toronto Press.
——. 1998. Ad Hominem Arguments. Tuscaloosa and London: The University of Alabama Press.
Wellman, Carl. 1971. Challenge and Response: Justification in Ethics. Carbondale & Edwardsville:
Southern Illinois University Press.
Wisdom, John. 1957. Explanation and proof. Lectures presented at the University of Virginia.
Unpublished manuscript. Cited in Govier, 1987, 1999.
Argument schemes and the evaluation of presumptive reasoning 61
ROBERT C. PINTO
Argument Schemes and the Evaluation of Presumptive Reasoning:

some Reflections on Blair’s Account
In “Presumptive Reasoning/Argument: An Overlooked Class,” Professor

Blair makes a largely convincing case that “presumptive reasoning/argument
represents a sui generis class of reasoning/argument” (Blair 1999, p.56). The
strength of Blair’s paper lies in the fact that
a.) he demonstrates through the use of examples that there is a large class
plausible arguments/inferences that don’t fit the usual models of good
deductive or of good inductive reasoning
b.) he sets out a list of four properties of presumptive reasoning – pro-
perties whose joint presence might be taken to be criterial for the class
in question.
I agree with Blair that there is a large number of arguments/inferences that
cannot be assessed in terms of the techniques of formal deductive logic or of
the probabilistic standards usually applied to so-called inductive inferences. I
agree, moreover, that recent work on argument schemes and the critical que-
stions associated with them promises to yield useful tools for dealing with
such reasoning. However, I have reservations about two aspects of the story
that Blair tells, and in what follows I will attempt to lay out the grounds for
those reservations.
1) In what sense does presumptive reasoning/argument constitute a sui

generis class?
Blair says that presumptive reasoning/argument “has not, at least in philoso-

phy, received much attention as a distinctive class with a wide variety of sub-
types that is on a par with, and deserving as careful attention as, ‘deductive’
and ‘inductive’ reasoning or argument” (see Blair, p. 57). Clearly, he wants to
contrast presumptive reasoning with both “deductive” and “inductive” reaso-
62 Robert C. Pinto
ning – though (wisely, in my view1) he puts the words ‘deductive’ and ‘inducti-
ve’ in scare quotes.
My reservation – albeit a minor one – is over the attempt to contrast pre-
sumptive reasonings with “inductive reasonings” (or perhaps better: to con-
trast “valid” presumptive reasonings with reasonings that are inductively
strong). That contrast is problematic, for at least three reasons.
First, the concepts of ‘inductive argument’ and ‘inductive strength’ are not
sharply defined. Blair himself acknowledges that presumptive reasonings are
inductive “in the vacuous sense that they are non-deductive”. But, he says,
“[n]one exhibits the standard kinds of inductive reasoning found in science”
(see Blair, p. 54). Yet there is, as far as I know, no precise criterion for de-
termining which of the reasonings that occur in science are inductive and
which are not.2 It might be tempting to identify inductive inference with those
inferences whose strength can be measured in straightforward ways by the
calculus of probability (and in particular with the help of inferential or projec-
tive statistics).3 But that I think yields too narrow a concept of inductive
inference – since (absent the success of the Bayesian program) it would exclude
the confirmation of theoretical hypotheses, which these days is a paradigm of
(so-called) inductive reasoning.
Second, there are kinds of reasoning which arguably belong to the class of
presumptive reasonings, but which are also standardly classed as inductive
arguments. Thus both Blair and Walton count argument from analogy as
presumptive, while Copi treats it as a species of inductive argument (see Copi
and Cohen 1990, p. 357), and even Trudy Govier recognizes two species of
argument from analogy, one of which she calls inductive analogy (see Govier
1989, p. 141). Moreover, Walton includes Argument from Evidence to Hypo-
1 In the literature of logic it is common to find people trying to classify arguments or inferences
as deductive or inductive, and maintaining that the appropriate standards for judging an argument
or inference depend on the category to which it belongs. With Brian Skyrms, I hold that to be a
mistake. Rather, we can ask about any argument, ‘Is it deductively valid?’ And if the answer to
that question is ‘No,’ we should then proceed to ask, ‘Is it inductively strong or is it inductively
weak?’ See Skyrms 1966. The usual attempts to classify arguments as deductive or inductive
inevitably bog down in questions of whether the argument’s author “claims” or intends it to be
taken as deductively valid or inductively strong – and in many cases there is simply no answer to
that question (e.g., where the arguer lacks the logical sophistication and conceptual apparatus to
think of arguments in terms of deductive validity and inductive strength).
2 Authors like Marcello Pera have argued, persuasively I think, that scientific discourse contains
a broad range of arguments that he calls rhetorical, at least some of which would fall into Blair’s
or Walton’s listings of presumptive arguments. See Pera 1994, esp. chapters 3 and 4. If that is so,
then we can’t define inductive (as opposed to presumptive) arguments/inferences as those that
occur in science.
3 Indeed, Balir’s examples of inductive support of are of this sort (Blair, p. 51).
thesis among his list of argumentation schemes for presumptive reasoning (see
Walton 1995, pp. 67-71 – though on p. 46 he expresses some hesitation about
its inclusion in the list), while Blair’s suggestion (Blair, p. 54) that inductive
inferences are those found in science would seem to lead to the view that
Argument from Evidence to Hypothesis is inductive.
Finally, one of the principal features of presumptive reasonings is the fact
that they are defeasible or that they are instances of nonmonotonic1 reasoning.
That is to say, additional information, consistent with the premisses of the
original argument or inference, can undermine or override the support that
those premisses give to their conclusion. (Thus ‘Tweety is a bird’ supports the
conclusion that Tweety can fly; but the additional information that Tweety
has a broken wing cancels or overrides the inference from ‘Tweety is a bird’ to
the conclusion that Tweety can fly.)
But defeasibility has long been recognized to be a feature of “inductive” or
probabilistic inferences as well (see, for example, Chisholm 1977, pp. 71-72).
And indeed, it is necessarily a feature of any argument or inference that is not
deductively valid.2 In virtue of this, most of the things that Blair says about
presumptive reasonings on pp. 54-56 seem also to apply to inferences that are
inductively strong (at least if patterns of “inductive inference” can be counted
as argument schemes). That leaves us with the question of whether and how
the defeasibility of presumptive inferences differs from the defeasibility of
inductive and/or probabilistic inferences.
These considerations do not, I think, undermine anything central to what
Blair is saying about presumptive reasoning. And I am inclined to think pre-
sumptive reasonings should be contrasted with the inferences usually deemed
inductive. But these considerations indicate the need to elaborate the precise
relation between presumptive reasoning and inductive inference and to clarify
the force of the claim that presumptive reasoning is sui generis.
2) Are there normative argument schemes?
In characterizing the relationship of argument schemes to presumptive reaso-

nings, Blair distinguishes between descriptive schemes and normative schemes.
The latter in contradistinction to the former are described as “patterns of good
reasoning” (Blair, p. 54, emphasis mine). Moreover, he characterizes the nor-
1 On nonmonoticity, see the discussion in Walton 1995, pp. 21-23, and van Bentham’s explanation
in van Bentham 1996, esp. pp. 29-30.
2 Let [p, q, r] be a set of premisses that support but do not entail a conclusion c. From the fact that
they don’t entail c, it follows that not-c is consistent with that set of premisses. But then the
enhanced set of premisses [p, q, r, not-c] will fail to support c, since it will in fact entail not-c.
64 Robert C. Pinto
mativity of such schemes in the following ways (which he appears to treat as

roughly equivalent):
C.1 Each scheme “describes a pattern of reasoning or argument the pre-
misses of which, if true, provide support for the conclusion, but only
if other things are equal.”
C.2 Each scheme “is a default form of good reasoning,” though “any such
scheme is subject to override in a variety of circumstances.”
C.3 Each instance of a normative scheme “is prima facie good reasoning,
but whether it is good reasoning all things considered depends on the
circumstances of the case.”
C.4 “The grounds that instantiate the premise-types of a given normative
scheme create a presumption if favor of the proposition instantiating
the conclusion….”
C.5 In a dialogue, “the party with grounds instantiating a normative
scheme thereby shifts the burden of proof to the other party.”
Implicit in these claims, I think, is Walton’s description of argumentation
schemes as “a formal pragmatic structure of arguments that is the counterpart
to logical forms of inference in semantics” (Walton 1995, p. x). In discussing
arguments of a deductive sort, Blair himself has said:
When the illative core structure [of an argument or inference] has the property
that the premisses entail the conclusion, it is, with rare exceptions, a good argu-
ment or good reasoning in that respect – that is, with respect to the support that
the premisses, taken as given, provide for the conclusion. (Blair, p. 50).
And of course in standard cases, we determine that premisses of an argument
entail its conclusion by determining that the argument instantiates one of the
logical forms of inference.1 Analogously, Blair seems to be suggesting in C.1-
C.5 that we can determine whether the premisses of an argument create a
presumption in favor of its conclusion by determining whether we are dealing
with an instance of a normative argument scheme.
Now I want to question whether this way of viewing the function of
argument schemes in the appraisal of arguments and inferences is correct
and/or fruitful. That is to say, I want to question whether the argument sche-
mes that Blair and Walton recognize really do have the normative force as-
1 Interestingly enough, as Blair himself concedes (Blair p. 39), the fact that premisses entail a
conclusion does not guarantee that they support it – circular or question-begging arguments often
entail but do not genuinely support them. (There are other exceptions as well; see Pinto 1994, pp.
121-122.). Indeed, one can say of instances of the logical forms of inference what Blair says about
instances of normative argument schemes in C.3 – namely, that each such instance “is prima facie
good reasoning, but whether it is good reasoning all things considered depends on the circums-
tances of the case.”
cribed to them. Since I am questioning whether these argument schemes are in

fact normative, I shall henceforth refer to the schemes in question simply as
recognized argument schemes, in order to leave open the question whether
they are indeed normative.
Let me begin by distinguishing between the two following two claims:
P.1 Any good presumptive argument/inference creates a presumption in
favor of its conclusion (and shifts to burden of proof to anyone who
would dispute that conclusion).
P.2 Any instance of a recognized argument scheme should be presumed
to be a good presumptive argument/inference (though that presump-
tion can be overridden in special circumstances).
I take (P.1) to be uncontroversial in this context – indeed, I think that it captu-
res part of the essence of what it is to be a good presumptive argu-
ment/inference. The second of these claims (P.2) is, I think, pretty much
equivalent to Blair’s C.2 – and, I suspect, to C.1 and C.3 as well. And it is (P.2)
that I want to question.
To help motivate my reservations here, let me use an example of a fairly
simple piece of presumptive reasoning, and then canvas several of the kinds of
criticisms that would lead us to reject either the reasoning or its conclusion.
Let me take as an example the Argument from Sign, whose scheme is
represented in Walton 1995 (p. 49) as follows:
A is true in this situation.
B is generally indicated as true when its sign, A, is true, in this kind of
situation.
Therefore, B is true in this situation.
Consider now two examples that fit this scheme (the first is from Walton, p.
49):
Case 1: Bob is covered with red spots.
Therefore, Bob has the measles.
Case 2: When I purchased that ring from a reputable jewelry store, it
came in a box that was labeled “24 karat gold”
Therefore, the ring is made of 24 karat gold.
Notice that each of arguments requires an unexpressed generalization to the
effect that one thing is a reliable sign of another: that being covered with red
spots is a reliable sign of having the measles, that the label on a box in a repu-
table store is a reliable indicator of the contents of the box.
66 Robert C. Pinto
Consider now some of the potential criticisms that might be made of these
arguments, or some of the potential defects that might be found in them:
1. The first expressed premiss might be defective or might be challenged.
For example, Bob’s face and torso might have seemed to be covered with
red spots because Bob was viewed through a stained window pane.
2. The unexpressed generalization might be challenged, for a number of
different sorts of reasons
a) The validity or reliability of the generalization might be challenged.
For example, are red spots a reliable sign of measles, and/or is it rea-
sonable to assume that they are? As a matter of fact, measles is just
one of many medical conditions that could be responsible for red
spots, so that this generalization might well be challenged.
b) Is the strength of the correlation between the sign and the signified
great enough to warrant a presumption? For example, the presence of
red spots gives some support to the hypothesis that a subject has
measles; but in the estimation of many the support would be too
weak to create a presumption that measles is present, or to shift the
burden of proof to anyone who would resist that conclusion.
c) Does the generalization, even if reliable, provide the appropriate sort
of evidence for the context at hand. In the context of informal dea-
lings with friends and family members, the store label probably does
provide an adequate ground for presuming the metal to be 24 karat
gold. But for insurance purposes it would surely be considered insuf-
ficient; for insurance purposes something like written assurance from
a qualified expert, or the results of chemical tests, would be required.
3. Though the initial premiss is OK (the box was indeed labeled “24 karat
gold”), and though the unexpressed generalization is sufficiently strong
and of a sort appropriate to the context at hand, there are additional facts
that undermine the inference. Suppose, for example, that an earthquake
occurring near the jewelry store in question had emptied most containers
of their contents, and that a stock clerk with little knowledge of precious
metals was responsible for placing items of jewelry back in their contai-
ners. Knowledge that such a thing had occurred would undermine the
inference in Case 2 – without, it should be noted, providing evidence that
the conclusion of that inference is false.
4. Though the inference does not suffer from any of the kinds of defects
catalogued so far, there is additional evidence that overrides the inference
in question, by supporting the negation of its conclusion. For example,
when the ring in Case 2 is brought to a highly reputable jeweler for
repair, he reports after careful examination that it is made of gold that’s

only 14 karats.
Now I want to call attention to what I consider to be an important difference
between the defects catalogued in (1) and (2), and the defects catalogued in (3)
and (4). The defects catalogued in (1) and (2) render an inference one that
should not have been made in the first place. As I see it, a respondent who
objects to a proposed presumptive inference on grounds that fall under (1) or
(2) is claiming in effect that proponent has failed to shift the burden of proof.
Let me call the allegation that a proposed inference suffers from such defects
a Type I criticism of that inference. It is worth noting, moreover, that the
kinds of objections envisaged in cases (1), (2a) and (2b) are each of a sort that
can be raised against arguments or inferences normally be classed as “deducti-
ve” or “inductive.” They are in no way specific to presumptive reasonings.
However, as I see it, in cases (3) and (4), the initial presumptive inference is
not defective. The inference goes through, as it were – in the sense that it in a
dialogue it shifts the burden of proof to the respondent or in a monolectical
situation it gives a reasoner a genuine ground for holding the conclusion. In the
absence of additional factual information about the case under scrutiny, the
conclusion retains its status as a presumption. And when we say that argu-
ments and inferences which avoid the defects catalogued in (1) and (2) shift the
burden of proof, what we mean amounts simply to this: it is incumbent upon
a respondent who would dispute the conclusion of such an argument to bring
forth precisely additional factual information that will either undermine the
inference, as happens in (3), or override the inference, as happens in (4). Let me
call the allegation that a presumptive inference can be undermined or overrid-
den in these ways a Type II criticism of the inference.
Now, if you examine the examples at the beginning of Blair’s paper and his
gloss on those examples in section 3, I think you will find in every case that he
is calling attention to inferences and arguments that avoid the defects catalo-
gued in (1) and (2), but have conclusions that are open to revision on the basis
of additional information that undermines or overrides those inferences, on the
pattern of (3) and/or (4) – he is calling attention to the possibility of Type II
criticisms of such inferences and arguments.
Let me return now to the claim I want to dispute, namely
(P.2) Any instance of a recognized argument scheme should be presumed
to be a good presumptive argument/inference (though that presump-
tion can be overridden in special circumstances).
I would maintain, rather, that when confronted with an instance of Argument
from Sign, we ought not to assume it is a good presumptive argument until we
68 Robert C. Pinto
have satisfied ourselves that it is not subject to a Type I criticism – anymore

than we should assume that a deductively valid argument is a good argument
until we have satisfied ourselves that its premisses are reasonable to believe,
that it does not beg the question, etc.1
Indeed, consider the two critical questions that Walton (following Ha-
stings) associates with Argument from Sign (Walton 1995, p. 48):
5. What is the strength of the correlation of the sign with the event signi-
fied?
6. Are there other events that would more reliably account for the sign?
I would maintain that until we’ve answered the first of those two questions,
we should not presume that an Argument from Sign even gets off the ground.
Question 1 does not ask after information that would “override” the presump-
tion that the argument in question is a good one; rather it asks for information
without which we can’t form a reasonable opinion about whether the argu-
ment is a good one. Question 1 belongs, I think, to Type I criticism. By con-
trast, question 2 could well be interpreted as a request for additional informati-
on that would undermine or override any presumption afforded to the con-
clusion by the sign cited in the premisses of the argument. Question 2 is an
attempt to instigate Type II criticism.
Yet if argumentation schemes aren’t normative in the way Blair claims
them to be – if they are not the counterparts of the logical rules of inference in
semantics – what is their role in the evaluation of arguments and inferences?
As I’m inclined to see the matter, the point of identifying argument schemes
lies in the critical questions associated with them. It isn’t the schemes that do
the evaluative work, it’s we who do the evaluative work by judiciously chal-
lenging premisses, identifying the risks of basing conclusions (if only pre-
sumptive conclusions) on those premises, and enriching our grasp of potential-
ly relevant fact that might alter the significance of the facts brought to light in
the premisses. Evaluating an argument or inference is a matter of probing its
strengths and weaknesses, and we do well if the initial impetus for our probing
comes from the critical questions that have, in the past, proved fruitful in the
probe of this, that or the other sort of argument. Associating arguments with
schemes is an effective way of associating them with particular sets of critical
questions – questions that will initiate the probes that may issue in evaluative
verdicts.
1 Nor should we presume that an argument which resembles inductively strong arguments in a
general way is itself inductively strong, until we have satisfied ourselves both that its premisses are
reasonable, and that the degree of support it lends to its conclusion is such as to warrant acceptan-
ce of its conclusion as probable.
I don’t claim to have shown that Blair’s account of the force and function
of argumentation schemes is wrong, or that mine is right. I do hope, however,
that I’ve brought to light an alternative to Blair’s way of conceiving the matter.
And I hope, as well, to have made it clear that at this stage more work needs to
be done on these issues.
References
Benthem, Johan van (1996). “Logic and argumentation”. In van Benthm, et al. 1996, pp. 27-42.
Benthem, Johan van, et al. (1996). Logic and Argumentation. North Holland, Amsterdam: Royal
Netherlands Academy of Arts and Sciences.
Blair, J. Anthony (1999). “Presumptive Reasoning/Argument: An Overlooked Class”. Protosocio-
logy 13 (1999).
Chisholm, Roderick (1977). Theory of Knowledge, second edition. Englewood Cliffs, N. J.:
Prentice-Hall, Inc.
Copi, Irving and Carl Cohen (1990). Introduction to Logic, Eight Edition. New York: Macmillan
Publishing Company.
Govier, Trudy (1989). Analogies and Missing Premisses. Informal Logic XI,3, pp. 141-152.
Johnson, Ralph and J. Anthony Blair (1994). New Essays in Informal Logic. Windsor, Ontario:
Informal Logic.
Pera, Marcello (1994).The Discourses of Science, trans. Clarissa Botsford. Chicago and London:
University of Chicago Press.
Pinto, Robert (1994). Logic, “Epistemology and Argument Appraisal”. In Johnson and Blair 1994,
pp. 116-124.
Skyrms, Bryan (1966). Choice and Chance; an introduction to inductive logic. Belmont, Calif.:
Dickenson Pub. Co.
Walton, Douglas (1995). Argument Schemes for Presumptive Reasoning. Madison, Wisconsin:
Lawrence Erlbaum Associates.
70 Douglas Walton
DOUGLAS WALTON
The New Dialectic: A Method of Evaluating an Argument Used for Some

Purpose in a Given Case
The purpose of this article is to explain to the reader how to evaluate an argu-
ment critically with respect to how that argument was supposedly used for
some purpose in a goal-directed type of conversational exchange. Of course,
only so much can be explained in a short article. Nevertheless, by introducing
the reader to the recent literature on argumentation, and to the main methods
that are being developed in that literature, some insight into how to use the
new techniques can be given. The subject is controversial, as well. Some deny
that there can be any binding standards for judging when an argument is
reasonable or not, or whether one argument is better than another, as used in
a given case, in natural language discourse. Others feel that the only objective
methods that can or should be used to support such judgments are those of
deductive and inductive calculi of the kind that have for so long been central in
the field of logic.
Because of the controversial nature of the subject, many fundamental
logical and philosophical questions are raised in this article. What is the diffe-
rence between argumentation and reasoning? What are the forms of presump-
tive inference commonly used in everyday argumentation? How are such
inferences chained together to make up sequences of reasoning? How is such
reasoning used for different purposes in different kinds of arguments where
two parties are involved in a dispute? What are the relationships between
argument and explanation? How can presumptive arguments be seen as ins-
tances of inference to the best explanation that provide a tentative basis for
accepting a conclusion? How does arguing fix an arguer’s commitment to
specific propositions that can then be attributed to her, as representing her
position? How do new argument moves in a disputation change that com-
mitment? How should change of commitment be organized in different con-
texts of dialogue when two parties reason together for some collaborative
purpose? In particular, what are the rules for retraction of commitments? Can
these different normative models of dialogue be formalized, with precisely
defined components and sets of rules? These are the new kinds of questions
that are being asked.
In this article, an exposition of developments in argumentation theory is
presented that gives the reader a revealing glimpse into how these questions are
The New Dialectic : A Method of Evaluating an Argument 71
being answered. The framework in which the questions are answered is called
the “new dialectic”. The new dialectic is mainly concerned with the most
common kinds of everyday arguments, and is based on presumptive reasoning
rather than deductive or inductive logic. Presumptive reasoning takes the from
of an inference in which the conclusion is a guess or presumption, accepted on
a tentative basis, and subject to retraction as a commitment, should new infor-
mation come in. The new dialectic shares many common features with the old
dialectic of Plato and Aristotle, but is also different from it in other features. In
the new dialectic, argumentation is analyzed and evaluated as used for some
purpose in a type of dialogue underlying a conversational exchange. Each type
of dialogue has its own standards of plausibility and rationality against which
to measure the successful use of an argument. Thus the new dialectic has a
relativistic aspect that makes it different from the classical positivistic phi-
losophy. But it also has a structure with logical standards of evaluation of
argument use, which makes it different from postmodern anti-rationalism.
1. Old and New Ways of Thinking about Thinking
The positivistic philosophy that has been the orthodox way of thinking in the
universities since the Enlightenment, and that has become even more dominant
in the twentieth century took science, especially the hard sciences of ma-
thematics, physics and chemistry, as the models of correct reasoning. The
kinds of reasoning used outside science, like the kind of thinking used in
everyday deliberations, or the kind for reasoning used in law and ethics, were
simply dismissed as “subjective”. Scientific reasoning, seen as consisting of
deductive logic and inductive logic of the kind represented by the probability
calculus, was taken to represent all of logic. This positivistic philosophy has
failed to yield a theory of reasoning and argumentation that was useful for
cognitive science, to explain how human or robotic agents can reason in col-
laboratively carrying out practical tasks on the basis of communicating shared
assumptions, or in criticizing the views of another agent. This positivistic view
point saw deductive logic, of the kind one would find in the reasoning in
Euclidean geometry, as the model of correct reasoning. The positivistic
viewpoint saw reasoned thinking along the lines the receiving of knowledge
and the revising of beliefs. Even the notion of an agent or thinker came to be
abstracted out of the equation. An argument was seen as simply a “designated”
set of propositions. Beyond propositional and quantificational logic, further
conceptualization came to be based on the highly abstract notion of a possible
world. Reasoning was thought to be about various kinds of “accessibility
relations” between pairs of possible worlds.
72 Douglas Walton
In the positivistic viewpoint, concern with thinking centered on a highly

abstract subject called “epistemology”. In epistemology, knowledge and belief
existed, and the central problem was how to connect the two concepts. The
subject matter seemed to consist mainly of endless controversies about whet-
her knowledge could be defined as justified true belief.
Both these lines of advance turned out to be dead ends. Spectacularly so.
Traditional logic and analytical philosophy turned out to have not enough to
say that was useful to tell those in the field of artificial intelligence how ra-
tional thinking should work. The aging priests of analytical philosophy still
talk enthusiastically about possible worlds and justified true beliefs, at some
centers of learning. But judging from the little respect accorded logic, the
humanities and philosophy in recent years, not many are listening any more.
To fill the gap, postmodern theories came along to advocate throwing aside
any pretense to rationality. But this viewpoint did not turn out to be a success
either.
Then along came argumentation theory. A rational argument was now
described in terms of acceptance (commitment) instead of belief or knowledge.
Departing from the impersonal framework of deductive logic, argumentation
theory saw an argument as a dialogue exchange between two parties who are
reasoning together. No longer exclusively concerned with deductive and
inductive forms of argument, argumentation theory considered many forms of
presumptive inference based on intelligent guessing that leads to a tentatively
acceptable conclusion on one side of a dialogue. This new viewpoint is not
acceptable to traditional analytical philosophers who are so used to looking at
the world through positivistic lenses. But computing, and artificial intelligence
in particular, has taken to argumentation like a duck to water, finding this new
theory very useful for all kinds of purposes. Philosophers will also accept these
views at some point in the future, once they realize that scientists have accep-
ted it.
2. The New Dialectic
This article presents an introduction to and outline of a new dialectic designed

to be used to normatively evaluate any arguments used in a given case. Origi-
nating in the old dialectic of the ancient philosophers, the new dialectic is
centrally concerned with the most common kind of arguments used in ever-
yday conversations, which is based on presumptive reasoning rather than
deductive or inductive logic. Presumptive reasoning is always based on default
by an argument from ignorance that tilts a burden of proof one way or anot-
her on an unsettled issue. Presumptive reasoning can be used in a closed world
situation or an open world situation. In what Reiter (1980, p. 69) calls the
closed world assumption, all the positive information in a data base is listed,
and therefore negative information is represented by default. For example, in
case 1 below (Reiter, 1980, p. 69), the design of an artificial intelligence questi-
on-answering system is considered.
Case 1
Consider a data base representing an airline flight schedule and the
query “Does Air Canada flight 113 connect Vancouver with New
York?” A deductive question-answering system will typically treat the
data base together with some general knowledge about the flight do-
mains as a set of premises from which it will attempt to prove CON-
NECT(AC113, Van,NY). If this proof succeeds, then the system re-
sponds “yes”.
But, as Reiter indicates, the interesting fact is that if the system does not suc-
ceed, it will typically respond “no”. In other words (Reiter, 1980, p. 69),
“Failure to find a proof has sanctioned an inference.” Such an inference by
default has often traditionally been called a lack-of-knowledge inference or an
argumentum ad ignorantiam. The same kind of inference drawn by Reiter’s
artificial intelligence system might be drawn by a human reasoner who is
scanning the flight monitor at the airport. When such a person sees the listing
of flights on the airport monitor, he presumes that all the flights are listed, and
infers by default that any flight not listed is not offered. Of course, he can
always check by asking at the desk. But the inference by default is most likely
a pretty good guess, because the person is plausibly justified in assuming that
the closed world assumption is met in this case. If the closed world assumption
is in place, then the inference is more than just a hypothesis or assumption.
Once the data base is closed off, the negative inference by default from the
data base is such that we can say that the conclusion is known to be true
(assuming that all the data in the data base are known to be true).
But most argumentation in everyday conversational exchanges is based on
the open world assumption, where we are uncertain whether the data base is
complete, or think it is incomplete, and we have to make a guess about con-
clusion to infer. It is precisely in such a guesswork situation that presumptive
reasoning becomes most useful. Hence the real practical importance in argu-
mentation of the form of reasoning often called the argument from ignorance,
outlined in section five below.
Presumptive reasoning works by making a guess, in the form of drawing a
conclusion and accepting it on a tentative basis, subject to possible retraction
as a commitment, should new argumentation alter the case. A presumptive
74 Douglas Walton
inference gives an arguer a reasons for accepting a conclusion, even though

that conclusion may later have be withdrawn if critical questions are asked in
the dialogue. Nevertheless, although such forms of inference are neither de-
ductively valid nor inductively strong, they do have a certain standing or
bindingness in a dialogue. The form of inference does have a logical structure.
It tells you that if you accept the premises, and the form of the argument is
structurally correct, then unless you can ask the right critical questions, you
must accept the conclusion. Such forms of inference called argumentation
schemes represent the logical structures of these kinds of arguments.
In the new dialectic, reasoning is defined as a chaining together of inferen-
ces. Reasoning can be used for various purposes. It is used in explanations as
well as in arguments. An argument is a use of reasoning to fulfill a goal of a
dialogue, of one of six basic types. All argument is to try to settle some issue
that has two sides. The present view is called dialectical because every argu-
ment is seen as a case of two parties reasoning together for some purpose.
What is primarily important in the dialectical system of evaluating arguments
is not (at least centrally) knowledge or belief, but something called commit-
ment. This term refers to the acceptance of a proposition by a participant in a
dialogue. Commitments do not have to be logically consistent with each other.
But if a proponent’s set of commitments are apparently inconsistent in a case,
a respondent can challenge that inconsistency, and call for some resolution or
explanation of the apparent inconsistency.
The new dialectic is amenable to formalization, as shown by Hamblin
(1970), and Walton and Krabbe (1995). But the formal structure required is
quite different from that of the traditional deductive propositional and quanti-
fier logics. The formalization is a game-like structure in which, there are two
participants, a proponent and a respondent. Each takes turn, making moves -
generally asking question, replying to question, and putting forward argu-
ments. The type of dialogue, as a whole, has a goal, and each participant has an
individual goal (or role). The rules define what kinds of moves are considered
legitimate for the purpose of contributing collaboratively to the goal of the
dialogue. In some types of dialogues, the individual goals of the participants
are opposed to each other. Other types of dialogue are not adversarial in this
same sense, and the participants are supposed to cooperate with each other and
help each other to work towards the goal together.
3. The Old Dialectic
The ancient art of dialectic was a philosophical activity in which two persons
took part. The questioner first poses a problem, the respondent chooses a
position, and then the questioner draws inferences based on the respondent’s
answers (Kapp, 1942, p. 12). Evidently the questioner’s aim was to draw out
conclusions that raise doubts about the respondent’s position, perhaps even
conclusions that appear to be inconsistent with the respondent’s position. But
the exact purpose of dialectic as an organized and goal-directed activity is
obscure to the modern reader of ancient philosophical texts. It is an art that fell
into obscurity after the fall of the ancient world.
Plato called dialectic “the art concerning discussions” (Robinson, 1953, p.
69), but tended to shift his meaning of the term to describe “the ideal method,
whatever that might be.” (Robinson, 1953, p. 70). Zeno of Elea was supposed
by many in the ancient world to be the inventor of dialectic (Kneale and
Kneale, 1962, p. 7), but it is not known exactly what Zeno had in mind. Accor-
ding to the Kneales (p. 7), Plato in the Parmenides refers to Zeno’s claim to
have written a book in which he draws out some absurd philosophical conse-
quences of another person’s philosophical view. This reference may have been
the basis of Aristotle’s later remark, quoted by Diogenes Laertius and Sextus
Empiricus, that Zeno was the inventor of dialectic. For a clearly articulated
explanation or analytical theory of what dialectical argument is supposed to
be, Aristotle’s account is probably the most useful source.
Aristotle, in On Sophistical Refutations (165b3-165b4), defined dialectical
arguments as “those which, starting from generally accepted opinions (endo-
xa), reason to establish a contradiction.” According to the Topics (100b22),
generally accepted opinions are “those which commend themselves to all or to
the majority or to the wise - that is, to all of the wise or to the majority or to
the most famous and distinguished of them. “ For Aristotle then, dialectic was
the use of reasoning to draw logical consequences from premises that are
generally accepted opinions. What kind of activity was this? It seems to be an
art of the gadfly. The best example we can seem to come up from the ancient
world with is the critical use of argumentation by Socrates, as portrayed in the
Platonic dialogues. Socrates questioned the opinions of those who thought
themselves to be wise, and were assumed to be wise by others or the majority.
He also probed and questioned conventionally accepted views. He often drew
contradictions and logical problems from these views, using logical reasoning.
His method was of asking a sequence of questions, where each question is
based on the previous answer given by a speech partner.
The ancients attached quite a lot of importance to dialectic as an art. In
addition to its use to teach skills of arguing, and for arguing in casual con-
versations, Aristotle even saw dialectic as being useful for questioning and
discussing the axioms or first principles (archai)of the sciences (Topics, 101b4).
This idea is simply not acceptable to the modern way of thinking since the
76 Douglas Walton
so-called Enlightenment. Pascal and Descartes argued that the model of good
reasoning should be that of Euclidean geometry, where theorems are rigorous-
ly deduced from self-evident axioms. This paradigm of scientific reasoning as
the all-powerful method of reasoning represents the modern way of thinking,
where it is assumed that, time and time again, science has proved “common
sense” to be wrong. The idea of some kind of reasoning outside science, being
brought to bear on science and used to question or criticize the assumptions of
science, where such critical argumentation commands rational assent, is alien
to our modern ways of thinking. The ancient art of dialectic had no place in
the modern way of thinking. When ancient logic was rediscovered in the
middle ages, it was Aristotle’s theory of the syllogism (deductive reasoning)
that came to dominate as the paradigm of logical argumentation. Aristotle’s
fallacies retained a toe-hold in the logic textbooks, but was never again taken
seriously as a central part of logic. When the Stoic logic of propositions was
formalized around the beginning of the twentieth century, deductive formal
logic eclipsed all other branches of the subject. Dialectic was a lost art.
4. Types of Dialogue
The new dialectic is built on the pragmatic foundation introduced by Grice, in

his famous paper on the logic of conversation (1975). According to Grice, an
argument should be seen as a contribution to a conversation between two
parties, and should be evaluated as a good (useful) argument or not, on the
basis of whether it made a collaborative contribution to the moving forward of
the conversation towards its goal. This new approach was tied in much more
closely to how we use and judge arguments in everyday conversational ex-
changes. The argument was now to be evaluated with respect to how it was
used for different purposes in different types of conversational exchanges. But
what are these different types of exchanges, and what are their goals? By
specifying the precise rules and goals of the different types of conversational
exchanges, The New Dialectic (1998) offered a new method for evaluating
arguments that could be applied to the informal fallacies (sophistical refuta-
tions) that held such a place of importance in the early applied logic of Aristot-
le.
By going back to the Aristotelian roots of logic as an applied, practical
discipline, the new dialectic brought out and formulated, in modern terms
adequate for state of the art argumentation theory, many of the leading ideas
expressed in the ancient works on dialectical argument that heretofore appea-
red obscure, and were for so long treated as peripheral in logic. For the first
time it becomes possible to apply objective logical standards of evaluation to
arguments in everyday conversational exchanges on controversial topics where

real conflicts of opinions exist.
The new dialectic offers a framework of rationality for judging an argu-
ment as correct or incorrect insofar as it has been used adequately or not in a
given case to contribute to goals of dialogue appropriate for the case. An
argument is judged to have been used in a fallacious way in a dialogue insofar
as it has been used in such a way as to impede the goals of the dialogue. A
dialogue, to use our generic term, or a conversation to use Grice’s term, is
defined as a goal-directed conventional framework in which two speech part-
ners reason together in an orderly way, according to the rules of politeness, or
normal expectations of cooperative argument appropriate for the type of
exchange they are engaged in. Each type of dialogue has distinctive goals,
turn-taking moves, and methods of argumentation used by the participants to
work towards these goals together.
Six basic types of dialogue are described in the new dialectic - persuasion
dialogue, the inquiry, negotiation dialogue, information-seeking dialogue,
deliberation, and eristic dialogue.
TYPES OF DIALOGUE
Type of Dialogue Initial Situation Participant’s Goal of Dialogue
Goal
Persuasion Conflict of Opi- Persuade Other Resolve or Clari-

nions Party fy Issue
Inquiry Need to Have Find and Verify Prove (Disprove)

Proof Evidence Hypothesis
Negotiation Conflict of Inte- Get What You Reasonable Sett-

rests Most Want lement that Both
Can Live With
Informati- Need Informati- Acquire or Give Exchange Infor-

on-Seeking on Information mation
Deliberation Dilemma or Prac- Co-ordinate Decide Best Avai-

tical Choice Goals and Ac- lable Course of
tions Action
Eristic Personal Conflict Verbally Hit Out Reveal Deeper

at Opponent Basis of Conflict
Table 1
78 Douglas Walton
Each of these types of dialogue is put forward in the new dialectic as a norma-
tive model which specifies broadly how a given argument should be used, in
one of these contexts, in order to be correct, or to be defensible against the
criticism that it is incorrect, erroneous or fallacious. The most central type of
dialogue, for the typical purposes of applied logic as it is taught in classrooms
today, in courses of critical thinking, is the persuasion dialogue. In this type of
dialogue, the proponent has a particular thesis to be proved, and the respon-
dent has the job of casting doubt on that thesis by raising questions about it.
In some instances however, the dialogue can be symmetrical. Both participants
have a thesis to be proved, and each has the aim of persuading the other to
accept his or her thesis. In a persuasion dialogue, each party takes the initial
concessions of the other as premises, and then by a series of steps, tries to use
these premises in arguments designed to persuade the other party, by means of
using rational argumentation, to give up his original thesis.
One very common problem is that during the sequence of argumentation,
the dialogue may shift from a persuasion dialogue to another type of dialogue,
say to a negotiation or a quarrel. Such dialectical shifts can be very confusing,
and are associated with many of the major informal fallacies. Another problem
associated with the job of evaluating many common arguments is to apply the
dialectical method to cases of mixed discourse, like sales pitches, political
debates, and legal arguments in trials. Such cases are characterized by the
mixing of two or more types of dialogue. They also frequently involve dialec-
tical shifts from one type of dialogue to another, during the same sequence of
argument.
5. Presumptive Reasoning
Recent concerns with the evaluation of argumentation in informal logic and

speech communication have more and more begun to center around nonde-
monstrative arguments that lead to tentative (defeasible) conclusions, based on
a balance of considerations. Such arguments do not appear to have structures
of the kind traditionally identified with deductive and inductive reasoning.
However, they are extremely common, and are often called “plausible” or
“presumptive,” meaning that they are only tentatively or provisionally accep-
table, even when they are correct. These arguments shift a weight of evidence
to one side of a balance, thus supporting a conclusion that was previously in
doubt. But such a weight can, as the argument continues, be shifted back to
the other side.
Presumptive reasoning is based on pragmatic implicatures drawn out by a
hearer on the basis of what a speaker’s remarks can normally be taken to imply
in the context of an orderly, cooperative dialogue. Hence presumptive reaso-

ning is more subject to contextual interpretation, and also more subject to
defeat (and error) than logical reasoning of the more familiar deductive and
inductive sorts. Perhaps for these reasons, presumptive reasoning has generally
been ignored in logic, and excluded from serious consideration as inherently
“subjective,” in the past. However (Walton,1996) shows that the inference
structures of presumptive reasoning are well worth investigating, and do help
us to critically evaluate argumentation of the kind that powerfully influences
people in everyday speech, on all kinds of controversial issues where presump-
tive conclusions are drawn.
What kind of support is given to a conclusion on the basis of presumptive
reasoning? The kind of support given is different from that given by a deducti-
vely valid argument or an inductively strong argument. For logicians long
accustomed to working with deductive and inductive standards of argument
support, the move to a third standard is not easy to make, especially when the
new standard typically gives only a weaker kind of support that is tentative in
nature, and subject to withdrawal in many instances.
One way to introduce the new idea is through the idea of a generalization.
In deductive logic, the universal quantifier is used to stand for a kind of ge-
neralization, ‘For all x, if x has property F then x has property G’, in which a
single counter-example defeats the generalization. This type of generalization
could be called absolute, in the sense that it is equivalent to ‘There are (absolu-
tely) no x such that x has F, but does not have G.’ In contrast, inductive ge-
neralizations, of the form ‘Most, many, or a certain percentage (expressed
numerically as a fraction between zero and one) of things that have property
F also have property G’. This kind of generalization is not absolute, because it
allows for a certain number of counter-examples (but not too many).
The kind of generalization characteristic of presumptive reasoning is based
on a type of generalization of the form, ‘Normally, but subject to exceptional
cases, if something has property F, it may also be expected to have property
G.’ This kind of conditional is subject to defeat in unusual or unexpected
situations that are not normal, or what one would normally expect. Our
confidence in it is tentative, because, as we find more out about a situation, it
can come to be known that it differs from the normal type of situation. For
example, we normally expect that if something is a bird, it flies. But in a parti-
cular case, we may find out that Tweety, a bird, is a penguin, or has a broken
wing. This new information will defeat the inference based on the normal
presumption that Tweety, since he is bird, is an individual that flies.
Many (statisticians, in particular) feel that presumptive reasoning can be
shown to be a species of inductive reasoning, perhaps so-called “subjective
80 Douglas Walton
probability”. This claim appears dubious, because encountering something

that is not normal (and is an exception) in a particular case is often a surprise,
and does not appear to be based on statistical regularities. However, it should
not be entirely rules out, perhaps, that some sort of statistical model of infe-
rence may be found that fits presumptive reasoning. So far, however, no
numerical formula for evaluating presumptive reasoning appears to have been
found, or at least any criterion that fits all kinds of cases.
Presumptive reasoning is highly familiar in computer science, where it is
frequently identified with abductive inference, or what is often called “infe-
rence to the best explanation”. But abductive inference does not appear to be
the same thing as presumptive reasoning, even though the two kinds of reaso-
ning appear to be closely related. The terminology on these questions is not
settled yet, and there are many different theories about how these two kinds of
reasoning are related.
6. Abductive Inference
What is called abductive reasoning in computer science, or “inference to the

best explanation” in philosophy, is a distinctive kind of inference that goes
from given data to a hypothesis that best explains the data. An example from
ordinary conversation is given by Josephson and Josephson (1994, p. 6) :
Case 2
Joe : Why are you pulling into this filling station?
Tidmarsh : Because the gas tank is nearly empty.
Joe : What makes you think so?
Tidmarsh : Because the gas gauge indicates nearly empty. Also, I have
no reason to think that the gauge is broken, and it has been a long time
since I filled the tank.
Classified as an argumentation scheme, we would say that the argument in this
case is an instance of argument from sign. The gas gauge indicating “nearly
empty”, or being low, is a sign that the tank is nearly empty. Giving such a
sign or indication is what this instrument is designed for. But we can also see
the reasoning as an inference to the best explanation. Tidmarsh considers two
possible explanations for the indication on the gas gauge. One is that the gas is
low. The other is that the gauge is broken. But, as he says, there is no reason to
think that the gauge is broken. So the best explanation, from what is known in
the case, is that in fact the gas in the tank is nearly empty.
Abductive reasoning is common in science (Josephson and Josephson,
1992, p. 7). Some would even argue that the typical type of reasoning to a
scientific hypothesis from given data is abductive in nature. According to

Harman (1965), when a scientist infers the existence of atoms, she is reasoning
from the best explanation of the given scientific data. Peirce (1965, p. 375)
classified all inference as falling into three classifications - deduction, inducti-
on, and what he called “hypothesis”, which corresponds to abductive inferen-
ce. Peirce, in a work called ‘The Proper Treatment of a Hypothesis’ (Eisele,
1985, pp. 890-904) described abduction as a kind of guessing, characteristic of
scientific reasoning at the discovery stage, which can save experimental work
by narrowing down the possible hypotheses to be tested to the most plausible
candidates. Peirce clearly identified abductive reasoning as a distinctive type of
inference that is important in science, and described it as a kind of plausible
reasoning or “guessing” (Eisele, 1985, p. 898). He also frequently wrote about
abduction as a kind of “explaining” process in ‘The Proper Treatment of a
Hypothesis’ (Eisele, 1985, p. 899). Peirce ( 1965, p. 375) defined hypothesis as
occurring in the following kind of instance : “where we find some very curious
circumstance, which would be explained by the supposition that it was a case
of a certain general rule, and thereupon adopt that supposition.” He gives the
following two examples (p. 375).
Case 3
I once landed at a seaport in a Turkish province ; and, as I was walking
up to the house which I was to visit, I met a man upon horseback,
surrounded by four horsemen holding a canopy over his head. As the
governor of the province was the only personage I could think of who
would be so greatly honored, I inferred that this was he. This was an
hypothesis.
Case 4
Fossils are found ; say, remains like those of fishes, but far in the interi-
or of the country. To explain the phenomenon, we suppose the sea once
washed over this land. This is another hypothesis.
Case 4 is clearly an example of a scientific kind of hypothesis, while case 3 is
the kind of inference to the best explanation that is so common in everyday
reasoning. These two cases illustrate very well how presumptive reasoning is
typically based on a kind of inference to the best explanation that Peirce called
“hypothesis” or abductive inference.
One thing that is very interesting about abductive reasoning is that it
combines the two functions of argument and explanation. An abductive infe-
rence may be used as an argument to support a conclusion, but the basis of
that support utilizes an explanation, or a series of explanations. We normally
think of argument and explanation as two different speech acts, or uses of
82 Douglas Walton
discourse. The purpose of putting forward an argument to a hearer is normally

to prove some proposition that is in doubt to the hearer. The purpose of
offering an explanation of to a hearer is to take a proposition that both the
speaker and hearer presume is true, and to make it understandable to the
hearer. These two speech acts are inherently different, but in abduction they
are combined. The explanation function is part of what supports or makes
possible the carrying out of the argument function. The distinction is often
explained in computer science as one of the direction of the reasoning. Nor-
mally, the reasoning in an argument moves forward. That is, the inference goes
from the premises to the conclusion. But in abductive reasoning, there is also
a backwards movement of inference. The conclusion is taken as a given data,
and then a search back is made to try to determine the best explanation for this
data.
The general form an abductive inference can be represented as follows -
compare (Josephson and Josephson, 1994, p. 14).
Form Abduct:
D is a collection of data.
Hypothesis H explains D.
No other hypothesis explains D as well as H.
Therefore H is plausibly true (acceptable).
Another kind of scientific reasoning that has been recognized as being based
on abductive inference is medical diagnosis, a species of argumentation from
sign that reasons from the given data to the best explanation. For example, in
the case of a diagnosis of measles, the physician might reason as follows.
Measles:
If the patient has red spots, then the patient has measles.
The patient shows red spots.
Therefore, the patient (plausibly) has measles.
The conclusion drawn in the Measles Case is tentative, and based on the
assumption that there is no better explanation of the red spots. The converse
of the major premise, ‘If the patient shows red spots, the patient has the meas-
les’ is not true, since showing red spots is only one sign of having the measles,
and it s not a conclusive sign, by any means. It is just one sign that can be used
abductively as evidence for measles, in the absence of any better explanation of
the red spots. At any rate, it is this sort of case analysis that is the basis of the
often-expressed theory that abductive reasoning takes the form of argument
called affirming the consequent. This analysis has its tricky aspects, however,
and the reader might be referred to (Walton, 1996b, pp. 264-281) for a fuller
discussion of the forms of abductive inference and argumentation from sign.
Josephson and Josephson (1994, p. 266) report that use of a seven-step scale
of plausibility values worked very well in modeling plausible reasoning in
medical diagnostic systems. Such use of numerical values might suggest that
plausible reasoning could be formalized using the probability calculus. But
they report (p. 268) that interpreting plausibility as probability just didn’t
work out very well. After going into the various technical possibilities in such
a modeling, they conclude (p. 269) that there is no “significant computational
payoff” in it. They conclude (p. 270) that there is “a need to go beyond proba-
bility”, and look in some other direction for a way to model plausible reaso-
ning. They conclude (p. 272) that is unlikely that plausibility, of the kind
characteristic of abductive reasoning, can ever be quantified, in the way that
probability is quantified in the probability calculus. To make “smart ma-
chines” that can reason plausibilistically, they conclude, we need to go in a
different direction.
How is one to judge, by some clearly defined standard, whether a particu-
lar presumptive inference is structurally correct or not, in a given instance?
This has become an extremely important question in recent times, and could
even rightly be called the central question of argumentation theory. The
problem is that while we in the field of logic are highly familiar with deducti-
vely valid forms of argument, and somewhat familiar with inductively strong
forms of argument, we appear to lack forms of argument corresponding to
cases of presumptive reasoning. However, there is a literature on what are
called argumentation schemes.
7. Argumentation Schemes
Argumentation schemes are the forms of argument (structures of inference)

that enable one to identify and evaluate common types of argumentation in
everyday discourse. In (Walton, 1996), twenty-five argumentation schemes for
presumptive reasoning are identified. Matching each argumentation scheme, a
set of critical questions is given. The two things together, the argumentation
scheme and the matching critical questions, are used to evaluate a given argu-
ment in a particular case, in relation to a context of dialogue in which the
argument occurred. An argument used in a given case is evaluated by judging
the weight of evidence on both sides at the given point in the case where the
argument was used. If all the premises are supported by some weight of evi-
dence, then that weight of acceptability is shifted towards the conclusion,
subject to rebuttal by the asking of appropriate critical questions.
84 Douglas Walton
One premise of an argumentation scheme typically takes the form of a

presumptive generalization, of the kind described above, to the effect that if x
has property F, then normally x will also have property G. For example, the
argumentation scheme for argument from sign is the following (Walton, 1996,
p. 49) :
A is true in this situation.
B is generally indicated as true when its sign, A, is true, in this kind of
situation.
Therefore, B is true in this situation.
The second premise is a presumptive generalization which says that if A is
true, then generally, but subject to exceptions, B is also true. But such a ge-
neralization is defeasible. It, taken with the other premise of the scheme, shifts
a weight of acceptance to the conclusion. But counter-argumentation in a case
may subsequently overturn acceptance of the argument by withdrawing that
weight, or even introducing new evidence that places a weight against it.
The list of presumptive argumentation schemes given in (Walton, 1996)
offers a useful, modern, accessible, systematic and comprehensive account that
the reader can use as an aid in interpreting, analyzing and evaluating natural
language argumentation in everyday conversations. Perelman and Ol-
brechts-Tyteca (1958) identified many distinctive kinds of arguments used to
convince a respondent on a provisional basis. Arthur Hastings’ Ph.D. thesis
(1963) made an even more systematic taxonomy by listing some of these
schemes, along with useful examples of them. Recently Kienpointner (1992)
has produced an even more comprehensive outline of many argumentation
schemes, stressing deductive and inductive forms. Among the presumptive
argumentation schemes presented and analyzed in (Walton, 1996) are such
familiar types of argumentation as argument from sign, argument from ex-
ample, argument from commitment, argument from position to know, argu-
ment from expert opinion, argument from analogy, argument from precedent,
argument from gradualism, and the slippery slope argument. Helpful examples
of each type of argumentation are given and discussed. In other recent writings
on argumentation, like van Eemeren and Grootendorst (1992),there is a good
deal of stress laid on how important argumentation schemes are in any attempt
to evaluate common arguments in everyday reasoning as correct or fallacious,
acceptable or questionable.
The exact nature of the relationship between argument from sign and
abductive inference is an interesting question. The measles inference above is
clearly an instance of argument from sign, and it is also an instance of abducti-
ve inference, or inference to the best explanation. Many instances of argument
from sign can also be very well analyzed as cases of inference to the best
explanation. Consider the example of argument from sign cited in (Walt-
on,1996, p. 47).
Case 5
Here are some bear tracks in the snow.
Therefore, a bear passed this way.
In this case, the premise is based on the observing of a particular shape and
appearance of imprints in the snow identified as a bear tracks. The best ex-
planation of the existence of such tracks would be (in the right context) that a
bear passed this way. But not all cases of argument from sign appear to fit the
inference to the best explanation format this well. For example, dark clouds
could be a sign of rain, but can we say that the rain is the best explanation of
the dark clouds? Not without some twisting and stretching, which leads one to
suspect that not all cases of argument from sign fit the abductive model. None-
theless, it is clear that abductive inference and argument from sign are very
closely related.
8. Arguments from Ignorance
A useful and encouraging aspect of the new dialectic is that it shows how the
presumptive argumentation schemes are the essential underlying structure
needed for the analysis of the traditional informal fallacies. Govier (1988, p.34)
has rightly stressed that the fallacies approach to argument evaluation is in-
complete, precisely because it needs to be based on a prior understanding of
the various types of good argument involved. She notes that the traditional
fallacies are most often based on good arguments that are not “propositionally
valid”, but nonetheless represent “ways of arguing well”. Not knowing exactly
what these “ways” are has been the biggest obstacle to the analysis of the
fallacies. This new dialectic represents a breakthrough by showing exactly
what these ways of arguing well amount to. Three important informal fallacies
that have been analyzed in depth by the dialectical method elsewhere can be
used to illustrate this point. These three are: argument from ignorance (argu-
mentum ad ignorantiam), hasty generalization (secundum quid), and argumen-
tation from consequences (argumentum ad consequentiam).
The argument from ignorance has the following simple form: it has not
been shown that propositions A is true, therefore it may be presumed that A is
false. To see how common this type of inference is, consider the following
dialogue.
86 Douglas Walton
Case 6
Bob : Is Leona Helmsley still in jail? She’s probably out by now.
Helen : Maybe she’s still in there, because we’d probably hear about it
if she got out.
How the argument from ignorance is used in this dialogue can be better un-
derstood by placing it in the following sequence of reasoning : (1) we would
probably hear about it if Helmsley got out (because the story would be widely
reported in the media), but (2) we haven’t heard about it, therefore (3) she’s
probably not out, i.e. (4) she’s still in there. Neither Bob nor Helen has any
definite evidence, one way or the other, yet Helen’s presumptive inference that
“Maybe she’s still in there.” seems to justify the drawing of a reasonable
conclusion by default. It is more of a conjectural than a solid conclusion, but
it would seem to be an exaggeration to call her argument fallacious. It is an
argument from ignorance that can be evaluated within the context of dialogue
in which it was used in the above case, as a nonfallacious use of presumptive
reasoning.
In the standard treatment of the logic textbooks, this type of argument has
traditionally classified as a fallacy. But in this case, and in many other cases
studied in (Walton, 1996a), the argument is evaluated dialectically as a pre-
sumptively reasonable argument that could function as tie-breaker in a ba-
lance-of-considerations case, thereby being used to draw a justifiable con-
clusion in a dialogue.
The argument from ignorance still continues to be condemned as a fallacy
by many commentators. Gaskins (1992) portrays it as a powerful and
all-pervasive argumentation strategy used especially influentially in modern
discourse to base conclusions on suspicion about all forms of authority. Ac-
cording to Gaskins, the skillful modern advocate uses the following form of
the argument from ignorance: “I win my argument unless you can prove me
wrong.” According to Gaskins, this fallacious form of argument has come to
dominate not only legalistic argumentation, but also scientific inquiry, and
modern moral disputes on public policy.
However, the argument from ignorance has had powerful defenders.
Socrates, in the Apology, was allegedly told by the Oracle at Delphi that he
was the wisest man of all, because he was the only one who admitted his
ignorance. Here a subtle form of ignorance, knowing what you do not know
– was used as a premise to support the conclusion that awareness of limitations
could be a kind of wisdom. Recently, Witte, Kerwin and Witte (1991) have
championed this Socratic attitude of trying to teach medical students an aware-
ness of the limitations of medical knowledge, instead of the more usual me-
thod of instruction, which they see as dogmatic memorizing of facts, as a

better method of medical education. They have set out a Curriculum of Medi-
cal Ignorance that advocates use of the argument from ignorance as a model of
medical reasoning.
Another common use of the argument from ignorance is in computer
science. It is a very familiar kind of reasoning in this area to search through a
knowledge base, find that a particular proposition sought for is not there, and
then presumptively conclude that this proposition is false. This is called de-
fault reasoning in computer science. A familiar example (Reiter, 1987) similar
to case 1would be the kind of case where you look at an airport monitor listing
all the stops on a flight between Vancouver and Amsterdam, and you see that
Winnipeg is not listed as one of the stops. When you infer that the plane does
not stop at Winnipeg, you are using an argumentum ad ignorantiam. But this
argument could be reasonable, assuming the convention that all stops are listed
on this monitor (what Reiter calls the closed world assumption).
The dialectical examination of the argument from ignorance, indicated by
the cases above, suggest that in many cases, it is a reasonable argument that has
legitimate and common uses in scientific and medical reasoning, and in the
kind of knowledge-based reasoning common in computer science. It can also
be shown to commonly used, and quite reasonable in many cases, in legal
reasoning, where, for example, it is reflected in the basic principle of criminal
law that a person should be presumed to be not guilty, in the absence of proof
of guilt.
9. Applying the New Dialectic to Cases
In the new dialectic, each case is unique, and a given argument needs to be
judged on the basis of the text of discourse available, representing the informa-
tion in that case. To evaluate the argument, we have to ask whether the closed
world or the open world assumption is appropriate. And we have to ask what
type of dialogue the participants were supposedly engaged in, as far as the
information given in the case indicates. If these facts are not determined by the
information given in the case, then the best we can do is to evaluate the case
hypothetically, based on assumptions that may hold or not, from what we
know about the case. Any assessment of this kind is contextual. We have to
look at the case as a whole, and then evaluate the argument in light of how it
was used in that case, as far as we can determine the relevant details of the case.
A dialectical assessment of a particular argument as used in a case appears
to be quite different from the usual use of deductive logic to assess whether a
given argument in natural language is valid or invalid. But it may not be as
88 Douglas Walton
different as is widely assumed. Even so, formal treatment would appear to be

more limited in the dialectical assessment, because in many cases of everyday
conversational exchanges, there is little or no explicit agreement between the
participants on exactly what type of dialogue they are supposed to be taking
part in. Political argumentation, for example, is typically mixed, being partly
persuasion dialogue, but also partly negotiation and eristic dialogue (as well as
involving information-seeking dialogue and deliberation, in many cases). Even
so, each type of dialogue does have a formal structure, and once the argument
is modeled in a given structure, formal techniques, of the kind currently in use
in AI, can be brought to bear on it. Nonexplicit premises can be articulated,
and so forth. Formalization is both possible and helpful, but since the data in
a case can be massive, how the formal structure is to be applied to the known
data in a case requires considerable preparatory work in interpreting what the
text of discourse should be taken to mean, in a given case.. So here is the
problem. It is not so much a problem of any difficulties of formalization, as it
is a problem of determining the body of data one takes to be the case.
The new dialectic has many uses. But among the foremost of these uses is
that of evaluating an argument found in a given text of discourse in a specific
case. Logic has long dealt with the evaluation of such arguments, but the
assumption has always been that the argument is just a designated set of pro-
positions -a set of premises and a conclusion - and that everything else about
it is trivial or unimportant from a viewpoint of its logical evaluation. In the
new dialectic, what is now important is not only the set of propositions, but
the context of dialogue in which these propositions have (presumably) used
for some purpose. Now each case needs to be looked at with respect to the
argument is supposedly being used - is it being used to persuade, to negotiate,
or to deliberate, for example? The same argument could be seen as quite
reasonable if it was supposed to a negotiation tactic, whereas it could be right-
ly judged to be fallacious if supposedly used as a contribution to a critical
discussion on some specific issue. For example, the same threat that is relevant
in a negotiation dialogue could be irrelevant if used as an argument if it is
supposed to be part of a critical discussion.
The first task in evaluating any given argument in a text of discourse is to
identify the sequence of reasoning using an argument diagram to pinpoint each
proposition, and to identify the inferences dawn from such propositions to the
conclusions that were derived. This task requires the filling in of nonexplicit
premises and conclusions. Hence the context of dialogue is vitally important
even at this stage, because judging how to charitably fill in such missing links
should be guided by dialectical factors, like the arguer’s commitments, as
known in the case. Once agreement is reached on what the premises and
conclusions of the given argument are supposed to be, as judged from the
textual and contextual evidence of the case, the next step is to examine each
inference in the chain of reasoning. Where weaknesses are found, the ap-
propriate critical questions need to be asked. Then finally, looking over the
whole sequence of reasoning exhibited in the diagram, the question of relevan-
ce needs to be raised. Where is the argument leading? Is it leading towards the
ultimate conclusion that is supposed to be proved in the type of dialogue
exchange of which it is supposedly a part? These are the questions relating to
dialectical relevance.
10. Uses of the New Dialectic
The new dialectic offers a practical method of identifying, analyzing and

evaluating authentic cases of everyday argumentation that does not require an
abstraction form the realities of discourse in a natural language setting. A new
option is offered. No longer do arguments have to be judged solely in relation
to standards of deductive and inductive reasoning. In the past, the dominance
of this more restrictive approach has led to a distorted view of many everyday,
presumptive arguments, often leading to the conclusion that such arguments
are somehow inherently defective, or even fallacious. In the new dialectic, the
different contexts of use of such arguments are taken into account - included
are such factors as the type of dialogue, the stage of a discussion, the com-
mitments of the discussants, and other factors that are specific to a case of
argumentation in which two speech partners are attempting to reason together
for some collaborative purpose. The targeting of the new dialectic to factors of
how an argument was used in a specific case gives a more practical way of
evaluating everyday argumentation. Judged by such practical standards, an
argument can be evaluated as weak in certain respects, and open to appropriate
critical questions, without being so badly off that it should be condemned as
fallacious, implying an error or defect that is beyond repair. As well as provi-
ding new tools for the analysis of arguments by teachers of courses on critical
thinking and informal logic, the new dialectic has other important fields of
application. It is clearly applicable to many common kinds of arguments, and
problems of argumentation, in fields like artificial intelligence, experts systems,
legal and medical reasoning, and use of evidence in academic research (not
excluding scientific argumentation).
The new dialectic is a framework for reasoning that strikes a healthy ba-
lance between descriptive empirical research on argumentation and normative
or abstract logical methods of setting standards for good arguments. Such a
balance, although lacking in the past, is healthy because neither the empirical
90 Douglas Walton
or the normative approach, by itself, can provide a method of argument eva-

luation that is both objective in standards and that fits the realities of real cases
of argumentation in a way that is practically useful. Real arguments in con-
versational exchanges are mixtures of different types of dialogue, and have
different standards of plausibility and rationality. Each type of dialogue has its
own distinctive goals, its own procedural rules, and its own standards of
burden of proof. An argument that could be appropriate and reasonable might
be highly inappropriate, or even fallacious, in another type of dialogue. Cases
also frequently involve mixtures of two or more types of dialogue, and shifts
from one type of dialogue to another. In the new dialectic, judging how an
argument was used in a given case is a contextual matter. Much depends on
what type of dialogue the participants were supposedly engaging in, when the
argument in question was put forward by one of the parties in the discussion.
The evidence on which to judge a case, therefore, must be sought in the con-
text of use of the argument. As shown in section nine above, sometimes there
is plenty of such evidence available in a given case, but in other cases, the best
that can be done is to make a conditional evaluation of the case, based on what
evidence is given in that case. Such conditional evaluations, despite their
hypothetical and incomplete nature, can, in many cases, be extremely helpful
in diagnosing the logical strengths and weaknesses of an argument.
Acknowledgements
I would like to thank Manfred Kienpointer for comments that enabled me to avoid several errors,
and helped me to improve the exposition of many points. This paper was supported by a research
grant from the Social Schiences and Humanities Research Council of Canada. I would also like to
thank the Canada-U.S. Fulbright Program and the Department of Communication Studies of
Northwestern University for funding that helped to support the work in this paper.
References
Aristotle, On Sophistical Refutations, Loeb Classical Library, Cambridge, Mass., Harvard Uni-
versity Press, 1928.
Aristotle, Topics, Loeb Classical Library, Cambridge, Mass., Harvard University Press, 1939.
Carolyn Eisele, Historical Perspectives on Peirce’s Logic of Science, vol. 1, Berlin, Mouton, 1985.
Richard H. Gaskins, Burdens of Proof in Modern Discourse, New Haven, Yale University Press,
1992.
Trudy Govier, ‘Ways of Teaching Reasoning Directly,’ Critical Thinking : Proceedings of the First
British Conference on Informal Logic and Critical Thinking, ed. Alec Fisher, East Anglia, Uni-
versity of East Anglia, 1988, 30-38.
J. Paul Grice, ‘Logic and Conversation’, in The Logic of Grammar, ed. Donald Davidson and
Gilbert Harman, Enrice, Carlifornia, 1975, 64-75.
Gilbert Harman, ‘The Inference to the Best Explanation’, Philosophical Review, 74, 1965, 88-95.
Arthur C. Hastings, A Reformulation of the Modes of Reasoning in Argumentation, Evanston,
Illinois, Ph.D. Dissertation, 1963.
John R. Josephson and Susan G. Josephson, Abductive Inference : Computation, Philosophy,

Technology, New York, Cambridge University Press, 1994.
Ernst Kapp, Greek Foundations of Traditional Logic, New York, Columbia University Press,
1942.
Manfred Kienpointner, Alltagslogik : Struktur und Funktion von Argumentationsmustern,
Stuttgart, Fromman-Holzboog, 1992.
William and Martha Kneale, The Development of Logic, Oxford, Clarendon Press, 1962.
Charles S. Peirce, Collected Papers of Charles Sanders Peirce, vol 2, ‘Elements of Logic’, ed.
Charles Hartshorne and Paul Weiss, Cambridge, Mass., Harvard University Press, 1965.
Chaim Perelman and Lucie Olbrechts-Tyteca, The New Rhetoric : A Treatise on Argumentation,
trans. J. Wilkinson and P. Weaver, 2nd ed., Notre Dame, University of Notre Dame Press, 1971
(First published, as La Nouvelle Rhetorique, in 1958).
Raymond Reiter, ‘A Logic for Default Reasoning’, Artificial Intelligence, 13, 1980, 81-132.
Raymond Reiter, ‘Nonmonotonic Reasoning,’ Annual Review of Computer Science, 2, 1987,
147-186.
Richard Robinson, Plato’s Earlier Dialectic, 2nd ed., Oxford, Clarendon Press, 1953.
Frans H. van Eemeren and Rob Grootendorst, Argumentation, Communication and Fallacies,
Hillsdale, N.J., Lawrence Erlbaum Associates, 1992.
Douglas Walton, Argumentation Schemes for Presumptive Reasoning, Mahwah, N. J., Erlbaum,
1996.
Douglas Walton, Arguments from Ignorance, University Park. Pa., Penn State Press, 1996a.
Douglas Walton, Argument Structure : A Pragmatic Theory, Toronto, University of Toronto
Press, 1996b
Douglas Walton, The New Dialectic : Conversational Contexts of Argument, Toronto, University
of Toronto Press, 1998.
Douglas N. Walton and Erik C. W. Krabbe, Commitment in Dialogue : Basic Concepts of Inter-
personal Reasoning, Albany, State University of New York Press, 1995.
Charles L. Witte, Ann Kerwin, and Marlys H. Witte, ‘On the Importance of Ignorance in Medical
Practice and Education,’ Interdisciplinary Science Reviews 16, 1991, 295-2.
92 Manfred Kienpointner
MANFRED KIENPOINTNER
Comments on Douglas Walton’s Paper: The New Dialectic: A Method of

Evaluating an Argument Used for Some Purpose in a Given Case
In his article, Walton outlines an alternative to both classical positivistic phi-

losophy and postmodern anti-rationalism (section 1). Starting from insights of
ancient dialectic and modern argumentation theory, he develops a framework
called ‘New Dialectic’ (section 2). New Dialectic is mainly concerned with the
most common kinds of everyday arguments, which is based on presumptive
reasoning rather than deductive or inductive logic. What is presumptive reaso-
ning? Walton’s answer is: ‘Presumptive reasoning works by making a guess, in
the form of drawing a conclusion and accepting on a tentative basis, subject to
possible retraction as commitment, should new argumentation alter the case’
(p. 73).
Walton then outlines characteristic features of the old dialectic developed
by ancient philosophers like Plato and Aristotle. He proceeds to criticize its
lack of practical applicability and Aristotle’s highlighting of common sense as
the basis of all reasoning, including scientific reasoning (section 3).
In section 4, Walton describes a central assumption of the New Dialectic,
namely, that argumentation has to be analyzed relative to the type of dialogue
underlying a conversational exchange. He establishes a typology of dialogues
with differing standards of plausibility and rationality (among them, the
persuasion dialogue, the inquiry, the negotiation dialogue, the eristic dialogue).
He also makes clear that real dialogues often are mixtures of two or more
types of dialogue and that a dialogue may undergo a dialectical shift from one
type to another.
Next, Walton provides a closer description of presumptive reasoning. It is
characterized by the fact that it only produces tentative conclusions, shifts the
‘weight of evidence to one side of a balance’, but is ‘more subject to defeat
(and error)’ and ‘to contextual interpretation’ than familiar deductive and
inductive types of arguments (p. 79). But he also states that ‘some sort of
statistical model of inference may be found that fits presumptive reasoning’ (p.
80), which would make it possible to see presumptive reasoning as a sub-type
of inductive reasoning. Walton also assumes that presumptive reasoning is
closely related to abductive reasoning (section 5).
In section 6, Walton deals with argumentation schemes as forms or structu-
res of presumptive reasoning. In Walton (1996), he has established a typology
Comments on Douglas Walton’s Paper: The New Dialectic 93
of 25 argumentation schemes which can be used as an aid in interpreting,

analyzing and evaluating everyday argumentation. Each argumentation sche-
me is accompanied with a set of critical questions which can point out potenti-
al weaknesses of inferences according to the schemes.
Walton then points out that the New Dialectic can provide a fruitful ap-
proach to study traditional informal fallacies. Recent developments in fallacy
theory have shown that many alleged instances of informal fallacies are defea-
sible arguments which show certain weaknesses, but are nevertheless accep-
table given certain contextual circumstances. It it therefore not acceptable to
dismiss all informal fallacies as completely unacceptable kinds of arguments.
Walton gives the example of the argumentum ad ignorantiam, which has the
following form: ‘it has not been shown that proposition A is true, therefore it
may be presumed that A is false’ (p. 86). He demonstrates that there are a
variety of cases and contexts where this type of argument can be used in a
plausible way, like other kinds of presumptive reasoning. One example is
default reasoning in computer science: if a particular proposition sought for in
a knowledge base is not there, one can presumptively – which does not mean:
deductively – conclude that it is false. Another example would be medical
reasoning, where even a Curriculum of Medical Ignorance has been developed
for teaching students an awareness of the limitations of medical knowledge
(section 7).
In the final section of his paper, Walton stresses that in the New Dialectic,
‘each case is unique and a given argument needs to be judged on the basis if the
text if discourse available’ (p. 88). This does not preclude formalization, but
creates the problem that in any given case a massive amount of data has to be
taken into account to answer the following questions: which kind of dialogue
underlies the argument? How can the sequence of reasoning be diagrammed?
How can the missing premises and conclusions be filled in? Which critical
questions can appropriately be asked? What is the relevance of the argument?
(section 8).
I now would like to give some comments concerning strenghts and wea-
knesses of Walton’s approach. In general, I would like to stress that Walton’s
New Dialectic is a highly interesting and fruitful synthesis of traditional
approaches and modern argumentation theory, including his many important
works on fallacies, dialogue games, argumentation schemes and other areas of
the study of argumentation (cf. e.g. Walton 1989, 1992, 1996).
The New Dialectic offers a promising and realistic framework for looking
at authentic instances of everyday argumentation, which does not require to
idealize away the real world complexities of discourse. With the important
concept of presumptive reasoning, New Dialectic is able to take a positive
94 Manfred Kienpointner
view towards the specific merits and deficiencies of everyday arguments. This
means that they are no longer solely judged in relation to established kinds of
reasoning like deductive or inductive resoning, which in the past often has led
to the conclusion that they are always somehow deviant or even fallacious.
Both for the analysis and the evaluation of everyday arguments it is in-
dispensable to take into account the various contextual factors like the type of
dialogue, the stage of the discussion, the commitments of the discussants etc.
Therefore, Walton’s insistance on the individual case is to be appreciated.
Moreover, the New Dialectic has many uses, as Walton rightly points out
(p. 89). It provides a set of tools for the analysis and evaluation of everday
arguments which can be used by teachers of informal logic and critical thin-
king courses. Furthermore, it has other important fields of application like
artificial intelligence, expert systems, legal and medical reasoning. To me, the
possibility of a broad practical application of a theoretical approach is far from
being of only secondary importance.
Finally, like other recent approaches in argumentation theory (e.g. the
Pragmadialectics of the Amsterdam school; cf. Van Eemeren/Grootendorst
1994), the New Dialectic tries to establish a fruitful balance between des-
criptive and normative perspectives on argumentation. This is to be praised
because neither purely empirical research nor ideal models of rational argu-
mentation can do justice to all aspects of argumentative discourse.
My more critical remarks concern only three minor points, which could be
easily integrated into the research program of the New Dialectic.
First of all, I think that Walton’s criticism of ancient dialectic is a bit too
harsh. He writes that ‘the exact purpose of dialectic as an organized and goal-
directed activity is obscure to the modern reader of ancient philosophical texts’
(p. 75). It is true that the dialogue rules listed in the eighth book of the Aristo-
telian Topics (155b3ff.) could not easily be transferred to everyday debates and
discussions. But they were intended as part of the teaching practice at phi-
losophical schools and not as principles for, say, political debates (cf. Moraux
1968).
Moreover, when Walton criticizes that Aristotle even recommended dialec-
tic as being useful for the discussion of the axioms and first principles of the
sciences (‘This idea is simply not acceptable to the modern way of thinking,
where it is assumed that, time and time again, science has proved ‘common
sense’ to be wrong’, p.5), one could defend the Aristotelian position by poin-
ting out that there is hardly another way of discussing axiomatic principles of
sciences apart from using common sense arguments.
Second, I would like to suggest that the relationship between presumptive
and abductive arguments is clearified in some more detail. Peirce’s description
Comments on Douglas Walton’s Paper: The New Dialectic 95
and examples (1965: 372ff.) show that his abductive arguments are very close
to presumptive arguments, and also follow some of the argumentation schemes
(e.g. argument from sign) which are introduced by Walton (p. 81). A closer
look at the similarities and differences of abductive and presumptive reasoning
could be set on the agenda of future work on the New Dialectic.
Third, I would like to make clear that in my book on argumentation sche-
mes (Kienpointner 1992: 230, 245) I remain completely open to the idea that
the classical deductive-inductive dichotomy might not be sufficient to deal
with all types of arguments which can be empirically established. Therefore, it
was not my primary intention ‘to stress deductive and inductive forms’ of
arguing (cf. Walton, p. 85).
References
Aristotle (1960): Posterior Analytics. Ed. and transl. by H. Tredennick/Topica. Ed. and transl. by
E.S. Forster. London: Heinemann.
Van Eemeren, F.H./Grootendorst, R. (eds.)(1994): Studies in Pragma-Dialectics. Amsterdam:
SicSat.
Kienpointner, M. (1992): Alltagslogik. Stuttgart: Frommann-Holzboog.
Moraux, P. (1968): La joute dialectique d’après le huitième livre des Topiques. In: G.E.L. Owen
(ed.)(1968): Aristotle on Dialectic. The Topics. Oxford: Oxford UP. 277-311.
Peirce, Ch.S. (1965): Deduction, Induction and Hypothesis. In: Ch. Hartshorne & P. Weiss (eds.):
Collected Papers. Vol. II: Elements of Logic. Cambridge, Mass.: Harvard Univ. Press. 372-388.
Walton, D. N. (1989): Informal Logic. Cambridge: Cambridge Univ. Press.
Walton, D.N. (1992): The Place of emotion in Argument. University Park: Pennsylvania State
Univ. Press.
Walton, D.N. (1996): Argumentation Schemes for Presumptive Reasoning. Mahwah, N.J.: Erl-
baum.
96 Christopher W. Tindale
CHRISTOPHER W. TINDALE
The Authority of Testimony
‘Be wary of the argument from authority.’

– Sokal and Bricmont (1998:178)
Introducion
One of the older and, perhaps, more ill-traveled forms of argument is what is
known as the ‘appeal to authority’, the simple strategy of using the say-so of
a recognized authority (person, source, or institution) as evidence in support
of a claim or belief. I say ‘ill-traveled’ because for many people the ‘appeal to
authority’ has been judged inherently fallacious, as implied in the warning
from Sokal and Bricmont above. But there has developed in the textbook
treatments a general sense that such appeals can be reasonable if they meet
certain criteria. To take only one such textbook account as an example: the
authority’s credentials must be stated; those credentials must be relevant to the
claim/case in question; the authority must be free from bias in the case; there
should be wide agreement among relevant experts; and the claim must concern
an area of knowledge wherein such consensus would be possible (Groarke,
Tindale, & Fisher, 1997: 275)1.
Such criteria work well up to a point (usually helping students develop
reasoning skills and recognize the better ways to proceed); but they sketch
over or simply avoid some of the more difficult questions that can arise, for
example: Should we distinguish between authorities and experts? How do we
appraise competing claims of ‘authorities’ in cases where there is obviously no
consensus, but consensus might be in the process of being formed? And how,
most importantly, do we deal directly with authorities and evaluate their
responses?
The last question in particular points to an important problem involving
the impact of authority on argumentation that takes us beyond the simple
‘appeal to authority’, narrowly defined. While the layperson may find herself
listening to an arguer who appeals to an authority and then making a judge-
ment about the appropriateness of that authority, other relevant discussions of
authority will often involve cases where the layperson is directly confronted
1 Similar accounts can be found in a tradition stretching back to Bentham’s Handbook of Political
Fallacies (1816).
with the authority in person, with no intermediary ‘appealer’. Some of the

basic conditions for assessing an authority can still apply here, but the layper-
son also needs to be able to assess the authority’s responses, and there the gulf
between them may seem impenetrable: whether we refer to authorities or
experts, they are by definition more knowledgeable on the question at hand
than is the layperson, who may simply be reduced to silence.
An apparent paradigm case of this could be seen in the following: In a
courtroom, the layperson takes the form of a jury member who must evaluate
the testimony of an expert witness. But how is this to be done? What kind of
background experience can assist the juror?
I raise this case because it is not the kind of case that principally interests
me in this paper. In this case, obviously the ability of the juror to appraise the
testimony of the expert witness depends on the nature of the witness’ testimo-
ny. If he/she testifies to general matters of perception, like what can be seen in
certain lighting or what can be lifted by a woman with a certain build, then the
juror can rely on notions of plausibility and likelihood in judging the testimo-
ny. If the expert testifies on technical knowledge, then the juror is in a weaker
position. But in fact, this ‘paradigm’ case is, in this respect, artificial in the
sense that the expert witness in such a trial has a burden of proof upon her or
him to make the testimony understandable to the juror. Properly conducted,
the structure of the trial should facilitate that understanding.
More difficult are the ‘everyday cases’ where we consume the advice of
experts and authorities without those experts feeling compelled to explain
themselves. Two things help here, and these are what I intend to explore
throughout the paper: (i) Expert knowledge must in some way conform to
objective conditions. In fact, increasingly in accounts of the argument form,
we see an implicit recognition that authority is subordinate to objective evi-
dence. But this evidence, this ‘objectivity’ has its source in general testimony
as a primary way of knowing. (ii) Secondly, we can consider what kinds of
questions need to be asked by the layperson of both the expert and the ex-
pert’s testimony. And, more importantly, how the gradual understanding of
such questions and the ability to employ them factor in the development of a
certain type of autonomous, competent knower.
The Nature of the Problem
The power of authority permeates all sectors of life and any appreciation of
history indicates that this is no modern feature but one rooted in human
development and culture. Of course, there is an important difference between
authority that is derived from political or social (or institutional) standing, and
that which is derived from knowledge. But at a deeper level, these have often
been intertwined as those ‘who knew’ were in a position to assume other types
of authority. Along such lines Alberto Manguel speaks of ‘the argument of
authority’ (1996:270), whereby a classical text and its author were often dee-
med infallible (Aristotle is an obvious example). Thus it became a common
argumentative move to attribute one’s opinions to others, thus rendering them
immune to criticism.
Latter-day disdain for the ‘argument from authority’ may well be rooted in
such early practices, even though we may recognize that the ‘argument from
authority’ as a strategy for providing contributory evidence for a claim, and
the ‘argument of authority’ to foreclose debate on a question are by no means
synonymous. Furthermore, modern rejections of the argument form may well
be based on apparently reasonable grounds. Consider two examples, one from
literature and another from science.
In E.M. Forster’s 1934 obituary of art critic Roger Fry, he eulogized him in
the following terms: ‘What characterized him and made him so precious in
twentieth-century England was that, although he was a modern, he believed in
reason....[He] rejected authority, mistrusted intuition’(1936:50-51). As Forster
explains this, Fry, when presented with an appeal to authority would express
a preference to rather think about it. Hence, the contrast of appeals to authori-
ty and ‘reason’. There is an irony in this of course, since Forster’s plea for the
mistrust of authority is based in part on the appeal to Fry’s practice. But the
point is well made that accepting the views of an authority somehow involves
circumventing the task of thinking through the matter oneself, and if one could
do this thinking, it would be preferable. I emphasize this phrase because it
captures the crux of the problem that should appear: can we (laypeople) think
through the matter in away that obviates any appeal to authority? It is after all
the nature of the authority as the possessor of some particular knowledge or
experience which distinguishes them from the layperson in the matter at hand.
If Fry can equally (or sufficiently) think through the matter for himself, then
the point of the appeal is made redundant in his case.
The point is illustrated further in the domain of science. By way of clarify-
ing their invocation to be wary of the argument from authority1, Sokal and
1 This is probably the place to seriously consider one of the questions raised earlier, whether we
should draw a distinction between ‘an authority’ and ‘an expert’, because here in the domain of
science it may be thought to matter. Popular science writer, Carl Sagan insisted, for example, that
because authorities have made mistakes in the past and will continue to do so, ‘[a]rguments from
authority carry little weight’ (1995:210). Or, as he also puts it: ‘in science there are no authorities;
at most, there are experts’ (Ibid.). We might legitimately wonder why experts would not succumb
to the same fallibility that seems to plague authorities. But, this aside, ‘expert’ has a narrowness
that in cognitive, and related argumentative, matters seems to have a clarity that is more helpful
Bricmont recommend following ‘the best of the natural sciences’ methodologi-

cal principles, starting with this one: to evaluate the validity of a proposition
on the basis of the facts and reasoning supporting it, without regard to the
personal qualities or social status of its advocates or detractors’ (1998:178).
This remark has both a positive and negative aspect to it. Negatively, it appears
to interpret the ‘argument from authority’ in its non-cognitive, institutional
sense, where the authority carries weight not necessarily because of what they
know but because of their position or some feature of their character. Yet at
the same time, they seem aware of the real nature of such an appeal because,
positively, they advocate a personal investigation of the facts and reasoning
supporting a proposition. Again, a commendable sentiment, but one which
founders on two fronts.
In the first instance, it does not seem to conform with actual scientific
practice. Take as an example the investigation of the rock sample ALH84001,
deemed to be a martian meteorite. Scientists working on the case used various
forms of reasoning, from causal to analogical reasoning, in arriving at the
conclusion that phenomena connected to the rock were ‘evidence for primitive
life on Mars’ (McKay, et.al.,1996). But they also relied upon the ‘appeal to
authority’ in its cognitive sense. While the thrust of their experimentation was
on the chemical make-up of the rock, they based their research on the assump-
tion that the rock was of martian origin and supported this assumption by
citing the work of other scientists as reported in a paper by a further scientist
in the journal Meteoritics. The correctness of that earlier research is crucial to
the acceptability of McKay et.al.’s conclusions about the rock sample. If this
case is representative of the kinds of activities that go on in scientific practice,
and I submit that it is, then it illustrates that not only do scientists appeal to
authorities, or experts, they have to make such appeals in order to draw infor-
mation from cognate fields and avoid replicating studies that it would be
impractical to repeat.
The second problem relates back to the Roger Fry case. Here it is not a
matter of whether laypeople should investigate for themselves rather than
relying on the say-so of experts; it is a question of whether they able to do
otherwise than rely on authorities. Charles Willard (1990) captures what may
be considered the crux of the problem of the layperson’s confrontation with
authority and what is troubling in both the Fry and Sokal/Bricmont recom-
mendations.
than the broader ‘authority’. Experts do not possess the wide authority of an Aristotle that
transcends fields, but invariably have a more particular or specific knowledge of a limited field of
inquiry or experience. While continuing to refer to the argument form ‘argument from authority’
I will understand it to refer to experts in this narrower sense.
As Willard presents the problem of authority (and twentieth century

dependence upon it), it is not so much that we have difficulty assessing the
claims made, but more that the predisposition to acquiesce to authorities
threatens autonomy.
Willard argues that public decision-making is firmly rooted in a reliance on
authorities: ‘Public decision-making doesn’t use knowledge, it uses testimony
– a tapestry of positions maintained by authoritative representatives of
knowledge domains who presumably bridge the gap between disciplines and
public decision-making’(16). But this poses a real dilemma, because while
Willard believes that ‘it is presumptively rational in a consensualist world to
argue from and acquiesce to authorities’ (11), this has the tendency to under-
mine personal autonomy. This is seen in the problems of appraisal. Willard
makes the important observation that we are not concerned here with the
appraisal of whether experts are correct – that is the job for experts themselves.
We are concerned with deciding how expert testimony should be taken; whet-
her it’s the direct attesting to some fact or experience, or the challenge of one
expert by another.
The extent of this concern emerges when he turns to the responses that
have been made by argumentation theorists and critical thinkers. It has beco-
me the norm for such scholars to argue that the individual must learn to de-
fend themselves against the ‘tyranny’ of authority by developing certain skills.
Willard quotes Mickunas (1987) as typical of the kind of thing advocated by
such people:
While submitting to the authority of what has been attained and established by
scientific enlightenment – the formal and technical disciplines – the person should
reach through education, a level of understanding, independent judgment, and
creativity, at which the scientific-technical values and norms are established.
Only at this level the person can become competent to understand the rules, and
their embodiment in concrete social life, and be in a position to pass an autono-
mous judgment and engage in free debate. (Mickunas, 1987: 336-337; cited in
Willard: 19)
It will be recognized how well this conforms with the sentiments expressed by
Fry and Sokal & Bricmont: the ‘problem’ of authority is that it subverts the
power and opportunity of the individual to think/decide for her- or himself.
But as Willard suggests, the real problem lies in this very kind of thinking
about authority because it seems to advocate an ideal which is virtually unat-
tainable. Willard writes: ‘If it puts a burden on individuals that no individual
in the late 20th century can discharge, then it is an empty posture – a kind of
arm-waving in the face of the problem of modernity’ (19). In this sense, we see
in the appeal to experts a challenge to the concept of autonomy which has
significant implications for what we mean by ‘competence’ (for the layperson)

and is an indictment of the activity of most critical thinking advocates.
So we face a dilemma: expertise is indispensable, but with its rise comes an
almost commensurate loss of autonomy. This is a dilemma that should be
astutely felt by those working in argumentation and Informal Logic. As Wil-
lard notes:
Argumentation and Informal Logic pedagogies won’t come to much unless based
on a coherent stance toward authority. They are premised on rhetorics of mass
enfranchizement that often do not square with their commitments to acquiescen-
ce to expertise, and – more important – they build a naive picture of the compe-
tent citizen (20).
Implied here is a considerable challenge. But it also announces an important
shift of focus in discussions of the ‘appeal to authority’ argument, away from
the authorities or experts and onto the layperson who must operate in their
midst. Argumentation, with its employment and study of argument forms, is
not a set of principles or tools which leave the user unaffected. Reasoning,
specifically evaluating arguments, affects the evaluator. And, it seems, the
‘appeal to authority’ is an argument form that brings this into focus, because
it allows, calls for, an entry into considerations of the competence and autono-
my that should underlie effective argumentation.
Expertise and Experience
I want to take this pursuit of the competent layperson into an examination of

Douglas Walton’s treatment of the ‘argument from authority’ (1997), where
we can begin to develop the two threads of this paper. Walton takes a dialecti-
cal approach to argumentation (1995, 1998), focusing principally on the con-
texts in which a two-person dialogue frames the discourse. Thus he identifies
different types of dialogue that can be involved in an exchange (1998), and has
defined a fallacious argument as one that involves the inappropriate shift from
one dialogue type to another (1995: 255-56). His treatment of the argumentum
ad verecundiam, or ‘appeal to authority’, falls squarely within this perspective.
He stresses the dialectical nature of discourses involving laypersons and ex-
perts, and argues that ‘[t]he dialectical factor is a subtlety that appears to be
overlooked by the rather simplistic accounts of the logical form’ of the argu-
ment (1997: 153). Likewise, he discovers that at the heart of the distinction
between legitimate and fallacious appeals to authority ‘is a dialectical failure...
[where] the asking of appropriate critical questions is blocked by making it
seem improper’(249). In Walton’s treatment, the common conditions for
assessing the argument form that were noted above are filled out with a series
of critical questions related to each condition. These questions are the means
by which the layperson can interrogate the expert, and it is behind this set that
I will later want to look to consider what kind of competence must be held by
a person capable of asking such questions. In all fairness, this is to lay the
stress where Walton does not. He writes as if his interest is in the argument
form as it sits out there between expert and consumer (or appealer and consu-
mer). His critical questions have a certain “objectivity” to them, even if they
are used only in specific contexts. But they still raise the question of whether
there is a developing model of a competent individual behind Walton’s analy-
sis.
Walton’s treatment includes a discussion of the history of this argument
form, and it is within such a history that we can find the tension between
expertise and experience. Not surprisingly, Walton’s history accentuates the
dialectical perspective as it can be uncovered from Aristotle through to Locke.
Strictly speaking, the ad verecundiam begins with Locke in the Essay Concer-
ning Human Understanding (1690), where he identifies it as backing one’s
‘tenets with such authorities [he] thinks ought thereby to carry the cause’
(cited in Walton, 1997: 53). Through a detailed analysis of Locke’s account,
Walton comes to describe his version of the ‘argument from authority’ as ‘a
characteristic type of exchange of argument: moves between two participants
(or three, if you include the authority) in a dialogue framework’ (61). Hence,
it becomes clearly a dialectical account, with the principal participants being
the one making the appeal, and the one receiving it (being asked to accept it).
In spite of this ‘origin’ in Locke, the strategy predates him in a number of
forms, particularly as a feature of Aristotle’s collection of dialectical proposi-
tions (41). In fact, these two serve as important moments in the history becau-
se, although Walton details other relevant accounts in those of Plato, Galileo,
the Port-Royal Logic and Bentham, the Lockean account, as he conceives it,
that represents a return to the dialectical model that has been lost since Aris-
totle and anticipates what Walton himself will do with the argument form.
Another important feature of this history, perhaps moreso for our purpo-
ses, is the way it acknowledges a clash between the acceptance of dogmatic
appeals to authority and other criteria of evidence, principally scientific experi-
mentation (34). Such a tension can even be seen in the early accounts of Plato
and Aristotle.
Plato’s treatment of experts is far more ambiguous than Walton’s discussi-
on allows. The ‘one who knows’ is often a central concern. But while the
expert is lauded in Plato’s dialogues (with the ultimate expert appearing in the
form of the philosopher ruler), his early philosophy is essentially an attack on
certain experts.1 In fact, it is difficult to accept Walton’s claim that the Socratic
dialectic indicates Plato’s belief that the layperson can critically assess the
expert. Those who are critically assessed in these early dialogues are not ex-
perts. The competent layperson, if such we can imagine Socrates (and this
would be a difficult model to emulate), assesses the non-expert in search of the
expert, and consistently fails to succeed in the quest. An equally plausible
reading of these dialogues would suggest that their intent is not to allow
Socrates to assess the expert, but to provide the alleged expert with the oppor-
tunity to assess themselves and discover their failings as experts. Thus the
rationale of the method is not to benefit the layperson but to guide the ‘expert’
in self-critique. On the other hand, in the Socratic figure we do have a model
of epistemic competence, tied to the asking of critical questions, and it is
interesting (even surprising) that Walton does not make more of this. This is a
moment in the history that is left and not carried forward, in spite of its appa-
rent relevance to Walton’s own account.
In the ancient model that is chosen, that of Aristotle, we find another hint
of trouble that deserves to be explored. Walton (40-41) identifies in Aristotle’s
account an appeal to both the wise and the many. As he notes, these later will
become the appeal to authority and the ad populum. But this should lead us to
consider the ad populum as a type of appeal to authority. That is, what is the
relationship between these two argument forms? Certainly, there is a rela-
tionship that interests Aristotle. In the Topics, the wise, or philosophers, are
identified as the most notable of men. One can assent to their views ‘if it be
not contrary to the views of most men’ (I.104a). This is an explicit condition
that immediately limits the authority of the wise. It also flies in the face of
Socrates’ insistence that the one who knows should be trusted over the many.
But here Aristotle seems to provide a concession to the much broader expe-
rience of the majority. While there are certainly problems in such a suggestion,
it proposes an empirical base for evidence that will characterize later accounts.
This is also the first hint of the tension between the expert and direct ob-
servation that becomes more pronounced in later philosophers.
In Walton’s history, it is Galileo who first stresses this tension: ‘History
had reached a point of clear conflict between the institutional authority that
prevailed at the time and the claims of empirical science’ (46). Galileo
(1967:128) indicates clearly a preference for looking at how an authority, like
Aristotle, arrived at a conclusion, rather than simply accepting the conclusion
on the authority’s say-so, where this can be determined. In such a case, the
layperson’s access to relevant immediate experiences is superior to the word of
1 This is also to question Feyerabend’s claims(1987) about the ‘cult’ of the expert in Plato’s work.
the authority. The authority becomes a substitute for that experience. As

Walton notes, ‘...the say-so of any authority could come into conflict with
scientific findings, based on observation and experiment’ (48). But whose
observation is involved here? This very same point is emphasized in Arnauld’s
The Art of thinking (or Port-Royal Logic) (Walton, 1997: 51) and even in
Locke1:
Of course, Locke was an empiricist, and his general viewpoint reflects the empiri-
cist orientation of modern experimental science. In any conflict between a con-
clusion drawn from the opinion of an authority and a finding drawn from the
“light arising from the nature of things themselves,” the latter kind of ad judicium
argument wins. (54)
This appears to give us a tension between expertise and observation/experi-
ence. But it simply reduces to two sets of experience, one from the expert and
another from another source. That source might be the layperson, or Aristot-
le’s majority, but in either case it sends the individual back to her or his own
resources, to their own observations. In such cases, we might infer, the appeal
to authority or expertise has no sway because it produces evidence that is no
better than, and may be in conflict with, what others’ experience. However, an
alternative tact to take is to ask whether in fact the expert might not sometimes
be in a better position to interpret ‘direct evidence’. That is, someone’s status
as an expert qualifies them to speak on certain types of experience with an
authority that the layperson does not have. The non-knowledgeable indivi-
dual, after all, can be mistaken. And in the relevant areas, when there is a
conflict with the expert’s experience, the layperson is more likely to be in error
than the expert. But when the expert conflicts with general observation, then
we may have a problem.
A related concern that derives from this history involves the way in which
an appeal to an expert constitutes ‘evidence’ for a position. What is meant here
by evidence? For example, Walton tells us (72) that to be reliable, an appeal
must be based indirectly on objective evidence that can presumably be produ-
ced if requested. This puts “objective evidence” prior to the word of the
authority. But in confirming an empirical base, this move also gives testimony
a prior place, as we will see.
Authority and Testimony
The tension between the say-so of experts and the evidence of experience, and
in particular the question of whose experience is at stake, deserves further
exploration.
1 A similar point is identified by Walton in Bentham’s account (1997:57).

This question is involved, for example, in Hume’s appeal to ‘uniform expe-

rience’. In his ‘Essay on Miracles’1 the tension is expressed between reports of
miraculous events (testimony) and uniform experience:
It is no miracle that a man, seemingly in good health, should die on a sudden:
because such a kind of death, though more unusual than any other, has yet been
frequently observed to happen. But it is a miracle, that a dead man should come
to life; because that has never been observed in any age or country. There must,
therefore, be a uniform experience against every miraculous event, otherwise the
event would not merit that appellation. (1964: 93)
Earlier (91), Hume had observed the lack of any a priori connection between
testimony and reality. Rather, we find that they customarily conform to each
other and this is why we find testimony credible. But our understanding of
reality, presented as ‘uniform experience’ must itself be based upon testimo-
ny.2 Hume tests one type of testimony – reports of the miraculous – against
another – uniform experience. For example, countering the testimony that
Lazarus returned to life is ‘our’ uniform experience that the dead do not return
to life (Bitzer, 1998: 192). But this reasoning has been shown to involve a
vicious circularity: ‘the experience upon which our reliance upon testimony as
a form of evidence is supposed to rest is itself reliant upon testimony which
cannot be reduced in the same way’ (Coady, 1992: 81). As a type of inductive
inference, the paradox of induction being its own standard of legitimacy is
transferred to testimony. Hume is criticized for treating reasoning from testi-
mony as one kind of reasoning from experience (analogy being another, for
example) without differentiating its nature (Bitzer: 217). Moreover, a stronger
criticism, attributed to John Leland (Bitzer: 211-12) argues that it is a mistake
to reduce reasoning from testimony to reasoning from experience because the
latter concludes with expectation, conjecture or prediction, while the former
results in a judgment of a fact’s merits. Experience tells me what to expect;
testimony tells me whether something is the case. As Bitzer (212) notes, it is
unclear whether Hume did in fact conflate these two types of reasoning. But
the distinction in the criticism is not a fine one. If the difference between the
two is that between general inductions centered somehow on my own expe-
riences and reports that I gather from others, it overlooks the extent to which
‘my’ general inductions are dependent on a ‘uniform experience’ – that is,
essentially, reports gathered from others. Because I do not observe ‘uniform
1 Section 10 of Philosophical Essays and section 10 of An Enquiry Concerning Human Under-

standing. Page references are to the Enquiry in volume II of the Philosophical Works (1964), edited
by T.H. Green and T.H. Grouse.
2 It is common experience, of course, that validates the laws of nature that miracles are deemed by
Hume to violate.
experience’, I infer it from what I am told and read. Some reports are of such
a heightened status that I use them to judge facts. But they are not so different
in kind from the general reports which allow me to judge what to expect or
whether something could have been predicted. What Hume’s account masks
is what he recognizes elsewhere: testimony is our fundamental type of reaso-
ning. My ‘experience’ is drawn from others’ accounts.
Part of our failure to appreciate this stems from the way testimony is
interwoven with other sources of evidence, like perception, memory and
inference (Coady, 149). What one perceives is in part a function of what one
expects; and what one expects is in part a function of what one has been told.
It is also hindered by the modernist assumption that, epistemically, the self is
somehow independent and self-reliant. Yet testimony is not only my most
fundamental link to reality, essential for making sense of what I observe; it also
connects me to society. If we assume the absence of innate knowledge, we
need others for linguistic competence (Audi, 1997: 416). In a similar vein,
Coady (85-95) develops an extended example of a Martian community to
show how any ‘community’ relies upon testimony statements in order to exist.
On the other hand, the assumption of the epistemically isolated self is fraught
with problems. When asked how such a self would acquire knowledge, the
expectation is that the individual’s perceptions would be fundamental. Coady
invokes Condillac’s imaginary case of a marble statue1 gradually being brought
to life to explore such a scenario. Starting with olfactory sensations, the statute
slowly becomes fully attentive. Without overlooking its merits, Coady ob-
serves the way prejudices and assumptions of a certain cultural and philoso-
phical predilection inform the thought experiment. Importantly,
A statue does not belong to a species nor does it have a naturally communal life.
But if we shift focus in this way it will be natural to see our starting-point as
encompassing our knowledge and not exclusively my knowledge. There will be
no problem of the epistemological priority of my perceptions over our percep-
tions, though there will be plenty of room for discrepancies. (Coady, 150)
From the egocentric perspective, I assume things that cannot be the case: that
uniform experience is somehow rooted in and corroborated by my own,
rather than vice versa; that we come each to know the world and then turn to
communicate our understanding to others and contribute to the common
fund, rather than that our understanding is predicated upon the common fund
of experience. And testimony is our link to all this. This begins to suggest that
the reason appeals to empirical facts have historically been preferred over
expert testimony, when the two conflict, is because those empirical facts arise
1 This is drawn from Condillac’s A Treatise on the Sensations.

from general testimonial evidence and, hence, testimony is a more fundamental

source of knowledge than expert opinion.
Of course, reliance upon testimony alone would be naive. The point so far
has been to indicate the extensive dependence that we should have upon
testimony in argumentation, far beyond the more limited case of the ‘appeal to
authority’. But something else must be required to make this testimonial
evidence reliable, particularly if it is to form the foundation for our competent
layperson.
As we have seen, testimony, while fundamental, does not stand alone as a
source of belief or knowledge. It functions in combination with perception,
inference, and memory (Coady: 147; 175). In John Locke’s treatment of testi-
mony (1959/1690: Bk.4, Ch.15) he remarks that when one hears reports of
things which have been previously observed by the hearer, then we are ‘dispo-
sed by the nature of the thing itself to assent to it’. There is, as it were, a bur-
den of proof favoring the speaker because what is testified to conforms readily
to the hearer’s experience. But when what is reported falls outside the hearer’s
experience, testimony may cause belief if ‘the relators are more in number, and
of more credit, and have no interest to speak contrary to the truth’. However,
in the famous case of the king of Siam who is told by the Dutch ambassador
that water freezes in cold weather in Holland, Locke concedes that testimony
so contrary to experience ‘will scarce be able to find belief’.
Several things can be noted here that appear in later accounts of conditions
for judging testimonial and expert evidence: the number of testifiers can be a
factor; the perceived interests of the testifier are to be considered; and, impor-
tantly, the extent to which the report is contrary to the hearer’s experience will
weigh in the hearer’s judgement. As Locke implies, human psychology is such
that we cannot believe what falls too far outside our experience.
Hume does give more power to testimony. As he takes up the case it
involves an Indian prince (1964: 92). The freezing of water places nature in a
situation quite unknown to the prince such that even the rules of analogy
cannot help him. Still, Hume notes that it ‘requires a pretty strong testimony,
to render it credible to people in a warm climate’ (92fn). This is because it is
not miraculous, only extraordinary, since it is not
contrary to uniform experience of the course of nature in cases where all the
circumstances are the same. The inhabitants of SUMATRA have always seen
water fluid in their own climate, and the freezing of their rivers ought to be
deemed a prodigy: But they never saw water in MUSCOVY during the winter;
and therefore they cannot reasonably be positive what would there be the conse-
quence. (1964: 94fn)
Now we might observe that it is not the freezing of the water in Sumatra that
is in question. But this is Hume’s point: if it were, then they would be asked to
believe something miraculous. But instead they are asked to believe something
about which they can at most reasonably offer scepticism – they cannot assent
to it, but nor can they discount it. Thus, it is possible for ‘pretty strong testi-
mony’ to bring about belief.
Interestingly, this note from Hume (a late emendation of the text) turns the
‘uniform experience’ completely around and recognizes the value of general
testimony from a slightly different perspective. Extrapolating from what is
said about the Indian prince, it is not my experience which is being generali-
zed. Rather, I am being asked to consider that my experience may not conform
to what is uniform, as long as what is involved does not explicitly contradict
my experience. This reading of the role of ‘uniform experience’ conforms
nicely to the insights that derived from shifting from the egocentric to the
communal perspective above.
This critical scepticism is a necessary tool of the competent layperson. It
opens one to general testimonial reports and the more specific ones from
experts. Here, one’s autonomy is not threatened, but rather strengthened. It
involves the development of a judgment both of character and with respect to
the plausibility of events, two features not brought into the conditions for
assessing the ‘appeal to authority’.
In Robert Audi’s (1997) recent treatment of testimony, he suggests basic
conditions very similar to these. Testimony is a source of basic beliefs (409),
but not itself a generator of knowledge. For that transition to occur, other
things are required: ‘[A] belief that p based on testimony thereby constitutes
knowledge (i.e. counts as testimonially based knowledge) provided that the
attester knows that p and the believer has no reason to doubt either that p or
the attester’s credibility concerning it’ (412). These provisions effectively
capture what is involved in the hearer’s appraisal of expert statements, since
there is a difference between accepting something on the basis of an expert’s
say-so and stating that one thereby knows what it is that has been attested to.
It also indicates the complex context surrounding the ‘appeal to authority’
argument. Required here are further premises over and above those evaluating
the expert as expert (does she have credentials in a relevant field? etc.) and
what is said by the expert (is it agreed upon by other experts? etc.). We must
ask also whether the expert is credible: should she be believed? Is this a context
in which she would be expected to be serious? And we must ask whether what
is said is plausible, given what one knows about the world and what others, in
general non-expert reports, tell one about it.
The case of the prince/king and the frozen water also suggests another
matter about general testimony and expert testimony. Living as he does in a
warm climate, what is the consequence for him of believing or not believing?
Insofar as the Humean king can be persuaded, his general beliefs about nature
and his experience of it will change in a way not experienced by Locke’s
unconvinced prince. But this is a very general change that may not be evident
in any particular decision. After all, what can he do differently in light of this
information? In other cases, and particularly those involving experts, there
may be some immediate consequence for believing or not. If I choose not to
believe that rock specimen ALH84001 originated on Mars, no consequence of
any significance results.1 But if I choose not to believe my doctor’s advice that
I alter my diet, there is likely to be a clear consequence arising from my non-
belief. So the consequences of acceptance and non-acceptance must be entered
as a condition for deciding how much weight should be given to an expert’s
testimony. This, a competent layperson must do. But this, of course, is to draw
attention to the fact that we are talking here not about the competent layper-
son per se but about the tools (conditions) that person must employ. What we
want finally to focus upon is the person who stands behind these conditions,
the one using them.
The Competent Layperson
Douglas Walton’s dialectical treatment of the ‘appeal to authority’ argument

results in six critical questions that should be asked in the assessment of an
expert, along with a set of sub-questions for each. As the most extensive
treatment of the argument form so far, it is a good place to turn in pursuit of
the competent layperson.
As noted earlier, Walton’s interest is in the questions themselves, set, as it
were, between the evaluator and the expert.2 We need to step back and inquire
into the sort of person who is implicated in these questions. That is, what
kinds of competencies are assumed by the questions? I will discuss this with
respect to each major question and its subset, then collect the relevant ideas
and reflect on them in light of what has gone before.
The first question raises the matter of expertise: ‘Expertise question: How
credible is E as an expert source?’ (1997: 223). The assumption here is that the
1 Of course, the choice to not believe is not an arbitrary one; I still need good reasons for dis-
missing the claims of experts. In fact, just talking about choosing not to believe sounds strange,
since beliefs are not things that I can adopt or reject at will.
2 Whether the expert is involved directly or is appealed to by a third party is not a concern at this
point.
plausibility of the expert’s claim is directly related to the degree of credibility

conceded to that person as an expert (rather than as a trustworthy character,
for example). To this end five sub-questions are employed:
1. What is E’s name, job, or official capacity, location and employer?
2. What degrees, professional qualifications or certification by licensing
agencies does E hold?
3. Can testimony of peer experts in the same field be given to support
E’s competence?
4. What is E’s record of experience, or other indications of practiced
skill in S?
5. What is E’s record of peer-reviewed publications or contributions to
knowledge in S? (223).
Obviously these are highly demanding questions and Walton is not suggesting
that they must, or could, all be answered in every case. But they do cover the
kind of information required to satisfy the first critical question. Our question
in turn, is what is assumed here about the person who can ask such sub-que-
stions competently? Ironically, the layperson must here be able to judge
competence. To this end while they do not themselves have to be experts in
the field in question, they need a more general ability to research and evaluate
background. They need to know where to go to find records of publication
and the corroborative testimony of peers, and a sufficient level of education to
be able to measure the quality of what they read, particularly with respect to
sub-question #5. What critical question #1 involves are the general qualifica-
tions of the expert. The layperson must understand those enough to accept the
person as an expert. In some contexts, like a courtroom, much of this informa-
tion would be provided as the expert is presented, and any problems with the
credentials (sub-question #3, for example) would likely come to light through
the cross-examination of lawyers. This first critical question doesn’t directly
help the person to decide on the issue at stake; it merely establishes that what
is involved is an appeal to expertise. But it still tells us something about the
competent layperson.
In a similar way to #1, critical question #2 helps to further establish the
kind of argumentation that is at stake: ‘Field question: Is E an expert in the
field that A is in?’ (223). Here, A is a proposition in a certain field, and the
intent is to locate E and A in the same field. The sub-questions for this are as
follows:
1. Is the field of expertise cited in the appeal a genuine area of knowled-
ge, or area of technical skill that supports a claim to knowledge?
2. If E is an expert in a field closely related to the field cited in the appe-

al, how close is the relationship between the expertise in the two fields?
3. Is the issue one where expert knowledge in any field is directly rele-
vant to deciding the issue?
4. Is the field of expertise cited an area where there are changes in tech-
niques or rapid developments in new knowledge, and if so, is the expert
up-to-date in these developments? (224).
Clearly, these questions demand far more of the competent layperson. They
assume an understanding of what counts as a genuine area of knowledge,
which in turn implies a general comprehension of such fields. This is particu-
larly difficult when the very question of whether a field is legitimate is so often
in debate, particularly in law courts or public fora. Coady (1992: 286-88), for
example, draws distinctions between sciences, pseudo-sciences, and nascent
sciences, primarily in terms of their methodologies and agreed results. The last
category (the nascent) allows that a field about which there may still be some
disagreement may yet grow into a mature science. He cites stylometry as a
plausible candidate for such a designation. Walton himself raises similar kinds
of concerns (176-181). But this in turn does help us to a certain degree. What
the competent layperson needs to know about fields is that they require an
accepted and recognized set of methodological principles. They must be
subject in principle to challenge and testability. And the basic understanding
we already have from observing the world and receiving testimony about it
should help us here. What is at stake in a field of knowledge is no different in
kind from these general underlying experiences that we have.
In critical question #3 Walton turns to what is said by the expert: ‘Opinion
question: What did E assert that implies A?’ (223). The sub-questions associa-
ted with this clearly do not imply that a direct dialogue between the expert and
layperson is being engaged, but allow more generally for reports about ex-
perts:
1. Was E quoted in asserting A? Was a reference to the source of the
quote given, and can it be verified that E actually said A?
2. If E did not say A exactly, then what did E assert, and how was A
inferred?
3. If the inference to A was based on more than one premise, could one
premise have come from E and the other from a different expert? If so,
is there evidence of disagreement between what the two experts (separa-
tely) assert?
4. Is what E asserted clear? If not, was the process of interpretation of
what A said by the respondent who used E’s opinion justified? Are
other interpretations plausible? Could important qualifications be left

out? (225)
These sub-questions assume a basic logical competence. Sub-questions #2 and
#3 in particular require an understanding of how inferences work. And in a
related way, #1 requires some sense of the difference between direct and
indirect communication and attribution, and sub-question #4 assumes an
appreciation of language and meaning. Apropos Willard’s concerns about the
ability of argumentation and Informal Logic courses to provide the skills
necessary for competence in the face of experts, these are all basic skills that
are fundamental to the early stages of any such course.
The fourth critical question focuses on a feature that was recognized in the
earlier discussion: the character of the expert. ‘Trustworthiness question: Is E
personally reliable as a source?’ (223). Far from being an irrelevant considerati-
on when it is the matter of expertise that is being considered, we’ve seen the
personal credibility of the testifier integral to all testimonial evidence. This is
understood in Walton’s sub-questions:
1. Is E biased?
2. Is E honest?
3. Is E conscientious? (217; 227)
These questions, as Walton notes, (227), also recognize the close relationship
between the ‘appeal to authority’ argument and ad hominem argumentation,
here construed as the critical assessment of a person’s character as it relates to
what is said.1 This assumes in the competent layperson a general moral acumen
such that they can evaluate the qualities of character and in particular honesty.
They need to separate legitimate biases (which all reasoners have) from those
that are illegitimate and may cloud the judgment of the expert on a particular
issue. This involves an appreciation of what is relevant to the case – another
basic argumentation skill. The third question, according to Walton, ‘pertains
to E’s scholarly habits or carefulness and professionalism as a skilled technici-
an’ (217). Presumably, if E has sloppy work habits, then the research or results
may be unreliable. A reasonable assessment of this question would involve
more knowledge of E’s background and work history than most laypersons
would possess. But it is conceivable that such information might come to light,
and it would be relevant to the evaluation without requiring any specific skill
on the part of the layperson.
1 See my (1999: Chapter 3) for a discussion of how these argument forms are related under the
general umbrella of ethotic argumentation.
Critical question #5 is one of consistency: ‘Consistency question: Is A

consistent with what other experts assert?’ (223). It is one of the more difficult
challenges involved with the ‘appeal to authority’ argument form. The sub-
questions acknowledge this in that Walton offers two pairs. The first pair is
straightforward:
1. Does A have general acceptance in S?
2. If not, can E explain why not, and give reasons why there is good
evidence for A? (222)
If the answer to #1 is ‘no’, then the second question needs to be asked. Here
the competent layperson must be able to evaluate competing claims. But
insofar as the context of such debates puts the onus on the disputants to
support their position, the requirement reduces to the ability to judge ‘good
reasons’. This is not to sketch over the very real nature of the problem. Consi-
der the recent debate over ‘Cold Fusion’,1 one which is still ongoing. Here the
dispute between the original claimants, Fleischmann and Pons, and their
opponents was played out in a very public forum as the media recognized the
revolutionary nature of the claim if it was established. But the layperson was
left bemused by the researchers’ confession (one that still stands) that while
they had achieved the result of cold fusion in a bottle, they could not explain
the principles involved. Here, Hume’s critical scepticism would counsel wai-
ting for further evidence before accepting the claim.
The aspect of consistency is better captured in Walton’s provision of
further questions when the answer to critical question #5 above is ‘no’:
1. Is it because the question is one on which you would normally expect
experts in this field to disagree?
2. Is it a question where the preponderance of other highly qualified
experts agree on an opinion other than the opinion of the expert source
cited? (221)
The very asking of question #5 assumes the competent layperson can recogni-
ze inconsistency and thus has an understanding of what consistency entails.
This again is a general critical thinking skill. From this perspective the question
set aids in judging the apparent inconsistency and deciding on its importance
in accepting a claim. In the Cold Fusion debate, for example, it was crucial that
other experts could not replicate the results of the principal researchers.
1 The controversy erupted in 1989 when chemists Martin Fleischmann and B. Stanley Pons
announced at the University of Utah that they had succeeded in creating fusion in a bottle. In
November of that year, a panel of scientists reported negatively on the claims. See Eugene F.
Mallove’s ‘The Cold Fusion Mystery Continues’ Tech Talk (4/11/90).
This point takes us to the sixth and last of the critical questions for
assessing experts: ‘Backup evidence question: Is A’s assertion based on
evidence?’ (223). The failure to replicate an experiment is a matter of external
evidence and on this ground we return to the historic tension discussed earlier.
The sub-questions for this are as follows:
1. What is the internal evidence the expert used herself to arrive at this
opinion as her conclusion?
2. If there is external evidence, e.g., physical evidence reported indepen-
dently of the expert, can the expert deal with this adequately? (218).
3. Can it be shown that the opinion given is not one that is scientifically
unverifiable?
It might be objected here that if the critical layperson knew the answers to
these question, he or she would not need to depend on the expert’s say-so,
they would be on equal footing with the expert. Thus, these questions appear
to echo the criticisms of Fry and Sokal and Bricmont. But to be useful (and
our earlier discussion showed that such questions must be involved) we can
interpret the questions as ascertaining generally that such evidence exists
without requiring that the questioner know the evidence. In this sense what is
assumed is general epistemic competence: the ability to recognize what counts
as evidence and an appreciation of the principle involved in verifiability. (In
fact, the point behind question #3 is really subsumed by the background
required to ask the earlier question.) Again, sub-question #2 is a concession to
the power of external evidence, based on general testimony, such that the
expert must be able to account for it. In questions of conflict, the implication
is, the external evidence will be deemed superior.
Conclusion: Competence and Argumentation
Behind much of the discussion about experts and their relation to us we find
an assumption that specialized knowledge somehow distances such people
from us, and, insofar as we might be judged experts in certain fields, it
distances us from others. It is as if we were being looked down upon from
above and viewed each in isolation. But, as we have seen, the extent to which
this is a legitimate picture can be challenged by the lessons learned from
testimony and the pervasive way it draws us together and provides the fund
out of which we emerge as individuals. This unifying base is implicated not
only in the historical treatments of the ‘appeal to authority’ argument, but also
in the more developed contemporary accounts of argumentation theorists like
Walton. Several of his critical questions assume a competence on the part of
the inquirer that relates back to how we function in the world, how we emerge
and grow as epistemic beings.
It is because of the fundamental role that testimony plays in our lives that
the competent layperson is open to other people’s reports. From this derives a
critical scepticism toward the idea that our own experience is central and
representative. As such it cannot be assumed to conform with general expe-
rience. Rather, such conformity has to be established, especially with respect
to astounding and complex reports. This predisposition to believe the accounts
of others is a prerequisite for a reasonable attitude toward the testimony of
experts. Our self-governance, or autonomy, is developed against this back-
drop, conditioned by it, rather than threatened. But in order to evaluate the
testimony of experts, we have seen that more specific traits must be possessed.
Broadly conceived, these traits amount to basic skills in scientific methodo-
logy, reasoning, and ethics. The competent layperson draws on the general
epistemic competence developed in our general functioning in the world in
order to recognize what counts as evidence. Likewise, our general under-
standing of the world allows for the discrimination of fields of expertise as
they admit of methodological principles that allow claims related to them to be
assessed. The evaluation of an expert’s statements involves the logical compe-
tencies of inferencing, identifying relevance, identifying consistency, and a
general appreciation of what count as good reasons. The assessment of charac-
ter requires an understanding of basic moral categories and associated sensibi-
lities. These general and specific traits coalesce in a fluid working definition of
‘competence’.
Charles Willard has done the argumentation community an important
service in pointing out its tendency to idealize its role in society and exaggerate
its accomplishments. But that a problem could be identified in the postures of
some theorists and educators does not mean that answers cannot be found in
the same material. As Willard (1990: 21) himself implies in stressing that he has
offered diagnosis rather than therapy, a way forward often involves a going
back to see if we have been doing what we thought we were doing. It will not
be lost on readers that the model of competence advocated here is very much
rooted in the skills of argumentation and Informal Logic. As has been indica-
ted, a course in argumentation or Informal Logic is exactly the place where
many of the requisite skills are first identified, taught, and developed. In fact,
the ‘appeal to authority’ argument seen in its full complexity suddenly appears
as the perfect vehicle for the recognition and employment of these assorted
skills. It is also the place where we can come to reflect on the full impact of
testimony on our understandings of the world.
We do not need to learn to be experts as much as we need to learn to be

non-experts, to function competently in a world of competing claims from
specialized fields. The education and training we acquire that gives us speciali-
zed knowledge is all too obvious, hence it is easily recognized and promoted
by governments and the media. But the education that gives us the broader
kinds of competence often fails to receive the respect and recognition it de-
serves. It is, on reflection, very much the foundation of what has come to be
called a liberal arts and science education.
References
Audi, R. (1997) ‘The Place of Testimony in the Fabric of Knowledge and Justification’, American
Philosophical Quarterly 34:405-422.
Bentham, J. (1971) Handbook of Political Fallacies [1861]. New York: Thomas Crowell.
Bitzer, L.F. (1998) ‘The “Indian Prince” in Miracle Arguments of Hume and His Predecessors and
Early Critics’, Philosophy and Rhetoric 31:175-230.
Coady, C.A.J. (1992) Testimony: A Philosophical Study. Oxford: Clarendon Press.
Feyerabend, P. (1987) Farewell to Reason. London: Verso.
Forster, E.M. (1936) ‘Roger Fry: An Obituary Note’, from Abinger Harvest, London: Edward
Arnold (Publishers) Ltd.:50-53.
Groarke, L., et al.(1997) Good Reasoning Matters!: A Constructive Approach to Critical Thinking.
Toronto: Oxford University Press Canada.
Hume, D. (1964) An Enquiry Concerning Human Understanding, in Philosophical Works. Volume
II, ed. T.H. Green and T.H. Grouse. Scientia Verlag Aalen.
Locke, J. (1959) An Essay Concerning Human Understanding [1690]. ed. A.C. Fraser, New York:
Dover.
Mallove, E.F. ‘The Cold Fusion Mystery Continues’ Tech Talk (4/11/90).
Manguel, A. (1996) A History of Reading. New York: Alfred A. Knopf.
McKay, D.S. et al. ‘Search for Past Life on Mars: Possible Relic Biogenic Activity in Martian
Meteorite ALH84001’, Science Magazine (August 16, 1996).
Mickunas, A. (1987) ‘Perelman on Justice and Political Institutions’, Practical Reasoning in
Human Affairs, eds. J.C. Golden and J.J. Pilotta, Dordrecht: Reidel.
Sagan, C. (1995) The Demon-Haunted World: Science as a Candle in the Dark. New York:
Random House.
Sokal, A. and J. Bricmont (1998) Intellectual Imposters: Postmodern Philosophers’ Abuse of
Science. London: Profile Books.
Tindale, Christopher W. (1999) Acts of Arguing: A Rhetorical Model of Argument. New York:
State University of New York Press.
Walton, D. (1995) A Pragmatic Theory of Fallacy. Tuscaloosa: The University of Alabama Press.
Walton, D. (1997) Appeal to Expert Opinion: Arguments from Authority. University Park,
Pennsylvania: Penn State Press.
Walton, D. (1998) The New Dialectic: Conversational Contexts of Argument. Toronto: University
of Toronto Press.
Willard, C. A. (1990) ‘Authority’, Informal Logic 12:11-22.
JOHN WOODS
Peirce’s Abductive Enthusiasms
I am part of a team producing a multi-volume Handbook of Practical Reaso-

ning: Computational and Theoretical Aspects, to be published in the fullness of
time by Kluwer. One of our tasks is to get clear about the role of abduction in
practical reasoning. So it is natural to begin with Peirce.
In 1898 Charles Peirce gave a series of eight lectures at Harvard1. Delivered
“off-campus”, these astonishingly fruitful pieces were presented James’ behest,
in yet another gesture of generosity towards his beleaguered friend. Peirce’s
range was striking. He had original and prescient things to say about probabi-
lity, likelihood and randomness; a bout quantification (of which, with Frege,
he was independent co-discoverer); about causality, space and time, and cos-
mology. Of particular note is Peirce’s emphasis on abduction (or retroduction)
as the dominant method in science, in fact, as the only purely scientific method
beyond brute observation. For Peirce, scientific reasoning stands in sharp
contrast to practical reasoning or, as he also says, to reasoning about “matters
of Vital Importance”.2 In this contrast he was hardly alone, although it may be
mentioned in passing that it was a matter on which he disagreed fundamentally
with Mill, for whom “a complete logic of the sciences would also be a com-
plete logic of practical business and common life”.
Peirce saw abduction as hypothetico-deductive reasoning under conditions
of trial. Thus the by now familiar pattern:
If H were true, certain consequences C1,...,Cn would follow. We see that C1,...,Cn
do in fact obtain. So provisionally we suppose that H.
Supposing is not inferring, and neither then is abduction. It is rather a strategy
for making refutable conjectures. Abducted conclusions are not matters for
belief or for probability. In abduction “not only is there no definite probabili-
ty to the conclusion, but no definite probability attaches even to the mode of
inference [sic]. We can only say that ... we should at a given stage of our inqui-
ry try a given hypothesis, and we are to hold to it provisionally as long as the
1 Recently reprinted as Peirce [1992].

2 Peirce [1992], 110.
118 John Woods
facts will permit. There is no probability about it. It is a mere suggestion which
we tentatively adopt”.1
This paper investigates Peirce’s fledgling positive theory. “When [we] say
that a retroductive inference [sic] is not a matter for belief at all,. ..[an] example
is that of the translations of the cuneiform inscriptions which began in mere
guesses, in which their authors could have had no real confidence.”2
[We see, then, that Science] takes an entirely different attitude towards facts from
that which Practice takes.3
Practice involves belief ineradicably, “for belief is the willingness to risk a great
deal upon a proposition. But this belief is no concern of science ... .”4 Belief
contrasts with acceptance, as can be gathered from this passage, which might
well have been penned nearly a century later by Cohen:5
...whether the word truth has two meanings or not, I certainly do think that
holding for true is of two kinds; the one is that practical holding for true which
alone is entitled to the name of Belief, while the other is that acceptance of a
proposition which in the intention of pure science remains always provisional.6
Hence, I hold that what is properly and usually called belief, that is, the adop-
tion of a proposition as [a possession for all time].. .has no place in science at
all. We believe the proposition we are ready to act upon. Full belief is willing-
ness to act upon the proposition in vital crises, opinion is willingness to act
upon it in relatively insignificant affairs. But pure science has nothing at all to
do with action. The proposition it accepts, it merely writes in the list of premi-
ses it proposes to use.7
And since
1 Ibid., 142. Peirce distinguishes (1) induction, or the establishment of frequencies in populations
by sampling and (2) probable inference, or the inference from a known frequency in a randomly
selected sample. He here anticipates Neymann-Pearson statistical sampling theory with its notion
of likelihood. By requiring that inductions have a premiss to the effect that sampling is random,
Peirce thought that all inductions turn on prior discernment of lawlike statements. That a method
of sampling is random requires recognition of the equality of certain frequencies, and so is a kind
of lawlike knowledge, that is, knowledge of generals. Of course, Peirce didn’t get randomness
right. No one did or could until the development of recursion theory well into the present
century. Putnam is good here. See his “Comments on the Lectures” in Peirce [1992], 61 and 68.
See also Lecture Two, 123-142.
2 Peirce, op. cit., 176.
3 Ibid, 177.
4 Ibid.
5 Cohen [1992], 4: “...to accept that p is to have or adopt a policy of deeming, positing, or
postulating that p – i.e. of including that proposition or rule among one’s premisses for deciding
what to do or think in a particular context, whether or not one feels it to be true that p.”
7 Ibid., 112.
[n]othing is vital for science [and] nothing can be..., [t]here is.. .no proposition at
all in science which answers to the conception of belief.1
Whereupon,
...the two masters, theory and practice, you cannot serve.2
When writing about logic, Peirce anticipates Carnap and Brouwer (and
Kripke and Harman, too). And, of course, Quine.
My proposition is that logic, in the strict sense of the term, has nothing to do
with how you think ... . Logic in the narrower sense is that science which con-
cerns itself primarily with distinguishing reasonings into good and bad reaso-
nings, and with distinguishing probable reasonings into strong and weak reaso-
nings. Secondarily, logic concerns itself with all that it must study in order to
draw those distinctions about reasoning, and with nothing else.3
Concerning these things, “it is plain, that the question of whether any deducti-
ve argument, be it necessary or probable, is sound is simply a question of the
mathematical relation between.. .one hypothesis and.. .another hypothesis.”4
Concerned as it is with the presence or absence of that mathematical relati-
on between propositions.
[i]t is true that propositions must be expressed somehow; and for this reason
formal logic, in order to disentangle itself completely from linguistic, or psychi-
cal, considerations, invents an artificial language of its own, of perfectly regular
formation, and declines to consider any proposition under any other form of
statement than in that artificial language.5 ... As for the business of translating
from ordinary speech into precise forms,...that is a matter of applied logic if you
will....6
But applied logic stands to logic much as fool’s gold stands to gold, for it is
“the logical part of the science of the particular language in which the ex-
1 Idem.
2 Idem.
3 Ibid., 143.
4 Ibid.., 144.
5 A Carnapian emphasis.
6 Ibid., 144-145.
120 John Woods
pressions analyzed [or translated] occur.”1 That is to say, it is a part of the

linguistics of natural language.
Logic, then, has nothing to do with how we think. Still less has it to do
with how we think about vital affairs.
Once you become inflated with [what good deducers, i.e. mathematicians, are up
to,] vital importance seems to be a very low kind of importance indeed.
But such ideas [i.e. the procedures of deductive thinkers] are only suitable to
regulate another life than this. Here we are in this workaday world, little creatu-
res, mere cells in a social organism itself a poor little thing enough, and we must
look to see what little and definite task circumstances have set before our little
strength to do. The peformance of that task will require us to draw upon all our
powers, reason included. And in the doing of it we should chiefly depend not
upon that department of the soul which is most superficial and fallible, – I mean
our reason, – but upon that department that is deep and sure – which is instinct.2
All forms of reasoning, even deduction, require observation. What is ob-
servation? It is
the enforced element in the history of our lives. It is that which we are con-
strained to be conscious of by an occult force residing in an object which we
contemplate. The act of observation is the deliberate yielding of ourselves to that
force majeure, – an early surrender at discretion, due to our forseeing that we
must whatever we do be borne down by that power, at last. Now the surrender
which we make in Retroduction, is a surrender to the Insistence of an Idea. The
hypothesis, as The Frenchman says, c’est plus fort que moi.3
The passage is striking and at seeming odds with Peirce’s insistence that is not
useful in ordinary affairs and that it has nothing to do with belief or probabili-
ty. (After all, the most common fallacy of abduction is Bayesianism, that is, to
“choose the most probable hypothesis”4). But belief is like observation. It
presses in on us and demands our surrender. And what is so insistent about the
Insistence of an Idea if it does not somehow call for and eventuate in belief?
1 Ibid., 145. Peirce, like Mill, is somewhat ambivalent about the primacy of induction. “As for ...
Induction and Retroduction, I have shown that they are nothing but apogogical transformations
of deduction and by that method the question of the value of any such reasoning is at once
reduced to the question of the accuracy of Deduction” (Ibid.). On the other hand, the “marvelous
self-correcting property of Reason, which Hegel made so much of, belongs to every sort of
science, although it appears as essential, intrinsic, and inevitable only in the highest type of
reasoning which is induction” (Idem., 168). Of Peirce’s transformations we have the space to say
only that he regarded induction, probabilistic inference and retroduction as distinct forms of
reasoning of which Aristotle’s first three syllogistic figures were limiting cases. See Lecture Two,
131-141.
2 Ibid., 121.
3 Ibid., 170.
4 Putnam [1992], 78.
We begin to see the irreducibly practical aspect of abduction. It shows itself

in two ways, in the Insistence of an Idea and in the inductions that let loose
our sentiment, configure our habit and culminate in belief. This last aspect is a
scientific, an all too human acquiescence in a proposition by which our vital
affairs might be guided. But the first is purely scientific, an intrinsic part of
abduction itself, and a seeming contradiction of Peirce’s settled and repeated
views. Let us see.
Abduction is two things,
(a) surrender to an Idea, to a force majeure; and
(b) a method of testing its consequences.
In surrendering to an idea, say the existence of quarks, or the promise of
quantum electrodynamics, we do not come to a belief in quarks or to a belief
that QED is true, but rather to a belief that quarks are a good thing to suppose
and a good bet for testing, or that QED has a good explanatory record or at
least promise of such. In this, abduction resembles practical reasoning (and
observation, too). At the heart of this resemblance is the concept of a “vital
affair”. Here, too, there are two conceptions to ponder.
[]A vital affair is made so by its subject matter, i.e., its vitality is topic-
intrinsic (e.g., ethics possibly).
[]A vital affair is made so by its methods of enquiry irrespective of its
subject matter.
It seems to be Peirce’s view that the vitality of an affair is entirely a matter of
the manner in which we conduct its enquiry. There may be matters that espe-
cially conduce to such methods, but there seems to be nothing intrinsic to the
methods that preclude them from application to any subject matter.
In the case of abduction, these methods are applied to the important busi-
ness of making conjectures for subsequent test. Essential to this is the convicti-
on that a hypothesis H is a good enough conjecture to warrant our making it.
That is a vital affair for scientist and laymen alike. What this shows is that
belief is indispensable for abduction, but it does not show an inconsistency in
Peirce’s account. For here the object of belief is not hypothesis H, but rather
is some proposition of which H is the subject, viz., the proposition that H is a
good conjectural bet (or, in some variations, that H is a good explanation of
some set of target phenomena). Equally, a test of H is no good if the experi-
menter is in any doubt – to say nothing of a state of indifference – about
whether these things are consequences of H or about whether this other thing
implies the negation of a consequence of H.
122 John Woods
We have been suggesting that abduction has been revealed to have an

irreducibly practical component. if that were true, it would matter for of a
theory of abduction. And it would matter greatly if there could be such a thing
as a theory of practical reasoning (an affirmative answer to which being pre-
supposed by the title and subtitle of my Handbook). We note here for subse-
quent consideration the possibility that while abduction has been shown to
have a practical aspect, it has not been shown to be irreducible in the requisite
sense of being an aspect that any competent theory of abduction must take
note of and accommodate within its principled disclosures.
For the present we merely mark Peirce’s ambivalence about the possibility
of a science or a theory of practical reasoning. Certainly there is for him little
prospect of there being a logic of practical reasoning. Standing in the way is
logic’s necessary silence on how to represent a piece of practical reasoning in
the artificial language in which the logic transacts it proper business.
“In everyday business”, says Peirce, “reasoning is tolerably successful; but
I am inclined to think that it is done as well without the aid of a theory as with
it”.1 This suggests not the impossibility of a theory so much as its unhelpful-
ness, and it leaves open an option of a type that he has already acknowledged
in connection with deductive logic. Given that deductive logic doesn’t fix the
representation relation that takes natural language arguments into logic’s
artificial language, this is a task that is left for the science of the natural langua-
ge in question. By these lights, there is no logic of abduction, but it might well
have a scientific theory for all that. if so, the science of abduction will itself be
abductive – as all genuine science is. And therewith is met an interesting
problem. It is the problem of specifying what it is about our abductive practice
that calls out for abductive theoretical analysis and disposition. If we agree that
abduction is au fond the making of conjectures in the light of their consequen-
ces and the testing of those consequences in the light of new information, then
what are the abductive questions to put about these practices?
We are not here attempting to ready the reader for an assault upon the
Problem of Abduction, as we might say; but we might briefly pause over the
point before moving on. If we ask by what might abduction be justified, there
is little chance that we will like the suggestion that its legitimacy is the con-
clusion of a deductive proof or the conclusion of a forceful abductive argu-
ment. The first doesn’t exist and the second commits the fallacy of petitio
principii. Of course,
[a]fter a while, as Science [i.e., abduction] progresses, it comes upon more solid
ground. It is now entitled to reflect, this ground has held a long time without

showing signs of yielding. I may hope that it will continue to hold for a great
while longer... For a large sample has now been drawn from the entire collection
of occasions in which the theory comes into comparison with fact, and an overw-
helming proportion, in fact all the cases that have presented themselves, have
been found to bear out the theory. And so ... I can safely presume that so it will
be with the great bulk of the cases in which I shall go upon the theory, especially
as they will closely resemble those which have been well tried.1
This of course is induction, and induction has no place in science. If, on the
other hand, the justification is not scientific, it is practical. This is not sufficient
ground to disentitle induction, but it won’t work as a justification of abducti-
on. The reason is that induction itself is abductively justified. For what justi-
fies our inductive practices if not the conjecture that if nature were uniform
our practices would be successful, together with the observation that our
practices have been successful up to now? This leaves the justification of
abduction with a circularity problem and then some. (Perhaps we could call it
abduction’s double helix.)
With that aside, we return to our present question. Regarding our abducti-
ve practices, is there a subject matter for a scientific theory? If there were, there
would be something for that theory to explain by conjecture and refutation.
But what is this explanandum? Evidently that science proceeds by looking for
an explanation which it then tests. Then what is the explanation, the explanans,
of this fact? That this is how science is best done, presumably. But now: how
is this conjecture tested? If the conjecture were that this is how science is done,
it would be no conjecture, but a matter of observation; and without conjectu-
re, the theory could not be abductive. It matters that the conjecture at hand is
that this is how science is best done, that this is the right way to do science.
Now Hume hovers nearby, and chuckles. We find ourselves stretched out on
the prickly barbs of the ought-is divide.
What this suggests is that we are ill-served by any generalized abductive
enthusiasm for abduction as such. A better – or anyhow more hopeful –
option is to abandon our global ambitions in favor of a more provincial ap-
proach to abduction. What is needed are theories of the following things, one
by one.2
First is required a scientific account of hypothesis formation, itself a sub-
theory of belief fixation. Peirce insists that no logic is fitted for such a task, and
in this we may suppose him to be right. It may seem that in as much as belief
fixation involves, essentially, the holding of beliefs, and various other probabi-
listic raptures, no science is available for such phenomena. But this is quite
1 Ibid.., 177.
2 See Lipton [1991].
124 John Woods
wrong. It confuses subject matter with method. The belief-theorist is meant to

have or to hold no beliefs about belief, except, at the limit, his own beliefs
about what are the fruitful conjectures to make about belief. We may liken this
science to a branch of cognitive psychology, and nothing we have said here
prevents the theorist from proceeding as abductively as he pleases, as far as
abduction can in general go. All that is required is that the theory give an
explanatory account under conditions of trial of how abducers form the
hypotheses they provisionally adopt under conditions of trial.
Second comes a theory of consequence. It is necessary for the theorist to
have an account of that relation in virtue of which conjectured hypotheses are
testable. Nothing we have said here precludes our calling deductive logic into
service.
Thirdly, we must have a theory of refutation. Our friend modus tollendo
tollens will apply for the job, no doubt. He is not up to it. He will give hope-
lessly indeterminate guidance except for the simplest kinds of case. The best
general theory of refutation produced so far is Aristotle’s. It is a beautifully
crafted exploitation of properties of the syllogism and of conditions that bind
certain kinds of rational, even if adversarial, discourse. Two-and-a-half mille-
nia have come and gone since its launching. No doubt it could do with some
tarting up; but it is unlikely that its central core will require, or tolerate, much
change.
So conceived of, a theory of abductive reasoning is a triple AL=<an abduc-
tive theory of conjecture, a deductive theory of consequence, a logico-dialecti-
cal theory of refutation>. Of these three, only the theory of consequence lays
any claim to consensual maturity, while the third awaits contemporary revival.
This leaves the first, the theory of conjecture or of belief-fixation with respect
to, as we may now say, plausible conjecture. On this the polls are open, and
early results are only trickling in. Its various subcomponents show promise of
fruitful union, what with recent initiatives in cognitive psychology and M. We
would also conjecture a useful role for something like a logic (to speak loosely)
of plausibility as a minimal constraint, and a full-bore account of analogy,
perhaps along the lines of Woods and Hudak [1989] or along the rather diffe-
rent lines of Holyoak and Thagard [1994].
Pages back, we said that abduction was inherently practical in certain of its
aspects. The abducer is always driven by a belief that “this proposition is
worth entertaining”, a belief held by the abducer independently of whether he
or anyone else actually believes the proposition in question. Given that the
abducer’s belief is always a function of the Insistence of an Idea, abduction is
always practical and, as such, ascientific. if this is right, the same will hold for
any theory of abduction itself which aspires to be abductive. It may seem to

follow that there can be no theory of abduction, but we may think that this
goes too far. The fallacy of composition looms. It is here the fallacy of suppo-
sing that if a theory has an ascientific component the theory as a whole is
ascientific. That thinking so would be a fallacy in the present case is suggested
by the fact that the abducer’s sole belief (hence his sole practical contribution)
is his belief about what is a plausible candidate for testing. Of course, it may
turn out that the candidate tests negatively, but this wouldn’t show that it
wasn’t worth testing. And so we might find ourselves ready to concede that
such beliefs are ascientific in the degree that they are untestable. But ascientific
or not, such beliefs don’t discredit the scientific enterprise. We will be led by
such beliefs to no scientific theories of a sort not already admissible under the
testing provisions of abduction generally. If these let in bad theories, it won’t
have been for the ascienticity of the beliefs that set them up as conjectures in
the first place. So it cannot be supposed that abduction entire is ascientific just
on account of the ascientificity of that aspect.
Suppose, then, that we like this specification of our triple AL enough to be
getting on with. Getting on with it in dominantly the business of elucidating
the mechanisms of what we might call the theoretical imagination. If tellable at
all, it would tell the story of how in Quine’s words theories are free for the
thinking up, that they are the exercise of our conceptual sovereignty, no more
and no less.
Bibliography
L. Jonathan Cohen. 1992. An Essay on Belief and Acceptance, Oxford: Clarendon Press.
Keith J. Holyoak and Paul Thagard. 1994. Mental Leaps, Cambridge, MA: MIT Press.
Peter Lipton. 1991. Inference to the Best Explanation, London: Routledge 1991.
C.S. Peirce. 1992. Reasoning and the Logic of Things. Ed. Kenneth Laine Ketner. Cambridge,
MA: Harvard University Press.
Hilary Putnam. 1992. “Comments on the Lectures”, in Peirce [1992], 55-102.
John Woods and Brent Hudak. 1989. “By Parity of Reasoning”, Informal Logic XI, 125-140.
126 Henry W. Johnstone
HENRY W. JOHNSTONE, JR.
“‘Any,’ ‘Every,’ and the Philosophical Argumentum ad Hominem”
Aristotle obviously has in mind a geometrical proof a later version of which is

included in Euclid’s Elements . This proof concerns an arbitrary triangle – any
triangle – and shows that on the basis of postulates, axioms, and previously
proven theorems, such an arbitrary triangle has an angle-sum of two right
angles. Hence this property applies to every triangle.
That every triangle has a certain property is thus demonstrated by reaso-
ning about any triangle. “Every,” we see, belongs to the statement of a theo-
rem; “Any” belongs to a step in the proof of the theorem, and provides the
basis for the use of Universal Generalization.
The same machinery operates in the philosophical argumentum ad homi-
nem . If some arbitrary person any person holding the view in question
can be attacked ad hominem for holding this view, then everyone holding this
view can be similarly attacked, and the view itself comes into question.
-° C#/ Á%9'3J <-> -C-J, ²- %¤ -#º -/3C-#+ ¤
%'Ë-#/ J½- ... µ -#?/ -° -/3° %'Æ-# J?/- E#
±'qO 3# ´-#º x#, -#E-# %'Ë-Ç Á%w'3J C#/. (Aris-
totle, Posterior Analytics 73b3374a1)
Tredennick translates this passage: “An attribute only belongs to a subject
universally when it can be shown to belong to any chance (%¤ -#º
-/3C-#+) instance of that subject, and to belong to that subject primarily.1
. .. Thus that which can be shown in any chance (-/3C) instance to fulfill the
condition of containing the sum of two right angles, or any other requirement,
is the subject to which that universal attribute primarily belongs” (pp. 47-49).
But if we render -/3C and the prepositional phrase %¨ -#º -/3C-#+ as
“arbitrary” rather than “chance” or “by chance,” it is clear (as Barnes saw in
1994, when he used “arbitrary” instead of “chance” or “by chance” in his
translation of this passage on p. 8 as well as in his commentary on it on p. 119)
that Aristotle anticipates the Rule of Universal Generalization, according to
which what is true of an arbitrary individual is true of every individual. For
example, if the sum of the angles of an arbitrary triangle can be shown to equal
two right angles, the sum of the angles of every triangle must be equal to two
1 In this article , I will not be concerned with how “primarily” (%'Ë-#/, %'Ë-Ç) functions in
this passage, either here or in the rest of what I quote.
‘Any,’ ‘Every,’ and the Philosophical Argumentum ad Hominem 127
right angles. This is of course Aristotle’s example.1 (In an earlier translation in

1984, Barnes, too, had rendered -/3C and %¤ -#º -/3C-#+ as “chance.”
This rendition apparently prevented him from seeing that at this point Aristot-
le was anticipating the Rule of Universal Generalization. I don’t think the
question whether Tredennick’s identical rendition kept him from seeing Aris-
totle’s statement as anticipating this Rule arises. The Rule anticipated here had
not yet been generally recognized at the time Tredennick’s translation appea-
red.)2
The adjective “chance,” used by Tredennick in his renderings of -/3C
and %¤ -#º -/3C-#+ is a near-synonym of “random.” I am thus emphasi-
zing the difference, in the present context, between “random” and “arbitrary.”
One way of making this difference clear is to consider what would be involved
in showing that the sum of the angles of any random triangle equals two right
angles. This is a statistical problem, involving the use of a protractor or similar
instrument to measure the angles of a large number of triangles and show that
their sum turns out in all these cases to be that of two right angles. This is an
inductive argument, while the place of the Rule of Universal Generalization is
in a deductive context.
There are cases in which while “random” is appropriate, “arbitrary” is not.
“Random” is an adjective often applied to samples. Of a random sample of
say, wheat drawn from the hold of a ship (to borrow an example used by
Peirce in “The Doctrine of Necessity Examined”; see 1982, 6.40), it cannot be
said that nothing matters about it except that it is a sample of wheat. What
matters is the quality of the wheat in this particular sample. But if we ever had
occasion to speak of “an arbitrary sample of wheat,” all that would matter
the only premise from which we could reason would be that it was a sample
of wheat.
It is easy to understand the temptation to translate the adjectival form
-/3C and the prepositional phrase using its genitive -/3C-#+ as “chance”
or “random.” The adjectival form -/3C is the aorist participle of the verb
-/97, one important new meaning of which is “to happen by chance.”
1 Aristotle is here talking about -'?7C - -'?7#.,, “a triangle quâ triangle.” Perhaps
such an interpretation of this passage in the Posterior Analytics sheds some light on the meaning of
the celebrated phrase -°² ², ‘Being quâ being,” in , e.g.., Metaphysics 1003a21.
2 Tredennick’s translation appeared in 1924. The paper by Gentzen in which the Rule is for the
first time formalized. appeared in 1934. It is unlikely that Tredennick would have seen any need
to translate -/3C in an unusual way.
When Aristotle uses this participle in the sense of “arbitrary,” this may well be
the first time the word had been used in this sense.1
What Aristotle must have in mind in speaking of an arbitrary figure is the
part such a figure plays in a proof on the order of the one formalized later by
Euclid (I.e., Bk. I, Dem. 32) to show that the angles of every triangle are equal
to two right angles. This theorem is proved in terms of an arbitrary triangle
showing on the basis of axioms, postulates, definitions, and previously pro-
ved theorems that the arbitrary triangle under consideration (that is, any
triangle) does indeed necessarily have an angle-sum of two right angles.
Euclid’s proof begins “Let there be a triangle ABC.” (`)-7 -'?7#
-°
) It then proceeds to its conclusionthat the sum of the angles of
ABC is two right angleswithout making use of any property of ABC except
that it is a triangle. For ABC is an arbitrary triangle. Instead of the )-7
construction, Euclid might have spoken of a -/3° -'?7#. Euclid does
in fact often use -/3C to mean “arbitrary” (See, for example, Dem. 5, Line
11, where Euclid uses the expression -/3° )J¥#, “an arbitrary point.”
He seems, in fact, throughout the Elements to reserve the participle -/3C for
points. But there are, as we have seen, alternative ways in Greek of expressing
the idea of “arbitrary”; for example, using )-7 to introduce an arbitrary
entity like a triangle.)
Aristotle and Euclid each expresses in his own way the proposition that if
any arbitrary triangle has an angle-sum of two right angles, then every triangle
must have the same property. Inferences of this sort, falling under the rubric of
“Universal Generalization,” may seem to demand that we follow Aristotle and
Euclid in designating some term involved in the proof as “arbitrary.” But it
should be noted that the systems of logic introduced by Gentzen under a
heading that has acquired the name of “Natural Deduction” are formalized in
such a way that the characterization of a variable as “arbitrary” is inappropria-
te and in any event unnecessary, since the function of such a characterization
is performed by formal restrictions on the introduction of the variable to
which the universal quantifier may be applied. Some examples of such formali-
zation of Universal Generalization are to be found in Price (p. 95), LeBlanc
and Wisdom (p. 205) and Anderson and Johnstone (p. 170) as well as in Gent-
zen himself (p. 186). But formalizations of this sort, in my view, have no
tendency to show that there never was a need for the concept of the arbitrary,
or that there is not still a need for it. Euclid did not have to wait for Gentzen
to systematize geometry. And at the end of this article I want to exhibit the
1 I assume, as most do, that the Posterior Analytics is earlier than the De Caelo, where Stocks
translates-/3C as “arbitrary” in, for example, “[an] arbitrary fragment of earth.” (297b8)
function of the unformalized concept of the arbitrary in the analysis of certain

kinds of philosophical arguments and conclusions.
The examples I have given of the uses of the concept of the arbitrary in
Aristotle and Euclid are applications of the general principle that while “any,”
when it appears, does so in a step of a demonstration, “every” introduces the
statement of the theorem demonstrated. What I am presenting here is a sort of
ideal lexicography: in an ideal version of the English language, “every” would
introduce a theorem and “any” (provided it function as a universal quantifier1)
a step in its proof. An infraction of this ideal rule is perhaps a little more
painful when “every” is used to express “arbitrary” than when “any” in-
troduces a theorem. Thus when Heath’s translation of the theorem demon-
strated by Euclid’s Demonstration 32 begins “In any triangle ...,” (p. 316) I,
perhaps with linguistic hypersensitivity, but at any rate in conformity with
Euclid’s Greek, would have expected “In every triangle ...” (%-°O
-'I#/., as Euclid in fact has it.) .
I do not intend the scope of these remarks to be restricted to theorems and
proofs in Euclid. I don’t know what would prevent them from applying to any
mathematical theorem, provided only a generalization were being proved and
the proof hinged on the introduction of arbitrary instances. (But see below for
reference to a theorem in which this is not the case.)
(I suspect that the distinction I am trying to draw here is not unique to the
English language. Isn’t it the case that while “tout” normally introduces a
theorem, “n’importe quel”2 is more suitable in a step of a theorem? And that
the same relation holds between “jeder” and “irgendeiner”? It should be
recalled, however, that our English “any” in negative, conditional, and interro-
gative contexts behaves semantically in an idiosyncratic way.3)
In any event, leaving such idiosyncrasies aside, I am here distinguishing
“every” and “any” only within the context of proofs making uses of the Rule
of Universal Generalization, which can be expressed semi-formally as “If 1a,
where a is an arbitrary individual, then for every x , 1x .” I have not yet said
anything about Universal Instantiation, in a Gentzenian formulation a logical
1 In this phraseology I am attempting to take account of the possibility that “any” might express
an existential quantifier, as in “if any x is P, then every x is /g” – a example suggested to me by
Hans Hansen. (Regarding the phrase “if any”, see Footnote 3, below.
2 This phraseliterally “It doesn’t matter which”seems to expresses more vividly than “any” the
arbitrariness of what it governs.
3 Consider “Not any S is P.” If there is any S, it is P,” and “Is anything S?” In none of these
examples could “Every” be substituted for “Any” without radically changing the meaning. I know
of no other language in which such idiosyncrasies are exhibited. But this study is not linguistic,
and I don’t feel obliged to pursue the matter.
movement in a direction opposite to that of Universal Generalization. Accor-

ding to this rule, from “For every x , 1x ,” we can deduce “1a .” Here the
individual a can, but need not, be arbitrary. It can, for example, be either an
arbitrary triangle or a specific one. But if we wish now to continue our train of
reasoning by applying Universal Generalization to premises making use of
what we have inferred through Universal Instantiation, we can deal only with
an arbitrary individual say, an arbitrary triangle rather than a specific triangle
the properties or dimensions of which, over and above its being a triangle are
to be taken account of in our reasoning.
*****
In Johnstone 1996a (p. 95), I deal with the following problem: In an argumen-
tum ad hominem in philosophy, as well as many non-philosophical contexts,
the attempt is made through the use of this argument to show that on his or
her own principles, a person ought to reject a thesis he or she has asserted. If
this argument is valid, the person attacked is under an obligation to withdraw
or modify the thesis in question. But to attack a person in such a manner is not
yet by any means to establish a general philosophical position.
In my view, however, a valid philosophical ad hominem does establish a
general position, provided in attacking the holder of the position we have
taken account of nothing about the holder except that he/she takes the position
in question. In other words, we treat the holder of the position as an arbitrary
individual.1 And we apply Universal Generalization.
In the so-called “third man” (but more appropriately called “the third
greatness”) argument, an ad hominem addressed against Socrates by Parmeni-
des in that eponymous Platonic dialogue, (in 132ff), Socrates is shown that his
theory of forms leads to an infinite regress, and so is untenable. The argument
is ad hominem because each step of the regress is powered by Socrates’ own
assent to consequences invoked one after another by Parmenides. The form of
greatness must share some further form with the great objects to which it is
common. (I have written about this effect in Johnstone 1996b, and elsewhere.)
This argument appeals to no feature of Socrates except his espousal of a
certain theory of forms. Socrates is, then, an arbitrary holder of that theory.
By Universal Generalization it follows that it is not merely Socrates’ mainten-
ance of the theory in question that is under attack; it is the maintenance of the
1 To treat the person attacked as an arbitrary individual would completely undercut the very point
of an argumentum ad hominem in the abusive category, and would not often be appropriate for
one considered circumstantial .
theory by everyone who now maintains it or might be tempted to maintain it

in the future. It is the theory itself that is untenable.
A philosophical argumentum ad hominem has been used to reach, via a
Universal Generalization, a conclusion independent of the homines who hold
it. This conclusion has actually required two steps: from any thinker who
might hold the theory in question to every such thinker; and then from such
universal concurrence to the falsehood of the theory itself.
Aristotle, in his argument1 more appropriately than Plato’s called “the third
man” because it invokes, in addition to the illustrative men Socrates and Plato
(who are incidentally treated in this example as arbitrary men, since none of
their specific properties is at issue here except that of being men) the form in
which they both participate. Aristotle then sets this form up beside the two
men, considering that there must now be a further form common to all three.
He carries the inquiry still one step beyond the conclusion reached in the
Parmenides , basing on the general rejection of the theory of forms his own
positive view of substance and attribute.
To summarize: an argumentum ad hominem valid against any person who
is arbitrary in the sense that the only property of this person of which we are
taking account in arguing is his/her commitment to the point of view under
attack is valid against every such person, and hence refutes the point of view in
question itself. From this point we may proceed, as Aristotle did, to establish
positive conclusions in philosophy conclusions having their origin in argu-
menta ad hominem .
It might be objected that when a position is universally refutable ad homi-
nem, the situation can be characterized without recourse to the idea of an
argumentum ad hominem by simply stating that the position itself is refutable
in its own terms the homo need not be called on to participate. For example,
the version of the theory of Forms that gives rise to the third man simply
generates an infinite regress, and must be rejected for that reason regardless
whether anyone actually holds this theory and can, like Socrates confronted
by Parmenides, be attacked ad hominem for holding it. But in Johnstone 1989
I argue at length that all valid philosophical arguments are ad hominem. That
this consideration especially applies to infinite regress arguments I try to show
in Johnstone 1996b. But it is by no means restricted to them.
The underlying reason why I think that all valid philosophical arguments
are ad hominem is that none of them, in my opinion, is ad rem . For there is in
philosophy no generally accepted res no set of facts beyond controversy to
1 Summarized by Alexander of Aphrodisias in Commentarius in Metaphysica. 83.33-84.7 and

84.22-85.3
which one could properly i.e., validly appeal as a basis for arguing against
another view. Consider, for example, the issue between the Protagorean view
that man is the measure of all things and what is presupposed by the Biblical
question “What is man that thou art mindful of him?”1 To what res what
evidence could anyone appeal as evidence supporting either of these views?
But if the ad rem is unavailable, our only resource is the ad hominem; e. g., to
apply the Protagorean dictum against Protagoras himself. (But of course,
“Protagoras” can be taken as designating any arbitrary adherent to this dictum;
the argument establishes the falsity of the dictum itself.)2
References
Alexander of Aphrodisias. Commentarius in Metaphysica. In Barnes 1984a,
Anderson, John M., and Henry W. Johnstone, Jr. 1962. Natural Deduction, The Logical Basis of
Axiom Systems. Belmont, CA: Wadsworth.
Barnes, Jonathan, ed. 1984a. The Complete Works of Aristotle. The Revised Oxford Translation.
Princeton: Princeton University Press.
——, trans. 1984b. Aristotle’s Posterior Analytics. In Barnes 1984a.
——, trans. 1994. Aristotle, Posterior Analytics. 2d ed. Oxford: Clarendon Press.
Gentzen, Gerhart. 193435. “Untersuchen über das logische Schliessen.” Mathematische Zeit-
schrift 39: 176210.
Heath, Thomas L., ed. and trans. 1926. The Thirteen Books of Euclid’s “Elements.” Cambridge:
Cambridge University Press.
Heiberg, I. L., ed. 1883. Euclidis Elementa. Leipzig: Teubner.
Johnstone, Henry W., Jr. 1989. “Self-Application in Philosophical Argumentation.” Metaphi-
losophy 20.3&4:247261.
——, 1996a. “Locke and Whately on the Argumentum ad Hominem.” Argumentation 10.1: 8997.
——. 1996b. “The Rejection of Infinite Postponement as a Philosophical Argument.” The Journal
of Speculative Philosophy 10.2: 92104.
Leblanc, Hugues, and William A. Wisdom. Deductive Logic .Boston: Bacon and Allyn 1972.
Peirce, Charles S. 1982. Writings of Charles S. Peirce, A Chronological Edition. Ed. Max Fisch.
Bloomington, IN: Indiana University Press.
Price, Robert, 1962. “Arbitrary Individuals and Natural Deduction.” Analysis 22:94-96
Stocks, J. L., trans. 1984.”On the Heavens” (“De Caelo”). In Barnes 1984a.
Tredennick, Hugh, trans. 1926. Aristotle’s Posterior Analytics. Loeb Classical Library. Cambridge,
MA: Harvard University Press.
1 Psalms 8, 4
2 Those who have helped me in many ways all essential, but some more logistical or editorial
than dialectical with the preparation of this brief article include Ray Ayoub, Barbara Bennett,
Donna Black, Carl Hausman, Dale Jacquette, Mari Lee Mifsud, Dawn Osselmann, Robert Price,
Stephen Wheeler, and especially David Engel.
ON CONTEMPORARY PHILOSOPHY
HANS LENK
Interdisziplinarität und Interpretation
Im ersten Teil meines Beitrages möchte ich auf die Verflechtung der verschie-
denen Problembereiche in unserem komplexen und verflochtenen Zeitalter
hinweisen; im zweiten Teil werde ich auf Typen der Interdisziplinarität und
auf die Frage der Abgrenzung oder Abgrenzbarkeit der verschiedenen Diszi-
plinen eingehen. Im dritten Teil werde ich meinen Ansatz skizzieren, den ich
einem methodologischen Interpretationismus oder methodologischen Inter-
pretationskonstruktionismus nenne; dieser scheint mir eine Brücke zwischen
den Disziplinen zu ermöglichen. Zuletzt werde ich noch hinweisen auf be-
stimmte Tendenzen der Geschichte der Textinterpretation bzw. der methodo-
logischen Hermeneutik, die in gewisser Weise im Zusammenhang mit diesem
interpretationistischen Ansatz stehen.
I.
Zunächst beginne ich mit einem allgemeinen Überblick über Problemver-

flechtungen. Es ist weithin bekannt, daß die drängenden Fragen der Zeit sich
heute nicht mehr in sauberer fachlicher Abtrennung oder disziplinärer Ver-
packung stellen, sondern interdisziplinär nur zu bearbeiten sind. So sind etwa
die ökologischen Fragen nicht nur naturwissenschaftliche Probleme, aber
natürlich auch nicht nur geistes- oder sozialwissenschaftliche oder kulturelle
Probleme. Sondern wie bei allen unsere komplexen und übergreifenden auch
natürlich auch die Nationengrenzen übergreifenden Problemzusammenhängen
können wir eine innere Wirkungsverflechtung und eine Vermaschung der
entsprechenden Bereiche feststellen. Wir brauchen mehr und mehr abstraktere
fachübergreifende Methoden, Disziplinen und Erfassungstechniken, sozusagen
generalisierte operationale Techniken, um diese Probleme überhaupt präzisie-
ren und auch behandeln zu können. Das heißt, es gilt in gewisser Weise eine
abstraktere und methodologische Sicht einzunehmen, die ich im zweiten Teil
zu diskutieren versuchen werde. Das gilt übrigens auch für die Geisteswissen-
schaften – und auch für die Germanistik. Es ist deutlich, daß in den letzten
134 Hans Lenk
Jahren die Informationsverarbeitungsverfahren, die elektronische Datenver-

arbeitung und andere Techniken in den Geisteswissenschaften erheblich
zugenommen haben und immer mehr Relevanz gewinnen. Es geht um Infor-
mationen, genauer: um gedeutete Informationen oder Interpretationen bzw.
um Ergebnisse von Interpretationen. Man könnte von “Interpretaten” spre-
chen oder von “Interpretationskonstrukten”, die gleichsam zu einem neuen
höherstufigen “Rohstoff” der Wissenschaften, in diesem Falle der Geisteswis-
senschaften und der entsprechenden angegliederten Informationsverarbei-
tungsdisziplinen, geworden sind.
Zumal die Information und ihre Handhabung sind in den letzten Jahrzehn-
ten dem systematischen technischen Zugriff zugänglich geworden. Man kann
geradezu diagnostizieren, wie ich es bereits 1970 getan habe, daß die Informa-
tionsverarbeitung systematisiert wurde und zu einer Art von “informations-
und systemtechnologischen Zeitalter” Anlaß gegeben hat. Das gilt natürlich
sowohl für die Produktionsautomatisierung und die Einbettung in Systeme
wie für die systemhafte operationale Abläufe generell. Hier kann man von
einem Trend zu einer umfassenden Systemtechnik sprechen. Das gilt aber
natürlich auch für die repräsentationalen, also für die Darstellungen in Infor-
mationsdisziplinen, Informationssystemen; auch hier sprechen wir von Infor-
mationsverbundsystemen, die immer mehr Bedeutung gewinnen. Der Einsatz
des Computers ist hierfür natürlich auf allen Gebieten der wissenschaftlichen
Darstellung, aber auch der Produktions- und Verfahrenssteuerung und -kon-
trolle einschlägig. Alle diese Trends sind Gesichtspunkte einer umfassenden
Informations- und Systemrationalisierung in den hochindustrialisierten Ge-
sellschaften. Ich habe schon 1970 davon gesprochen, daß das technische oder
technisch-wissenschaftliche Zeitalter sich zum “informations- und system-
technologischen Zeitalter” – oder kurz zum “systemtechnologischen Zeitalter”
– gewandelt hat. Die Informations- und Systemtechnologien übergreifen die
Grenzen einzelner Disziplinen. Man kann sozusagen von einer systemtechno-
logischen oder sogar systemtechnogenen Interdisziplinarität sprechen. Die
scharfe Abtrennung in Disziplinen ist eigentlich nicht mehr in diesem Sinne
haltbar – und zwar schon aus operationalen methodologischen und quasi-
methodentechnischen Gründen nicht. Das führt natürlich auch zu entspre-
chenden Herausforderungen einerseits der wissenschaftlichen Methodologen,
andererseits natürlich auch der Sozialphilosophen und Moralphilosophen. Wir
alle kennen die Probleme, die sich mit den Dokumentationssystemen, der
Zugänglichkeiten von Daten, der Kombinierbarkeit von Daten in Hinsicht auf
Datenschutzfragen, Datenschutzgesetze usw. ergeben. Man kann fast befürch-
ten, daß wir so etwas wie eine Computerokratie erleben werden ; die Auf-
fassung, daß dieses das unaufhaltsame Schicksal der industriellen Massengesell-
schaften im systemtechnologischen Zeitalter sei, ist weit verbreitet. Heut-

zutage gilt das natürlich ganz besonders und wird bedeutsam durch neue
weltumspannende Informationssysteme, wie beispielsweise das INTERNET,
World Wide Web und die direkten Zugriffsmöglichkeiten allerorten, die dann
aber auch die Frage der ethischen Verantwortbarkeit für die darin enthaltenen
oder eingespeisten und oft gar nicht mehr zuordenbaren Daten aufwerfen. Die
humane Verantwortung auch für die Folgen und Entwicklungen in verwickel-
ten Informationssystemen kann weder ethisch noch rechtlich von einem
einzelnen getragen werden. Diese Fragen sind durchaus noch sehr offen und
man kann sich hierfür durchaus noch keine Lösung vorstellen, wie eine opera-
tionalisierbare, eine greifbare und handhabbare Ethik bzw. Rechtsprechung
oder Legislative bezüglich der weltweiten Informationssysteme aussehen wird.
In gewissem Sinne sind wir natürlich aber immer für eine konkrete Huma-
nität und deren Bewahrung im Umgang auch mit Informationen verantwort-
lich, ähnlich wie wir im Umgang mit unseren Handlungspartnern verant-
wortlich. Für die Fragen der Tradition und der Entwicklung der konkreten
Humanität sind natürlich die philosophischen und geisteswissenschaftlichen
Disziplinen zuständig, aber das ist eine Frage, die heute und hier nicht be-
handelt werden kann. Jedenfalls kann es nicht mehr angehen, daß die Problem-
bereiche und Disziplinen voneinander abgeschottet werden. Die Probleme
stellen sich nicht in einer schubladenartigen Auftrennung der Fächer, und
insofern können natürlich auch die einzelnen Disziplinen nicht mehr un-
abhängig voneinander operieren, sondern sie alle müssen die interdisziplinäre
Herausforderung aufnehmen. Das gilt zumal natürlich auch für die Geistes-
wissenschaften. Die Systemzusammenhänge im systemtechnologischen Zeital-
ter erfordern eine Anwendung von abstrakteren Verfahren und Verallgemeine-
rungen sowie von formalen und funktionalen Gesichtspunkten der Darstel-
lungsweisen: Es wird immer wichtiger, eine fachübergreifende Erfassung, d. h.
also eine Beschreibung und auch einen praktischen handelnden Umgang mit
Gegenständen, Verfahren, Systemen, zu entwickeln, der nun “interdisziplinär”
im echten Sinne genannt werden kann. Es gilt also eine praxisnahe und wirk-
lichkeitsangemessene Methodologie, aber auch eine Erkenntnistheorie, Episte-
mologie zu entwickeln, die diese Herausforderung annehmen kann und die
interdisziplinäre Verfassung und Verflechtung der Problembereiche angemes-
sen berücksichtigt.
Dabei ist es wichtig zu beachten, daß es in der Tat Tendenzen zu einer
Verselbständigung der Systemoperationen und Systeme selbst, der System-
technokratie bzw. der Computerokratie gibt, die natürlich auch nur durch eine
überfachliche Abstimmung, Sicherung, Kontrolle der vielfältigen Einfluß-
größen und Gesichtspunkte, die über die fachspezialistische Einseitigkeit
136 Hans Lenk
hinausgeht, aufgenommen werden kann. Demzufolge brauchen wir außer der

Teamzusammenarbeit unterschiedlicher Fachspezialisten und Experten aus
unterschiedlichen Fakultäten auch Generalisten, die abstraktere Methodolo-
gien, methodische und operationale Ansätze beherrschen und in verschiedenen
Gebieten anwenden können. Das gilt ferner auch für die “Spezialisten für das
Allgemeine”, die Universalisten, die übergreifende Ziel- und Wertprobleme
und die wissenschaftstheoretischen und methodologischen Grundlagen der
entsprechenden Disziplinen mitbehandeln können.
Diese interdisziplinäre Verfassung und Verflechtung der Probleme stellt
sich in dem Kreuzungsbereich vieler klassischer Einzelfächer dar. Es handelt
sich aber um überfachliche oder nur transdisziplinär zu erfassende Fragen-
komplexe, also um die Notwendigkeit, interdisziplinäre und gar supradiszipli-
näre Ansätze zu entwickeln. Die klassischen Einzelwissenschaften sind bei
diesen bereichsübergreifenden Problemen häufig überfordert. Die Vielfalt der
disziplinären Perspektiven und die Widersprüchlichkeiten der Expertenurteile
aus einzelwissenschaftlicher Sicht führen zu wichtigen organisatorischen und
methodologischen Problemen für die Anwendung der wissenschaftlichen
Ansätze, der Expertisen und Resultate, die nun über die Einzelwissenschaften
hinausgehen, die sich also der klassischen Schubfächereinteilung entziehen.
Hier gibt es eine Reihe von Beispielen auch von neuen Forschungsbereichen,
die sich von vornherein in gewisser Weise interdisziplinär angelegt haben bzw.
verstehen: einerseits beispielsweise – ganz aktuell – die Umweltforschung, oder
man denke an die Wissenschaftsforschung, Wissenschaftswissenschaft (oder
“Science of science”), die Peter Weingart eine “multidisziplinäre Aggregatwis-
senschaft” genannt hat, bei der weniger die Methode als der Gegenstand, eben
der Problembereich, der Untersuchung zugrunde liegt. Die Teilforschungs-
bereiche beispielsweise dieser Wissenschaftsforschung sind etwa Wissen-
schaftsgeschichte, Wissenschaftssoziologie, Wissenschaftsökonomie, Wissen-
schaftspsychologie, Organisationswissenschaft, Planungswissenschaft, Teile
der Politikwissenschaft und natürlich Wissenschaftstheorie, Methodologie,
also Philosophie der Wissenschaft. Es gibt hier offensichtlich noch keine
direkte Möglichkeit, eine wirkliche interdisziplinäre Theorie zu entwickeln,
deshalb “Aggregatwissenschaft”. Wie kann man nun unterschiedliche Diszipli-
nen und unterschiedliche Arten der Disziplinarität voneinander unterschei-
den? (Vgl. das Diagramm.)
Unterscheidung der Disziplinen

nach ihren/m/r
1. Gegenständen und Bereichen
2. Methoden(arsenalen)
3. Erkenntnisinteressen (Habermas)
4.“Theorien und deren systematischen und historischen Zusammenhän-
gen” (L. Krüger 1987)
5. Theorie-Praxis-Verhältnis
6. Substantivität versus Operationalität versus Formalität der Theorien
(Bunge)
7. Systemholismus versus Bereichsspezifik
8. Apriori-Analytik, Methodenformalität versus Empirik
9. Erklärungs- und Systematisierungsmustern (beschreibend versus
erklärend, historisierend versus systematisierend)
10. Kognitivität und Normativität (deskriptive versus normative Diszi-
plinen)
11. Fiktionalität und Sekundärrealitat (soziale “Geltung”) versus Pri-
märrealität (“Imprägnation”)
Die Disziplinen werden herkömmlich 1. nach Gegenständen und Bereichen
voneinander unterschieden oder 2. nach ihren Methoden oder Methoden-
kombinationen, Arsenalen, 3. nach “Erkenntnisinteressen”, (Habermas) was
immer das im einzelnen genauer heißen mag. 4. sind auch die Theorien und
deren systematische und historische Zusammenhänge entscheidend und unter-
scheidend für die Disziplinen. Diese Zusammenstellung bis hier ist von Lorenz
Krüger (1973, S. 111 ff.), dem leider verstorbenen Wissenschaftstheoretiker.
Wichtig sind aber auch noch die folgenden Punkte: 5. die unterschiedlichen
Bezüge zu Theoriepraxis in den unterschiedlichen Wissenschaften; man denke
beispielsweise an komplexe Systemtheorien in der Mathematik und der Um-
weltforschung. 6. ist hervorzuheben – was ich für besonders wichtig halte und
was bisher nicht genügend berücksichtigt worden ist – der Unterschied zwi-
schen inhaltlichen, den sog. “substantiven Theorien” (Bunge 1967) und “ope-
rativen Theorien”, die eher Verfahren betreffen. Substantive Theorien wären
beispielsweise die Gravitationstheorien nach Newton oder Einstein in der
Physik, operative Theorien wären beispielsweise die Informationstheorie, die
mathematische Spieltheorie oder generell formale Verfahren, analytische
Instrumente, die in verschiedenen Wissenschaften angewendet werden kön-
nen. Offensichtlich sind operative Theorien in der interdisziplinären For-
schung besonders sinnvoll verwendbar ; die formalen Theorien sind dabei
natürlich diejenigen, die nur formale idealsprachliche Konzepte erarbeiten, wie
138 Hans Lenk
beispielsweise mathematische Theorien. Ein 7. Gesichtspunkt wäre der Punkt

“Systemholismus gegenüber Bereichsspezifik”. Systemorientierte Wissen-
schaften achten auf Gesamtzusammenhänge, auf “holistische” Problemstel-
lungen im Gegensatz zu bereichsspezifischen Einzeluntersuchungen. 8. Ein
Unterschied, der auch für die Trennung zwischen Disziplinen wichtig ist, ist
die Apriori-Analytik, die Methodologie apriorischer (also von der Erfahrung
bereits vorausgesetzter) Art, z. B. formale Methoden gegenüber empirischen,
beschreibenden, beobachtenden Vorgehen. (Beispiele liegen auf der Hand,
sagen wir einmal in der Logik einerseits und der Paläontologie andererseits.) 9.
ist zumal in den Geisteswissenschaften dann bedeutsam und bekannt die
Unterscheidung zwischen “erklärenden” und eher “historisierenden” Theo-
rien. Es handelt sich um Erklärungs- und Systematisierungsmuster, also syste-
matisierend oder theoretisch-begrifflich verallgemeinernde erklärenden Theo-
rien auf der einen Seite gegenüber eher beschreibenden und historischen
Ansätzen. 10. gibt es den Unterschied zwischen kognitiven und normativen
Disziplinen. Kognitiv-deskriptive Disziplinen sind natürlich die Naturwissen-
schaften, während beispielsweise die Rechtswissenschaft als eine normative
Disziplin aufgefaßt werden kann. 11. Fiktionalität und soziale Geltung, also
sekundäre Existenz von “Gegenständen”, gegenüber einer Primärrealität, die
in den Naturwissenschaften unterstellt wird. Ich komme darauf im zweiten
Teil noch zurück. Ich nenne den Beitrag , den die Natur oder die Welt an sich
ausüben, der einen begrenzenden Einfluß auf die durchaus aktive Theorie-
bildung hat, die Wirkung der Imprägnationen im Gegensatz oder Unterschied
zu rein fiktiven produzierten Interpretationen im engeren Sinne.
Das alles sind also Gesichtspunkte, die zur Unterscheidung der Disziplinen
Anlaß geben, die aber dann auch zu verschiedenen Typen der Interdisziplina-
rität führen.
Ich habe im folgenden Diagramm zehn unterschiedliche Typen der Inter-
disziplinarität aufgeführt:
Typen der Interdisziplinarität
1. Interdisziplinäre Projektkooperation
2. Bidisziplinäres bzw. interdisziplinäres Forschungsfeld
3. Multidisziplinäre Aggregatwissenschaft (Sammeldisziplin) WEIN-
GART
4. (echte) Interdisziplin (Bidisziplin)
5. Multidisziplin (multidisziplinäre theoretische Integration)
6. Generalisierte interdisziplinäre Systemtheorie(n) (“Allgemeine Sys-
temtheorie”) V. BERTALANFFY
7. Mathematische Theorien abstrakter und komplexer dynamischer

Systeme (“Chaostheorie”)
8. Supradisziplinäre angewandte Struktur- und Operationsdisziplinen
(“Operations Research”)
9. Methodologisch-metatheoretische Supradisziplinen (Wissenschafts-
theorie, Wissenschaftswissenschaft)
10. Philosophisch-methodologisch-erkenntnistheoretische Metadiszi-
plin (“methodologischer Schemainterpretationismus”)
Zunächst 1. die einfache bloße Projektkooperation interdisziplinärer Art ; man
denke z. B. an die ursprüngliche Entwicklung in der Stadtplanung, wo bei-
spielsweise Experten ganz unterschiedlicher Ausrichtungen mitwirken bzw.
eine Rolle spielen. Ich habe das bei der Neuplanung eines Teils der Innenstadt
von Karlsruhe (des sog. “Dörfle”) z. T. miterlebt. Dann gibt es 2. eine bi-
disziplinäre oder interdisziplinäre Verortung eines Forschungsfeldes oder 3.
eine entsprechende multidisziplinäre Zusammenwirkung von Forschungs-
vorhaben, beispielsweise in der Umweltforschung, die jetzt eine Art von
Sammeldisziplin geworden ist, evtl. 4. eine multidisziplinäre Aggregatwissen-
schaft (nach Weingart) ist im Unterschied zu 5. einer spezifischen echten
Interdisziplin, wie beispielsweise der Molekularbiologie oder der Biochemie
oder besser vielleicht noch der physikalischen Chemie, um eine traditionelle
Interdisziplin zu nennen. Dann gibt es 6. die generalisierten interdisziplinären
Systemtheorien, wie etwa die seit den 30er Jahren existierende Allgemeine
Systemtheorie nach Ludwig von Bertalanffy. 7. sind die abstrakten rein forma-
len mathematischen Theorien der komplexen dynamischen Systeme anzufüh-
ren, die insbesondere in den letzten Jahren besondere Aufmerksamkeit erlangt
haben, z. B. die sogenannten Theorien des deterministischen Chaos und die
darüber hinaus auch erst zu entwickelnden probabilistischen Chaostheorien,
die bisher praktisch noch nicht existieren. (Die Anwendung der Chaostheorie
auf die Sozial- und Geisteswissenschaften steht also noch aus, denn es handelt
sich hier nicht um deterministische Theorien.) 8.wäre zu nennen eine supra-
disziplinäre angewandte Struktur- und Operationswissenschaft bzw. eine
Disziplin, wie sie etwa in den Verfahrenswissenschaften der Ökonomie in
Gestalt des Operations Research schon lange betrieben wird. 9. gibt es dann
methodologisch metatheoretische Supradisziplinen, durchaus auch solche
traditioneller Art wie die Wissenschaftstheorie und die erwähnte Wissen-
schaftswissenschaft, der Wissenschaftsforschung. 10. wäre schließlich noch zu
erwähnen und zu entwickeln eine philosophisch-methodologische oder meta-
theoretische Erörterung der entsprechenden Systemzusammenhänge bzw. der
betreffenden Disziplinen unter einem gewissen Gesichtspunkt, nämlich dem
140 Hans Lenk
methodologischen Interpretationismus, den ich im zweiten Teil diskutieren

werde.
Soweit also zunächst ein Überblick über zehn verschiedene Aspekte oder
Möglichkeiten, Typen von Interdisziplinarität aufzustellen und nach gewissen
Gesichtspunkten der Schärfe oder Stärke des Zusammenhangs zu unterschei-
den. Bloße Sammeldisziplinen, die lose ein Forschungsfeld abdecken, sind
etwas völlig anderes als eine exakte Interdiziplin wie die physikalische Chemie
oder als eine mathematische operative Theorie wie die deterministische Cha-
ostheorie. Hier muß man also meines Erachtens deutliche methodologische
Unterscheidungen einführen.
Für alle diese Gesichtspunkte gilt aber, daß Wissenschaftler, wenn sie sich
dieses interdisziplinären Vorgehens befleissigen, so etwas wie eine Zweit-
kompetenz in der betreffenden oder wenigstens einer entsprechenden Nach-
barwissenschaft brauchen. Hermann Lübbe sprach einmal von “Mitführ-
kompetenz” in einer anderen Wissenschaft, so ist es z. B. ganz klar, daß ein
Wissenschaftstheoretiker, der sich systematisch mit Methodenfragen der
Biologie befassen will, auch den Stand der Biologie einigermaßen übersehen
muß ; er muß zwar nicht ein produktiver Forscher in der Biologie sein, aber
doch in der Lage sein, den derzeitigen Entwicklungsstand der Biologie zu
beurteilen. Eine solche Mitführ- oder Zweitkompetenz müßte natürlich auch
dann für die Ausbildungsgänge gefordert werden. Diese Ausbildung von
“Mehrseitigkeiten” und Vielfachkompetenzen erfordert einerseits das Einge-
hen auf unterschiedliche Wissenschaften – das ist für den einzelnen natürlich
nur begrenzt möglich – , andererseits aber auch die Entwicklung gerade der
allgemeineren generellen Systemkompetenzen, die schön erwähnt worden
sind, insbesondere der abstrakteren und formaleren Methoden der Generalis-
ten und eben auch der Fähigkeit der Universalisten, über die spezifischen
Fachorientierungen hinaus beispielsweise übergreifende Wert- und Normen-
systeme diskutieren, erforschen zu können. Die relativ beste Lösung, die man
sich vorstellen kann, ist natürlich nicht diejenige, die nun dem enzyklopä-
dischen Gehirn des Universalisten entspringt, sondern eben eine Gemein-
schaftsschöpfung in Teamarbeit durch Wissenschaftler unterschiedlicher
Provenienz, insbesondere soweit es unterschiedliche Disziplinen betrifft, die in
einem Forschungsbereich relevant sind.
Heinz Heckhausen (1987, S. 135) hat einmal gesagt, daß in den Geisteswis-
senschaften sowieso im wesentlichen intradisziplinär geforscht werde, weil
eben alles historisiert werde und man im Grunde eigentlich immer nur unter-
schiedliche Textgrundlagen habe, aber die Forschungsmethode eigentlich
einheitlich dieselbe sei. Ich denke, daß das heutzutage nicht mehr gilt. Was
allerdings gilt, je stärker Kontexte, historische Traditionen usw. in Ansätze
eingehen, desto stärker wird natürlich auch dieser Gesichtspunkt der Be-
teiligung unterschiedlicher Disziplinen relevant werden. Es wird heutzutage in
den Geistes- und Sozialwissenschaften ein besonders hoher Anspruch an
interdisziplinäre Forschung und Lehre gestellt. Ich selber versuche in Karls-
ruhe seit fast drei Jahrzehnten in interdisziplinären Seminaren immer eine
Zusammenarbeit mit den Fachexperten der entsprechenden Disziplinen durch-
zuführen. Meine Erfahrungen zeigen, daß interdisziplinäre Lehre – insbeson-
dere an Technischen Universitäten – auf die Praxisnähe, die Problemorientie-
rung und die Projektnähe achten muß ; es ist am besten, im Rahmen eines
praktischen Forschungsprojekts oder eines begrenzten thematischen Opera-
tionsfeldes interdisziplinäre Veranstaltungen zusammen mit Experten der
entsprechenden anderen Fächer und Fakultäten anzubieten. Das heißt also, die
Zusammenarbeit muß eigentlich, wenn sie sinnvoll betrieben werden soll, auf
Dauer gestellt werden und auch in fortwährender Konfrontation und Kom-
munikation projektbezogen und problemorientiert ausgefächert werden. Dies
setzt bei den Beteiligten eine Art von Mitkompetenz voraus, zumindest eine
Art von Bereitschaft, die Fachsprache der anderen Disziplin wenigstens ver-
stehen zu lernen und sich damit vertraut zu machen. Man unterstellt dabei
auch, daß es so etwas wie eine gemeinsame Basis der Methoden und Zugangs-
weisen, der Methodologie der Wissenschaften, also der Wissenschaftstheorie,
gibt und auch, wie ich darüber hinaus sagen möchte, der “Methodologie des
Handelns”, des handelnden Forschens. Das impliziert per se, daß es interfakul-
tative und interdisziplinäre Ansätze und evtl. Institute, die möglichst praxisnah
solche Gesichtspunkte einbringen. Das hat unsere eigene Fakultät nun auch
versucht. Wir haben vor einem halben Jahrzehnt schon ein interdisziplinäres
Institut für Angewandte Kulturwissenschaft gegründet, ein interfakultatives
Institut für Kulturwissenschaft, das versucht, angesichts der unterschiedlichen
Fragen der Kulturerforschung sich solche Zielsetzungen zu setzen und eine
Art von Zusatzqualifikation, eine Art von “Begleitstudium” für Studenten
anbietet, die z. B. insbesondere in die Medienpraxis gehen wollen. Unsere
Fakultät hat neuerdings auch “berufsorientierte Zusatzqualifikationen” konzi-
piert, eine Art von kleiner Bestätigung von zusätzlichen Kompetenzen, welche
die Möglichkeiten ergeben, daß jemand beispielsweise eine berufsqualifizieren-
de Kenntnis erwirbt, die ihm als Geisteswissenschaftler beispielsweise gerade
bei der Arbeit für Rundfunk und Fernsehen, bei der Praxis in den Medien,
insbesondere Umgang auch mit Multimedia oder auch bei der interkulturellen
Kommunikation eine Hilfe bietet.
Entsprechende interdisziplinäre Initiativen finden wir beispielsweise auch
in Straßburg, in der Université Louis Pasteur: nämlich im Centre de Recherche
Transdisciplinaire sur les Sciences et les Techniques, das ja seit längerem ver-
142 Hans Lenk
sucht, für Naturwissenschaftler solche Fragestellungen zu diskutieren, die über

die disziplinären Gesichtspunkte hinausgehen und ein interdisziplinären
Diskurs zu pflegen und den interuniversitären Charakter der Interaktionen
zwischen harten Wissenschaften einerseits und Disziplinen der Humanwissen-
schaften andererseits zu betonen. Es geht, wie Pierre Karli (1993, S. 172 f.)
sagte, um eine “Entkapselung” der Einzeldisziplinen innerhalb der Universität
selber sowie um eine größere Öffnung der Universität auch ihrer Umgebung,
der sozialen Umwelt gegenüber. Man wolle sich Gedanken machen “über sich
selbst”, d. h. über die Naturwissenschaften und Technikwissenschaften und
über den “Sinn” des Beitrages zur Förderung von Wissenschaft und Technik.
Das sind ganz parallele Entwicklungen in unseren Universitäten, die, so glaube
ich, recht wichtig sind und die Transdisziplinarität der Problemstellungen auch
explizit hervorheben und darstellen.
II.
Der zweite Teil hat die Aufgabe, theoretische Folgerungen aus der skizzierten
Problemlage zu ziehen. Dabei möchte ich gerade auch die Implikationen für
die Sozial- und Geisteswissenschaften kritisch unter dem Gesichtspunkt des
herkömmlichen traditionellen Separatismus der Methoden zwischen Natur-
wissenschaften und Geisteswissenschaften diskutieren. Es handelt sich um die
bekannte Zweikulturentrennung im Anschluß an C. P. Snow (1967), die nur
zu begeistert von Profilneurotikern beider Seiten aufgenommen worden ist,
obwohl Snow eigentlich gar nicht die naturwissenschaftliche “Kultur” gegen-
über der geisteswissenschaftlichen “Kultur” gemeint hat, sondern von der
naturwissenschaftlichen und der “literarischen” Kultur sprach. Er kritisierte
nicht einen Gegensatz zwischen Wissenschaftsarten, sondern einen solchen
zwischen allgemeineren Aktivitäten intellektueller Art. Es ist natürlich ein
Unding – da hat Snow sicherlich recht – , daß nach wie vor der Zweite Haupt-
satz der Thermodynamik von Geisteswissenschaftlern oder Literaten eher
kühl und verächtlich als belanglos angesehen wird, wenn er überhaupt zur
Kenntnis genommen wird, und daß ebenso umgekehrt die Naturwissenschaft-
ler, nach ihrer Shakespeare-Lektüre befragt, meistens auch “passen”. Ich kann
aus meiner eigenen Erfahrung und von unserer Fakultät an einer Technischen
Universität Beispiele dafür anbringen: z. B. hat ein ehemals bekannter Mittel-
altergermanist unserer Universität auf einer Party gesagt, es gehöre ja wohl
nicht zur Bildung, die Einheit der komplexen Zahlen zu kennen. Solche Be-
hauptungen zeigen Arroganz und Ignoranz derjenigen auf, die solches be-
haupten. Hingegen hatte ich selber in der Schule einen sehr gebildeten Che-
mielehrer in der Schule, der beklagte, daß der Mythos von Achilles und der
Schildkröte, jedoch nicht die Jod-Stärke-Reaktion zur Bildung zählte. Ich

denke, daß solche Entgegensetzungen viel zu grob sind.
Ferner fordern manche, daß man mindestens eine “dritte Kultur” (Lepe-
nies), nämlich die der Sozialwissenschaften konstatieren müßte. Ich glaube,
nicht einmal das reicht, um die genannte Polarität zu überwinden. Es gibt
unterschiedliche Zwischenphänomene, Mischlingsdisziplinen und neuerdings
eben sehr viele interdisziplinäre Gebiete, wie ich erwähnt habe, z. B. Aggregat-
disziplinen, die eigentlich dieser herkömmlichen Einteilung glatt zuwider-
laufen. Man hatte ja schon immer und traditionellerweise Schwierigkeiten, z.
B. die Logik oder die Mathematik einzuordnen, denn diese sind eigentlich
Geisteswissenschaften, Präzisionsgeisteswissenschaften par excellence. Sie
werden aber sozusagen”natürlich”, aber eigentlich fälschlich in den naturwis-
senschaftlichen Fakultäten verortet. Dasselbe gilt natürlich in gewissem Sinne
ja für die Linguistik, die heutzutage z. T. auch als halb-mathematische Diszi-
plin unter Rückgriff auf Halbgruppentheorie und Halbverbandstheorie betrie-
ben wird ; auch bei ihr kann man nicht unmittelbar sagen, sie sei nur eine
herkömmliche Geisteswissenschaft. Sogar in den Naturwissenschaften gibt es
auch historische Disziplinen, wie z. B. die Paläontologie oder z. T. auch die
physikalische Kosmologie bzw. die Kosmologie generell. Der herkömmliche
Separatismus ist also sicherlich abzulehnen, faktisch und methodologisch
falsch, überholt und eher “ideologisch” motiviert.
Das gilt erst recht, glaube ich, für die traditionellen Unterscheidungen
zwischen “Verstehen” und “Erklären”. Wollte man sagen, der Naturwissen-
schaftler erklärt nur, aber versteht nichts? Das kann man ja wohl kaum sagen.
Und sollte vom Geisteswissenschaftler umgekehrt gelten: er “versteht” nur,
aber er “erklärt” nichts? Das ist pauschal gesehen ebenso unsinnig. Das heißt,
es geht also darum, diese unfruchtbare Dichotomisierung auf einer höheren
Metastufe, auf einer methodologischen Metastufe zu überbrücken, zu über-
schreiten oder gar zu überwinden.Ich denke, daß eine solche Möglichkeit
dadurch gegeben ist, daß wir eine allgemeinere konstruktive Theorie der
Interpretationen oder, wie ich besser neuerdings sage, der Schemainterpreta-
tionen oder Schemaaktivierungen aufstellen, nämlich einen methodologischen
Ansatz entwickeln, der gewisse Voraussetzungen der traditionellen methoden-
orientierten Philosophie mit bestimmten Überlegungen eben auch der Er-
kenntnistheorie der Naturwissenschaften verbindet oder kompatibel werden
läßt. Dies sind Überlegungen, wie sie etwa z.T. von Kant vorgegeben sind, bei
Peirce mit seiner Zeichen- und Symboltheorie sowie in Cassirers Symbol-
philosophie zu finden sind – und insbesondere bei Nietzsches Auffassung der
Grundtätigkeit des Interpretierens. Diese Wurzeln kann man, glaube ich, zu
einem konsistenten Ansatz zusammenführen, den ich einen methodologischen
144 Hans Lenk
Schemainterpretationismus nennen möchte. Der Grundgedanke ist, daß alles

Erfassen – und zwar sowohl das passive, das theoretische Ergreifen und “Fas-
sen”, z. B. Wahrnehmen, wie auch das Handeln unter Ordnungsmustern,
Schemata und in diesem Sinne unter Schema-Interpretationen steht, regelhaft
ist. Schemata werden etwa in der Psychologie verstanden als hierarchisch
strukturierte, relativ dauerhaft oder gar permanent gespeicherte Wissens- und
Handlungsabrufstrukturen, die etwas repräsentieren können, die unter Um-
ständen konventionell zustande gekommen oder erlernt sind, die insbesondere
Konfigurationen und Merkmalskombinationen repräsentieren. Sie enthalten
zu spezifizierende, d.h. durch konstante Größen oder Werte zu ersetzende,
Schemavariable, die dann geeignet sind, diese Schemata auf spezifische Situa-
tionen zu beziehen.
Die Schemata wurden eingespielt, sind interpretationsgebunden oder gar
interpretationsimprägniert in dem Sinne, daß unter Umständen äußere Welt-
faktoren (wie beim direkten Wahrnehmen) eine prägende, ja, determinierende
Rolle mitspielen. In diesem Sinne ist also das Schemainterpretieren oder das
schlichte “Interpretieren” (obwohl das etwas mißverständlich ist) etwas All-
gemeineres als das spezifische Textinterpretieren der Hermeneutik. Textinter-
pretieren ist ein Spezialfall des Schemainterpretierens. Ich komme darauf noch
zu sprechen. Bei diesem methodologischen Interpretationismus, genauer:
Schemainterpretationismus, wird ernst gemacht mit der Geformtheit aller
unserer Erfassungen, sowohl im Erkennen als auch im Handeln – und ins-
besondere natürlich bei symbolischen Repräsentationen und Erfassungen
sprachlicher Art und auch anderer Schematisierungen biofunktionaler oder
konstruktiver bzw. darstellender Art.
Es handelt sich hier also um einen methodologischen Ansatz, der relativ
abstrakt ist, genauer um eine Methodologie der Methodologie, also um Meta-
methodologie – wir hatten ja auch von Metakonzepten gesprochen. Man
könnte den Ansatz auch als einen für Erweiterungen offenen, hypothetisch-
tentativen (nicht dogmatisch fixierten) quasitranszendentalen Ansatz im Sinne
Kants bezeichnen. Die Grundidee ist, daß alles, was wir erfassen können, unter
der Bildung der relativen Stabilisierung, also Etablierung, Stabilisierung, und
der Aktivierung oder Reaktivierung von bestimmten Schemata geschieht wie
man sich neuerdings durchaus auch im Rahmen der Neuro- und Kognitions-
wissenschaften als durch die Einspielung von sog. Neuronenassemblies (Neu-
ronenensembles) plastischer Art getragen vorstellen kann. Die Neurowissen-
schaftler sprechen ja neuerdings sogar von “Hirnkonstrukten” (z. B. Singer
1990, S. 8) und meinen damit plastische, aber durch wiederholte Aktivierung
relativ stabilisierte Neuronenassemblies, die immer wieder reaktiviert werden
können. Zumindest kann man auf diese Weise eine “Wie-ist-es-möglich?”-

Erklärung der Bildung von Schemata verständlich machen. Damit ist natürlich
die semantische Lücke zwischen der naturwissenschaftlich-physiologischen
Beschreibung von Neuronenassemblies einerseits und der Deutung von eben
semantischen Gehalten andererseits noch nicht überdrückt, aber die Korrelati-
on ist etwas enger und nachkontrollierbarer geworden. Das ist ein recht
schwieriges Phänomen, das ich hier leider nicht ausbreiten kann. Ich denke
aber, daß in dieser Richtung eine Art von Ansatz und Kooperation interessant
wird, der für Natur-, Geistes-, Sozial- und eben auch Neurowissenschaften
eine Art von Überbrückung und gemeinsamem Ansatz liefert.
Ich möchte noch ein wenig auf die schematisierenden interpretierenden
Tätigkeiten, Aktivitäten eingehen und diesbezüglich einige Unterscheidungen
oder Schichtungen vorstellen. Das läßt sich am besten anhand einiger Dia-
gramme zeigen (vgl. Lenk 1993, S. 109 ; 1993a, S. 254):
INTERPRETATORISCH-SCHEMATISIERENDE AKTIVITÄTEN:
(SCHEMA-) INTERPRETATIONEN
konstituierende -> konstruierende -> rekonstruierende Aktivitäten
Konstituie- bewußtes Entwerfen, Anwenden, (Re-)Identifizie-

ren, unbe- Auslösen, Zuordnen, Projizieren, ren, (Wieder)er-
wußtes Aus- Diskriminie- Aufproje- Durchfüh- kennen, Unter-
lösen, Akti- ren, Kon- zieren, ren, Kon- scheiden,
vieren, Aus- trastieren, Variieren, struieren, Zuordnen durch
bilden, Ent- Vergleichen, Kombinie- Repräsen- Einsetzen, Sub-
wickeln, (Re-) Identifi- ren, Orga- tieren, Vor- sumieren, sortie-
Differenzie- zieren, Dar- nisieren, stellen, Kog rendes Klassifi-
ren, Stabili- stellen, Aus- Integrieren nizieren, zieren, Verste-
sieren wählen, Ver- Darstellen hen i.w.S. suk-
feinern zessives Weite-
ranwenden
von durch, mittels, mit nach, oder in

Schemata Schemata (von Konstanzen,
Formen/Strukturen/Gestalten,
Gegenständen, Ereignissen,
Prozessen, Fakten, Relationen,
Kontexten)
bei Textinterpretation: Re-Iden-

tifizieren Anwenden von Sche-
mata (Wieder)Erkennen Ver-
stehen i.e.S.
146 Hans Lenk
So steht das herkömmlich verstandene Interpretieren oder Schemaaktivieren

unter dem Gesichtspunkt der interpretatorisch-schematisierenden Aktivitäten
oder Schemainterpretationen. Dafür sind in der Abbildung Beispiele aufge-
führt und ein wenig nach ihrem Konstruktcharakter geordnet. Man findet auf
der linken Seite eher Tätigkeiten wie “Schemata etablieren” und “stabilisie-
ren”, die Gegenstandserfassungen und – formierungen konstituieren; das sind
also konstituierende interpretatorisch-schematisierende Tätigkeiten. In der
Mitte des Diagramms sind dann im engeren Sinne, konstruktiv bewußt-plan-
mäßig entworfenen Darstellungen oder Formen angeführt, und rechts die eher
wiederholenden, wiederaufgreifenden, reaktivierenden, rekonstruierende
Aktivitäten oder Schemaaktivierungen wie das Wiederlesen oder Erinnern.
Um einige Beispiele zu nehmen: Zum Konstituieren gehört sicherlich das
Stabilisieren eines Schemas bei einem Lernprozeß, zum direkten Wahrnehmen
aber speziell auch das unbewußte Auslösen durch einen äußeren Reiz. Zum
Konstruieren i. e. S. gehört das Variieren: z. B. eines Grundmusters von vor-
gegebenen Kriterien, das bewußte Variieren. Zum Rekonstruieren gehört das
Subsumieren unter einen bekannten Begriff, das einordnende Klassifizieren
oder Sortieren – und last but not least natürlich das Lesen, Erkennen, Deuten
von Texten. Textinterpretation ist natürlich eine rekonstruierende
interpretatorisch-schematisierende Aktivität. Insofern sind natürlich die
Textinterpretationen oder die traditionellen Schemata der Hermeneutik, die
auf das wiedererkennende Anwenden von Schemata und das Verstehen im
engeren Sinne ausgerichtet sind, ihrerseits Spezialfälle dieser schematisierenden
Aktivitäten. Darüber könnte und müßte man natürlich im einzelnen ausführ-
licher diskutieren und referieren, als das hier geschehen kann.
Stattdessen möchte ich, um den methodologischen Ansatz und auch die
philosophische Relevanz deutlich zu machen, auf einige Gesichtspunkte der
Schichtung der Schematisierungen eingehen. Man kann nämlich: Stufen oder
Ebenen des Interpretierens, des Schemainterpretierens unterscheiden.
(Ebenen) Stufen der Schema-interpretation
IS1 praktisch unveränderliche Urschematisierung oder die Imprägnation (gene-

tisch angelegte primäre Konstitution bzw. Schematisierung oder die Im-
prägnation in der direkten Wahrnehmung)
————————————
IS2 gewohnheits-, gleichförmigkeitsbildende Musterbildung und -reakti-
vierung (erlernte Muster – Schematisierung), habituelle Form- und
Schemakategorialisierung + vorsprachliche Begriffsbildung
IS3 sozial etablierte, kulturell tradierte, übernommene konventionalisierte
Schematisierung
IS3a vorsprachlich normierte “Begriffs”bildung und Schema-Interpretati-
on durch soziale und kulturelle Normierungen
IS3b repräsentierende sprachlich normierte Begriffsbildung (i. e. S.)
————————————
IS4 anwendende, aneignende bewußt geformte Einordnungsinterpretation
und Unterordnungs- oder Einbettungsschematisierung (Klassifikation,
begriffliche Subsumierung, Beschreibung, Artenbildung u. -einordnung;
gezielte Begriffsbildung)
IS5 erklärende, “verstehende” (i. e. S.) rechtfertigende, (theoretische) be-
gründende Interpretation Rechtfertigungsinterpretation, Einbettung in
Folgerungs-, Argumentations- und Begründungs-“Schemata” oder -
Metaschemata
IS6 erkenntnistheoretische (methodologische) Metainterpretation der Inter-
pretationskonstruktmethode
Ich unterscheide sechs verschiedene Stufen (vgl. Lenk 1993, S. 56 ; verändert

Lenk 1995, S. 103): einmal praktisch nicht von uns veränderbare, aufgrund
unserer biologischen Anlage Urinterpretationen (IS1) – beispielsweise bei
einem Normalsichtigen die Unterscheidung zwischen “hell” und “dunkel”:
das ist nicht vermeidbar. Das nenne ich also Urinterpretationen oder primäre
Schematisierung. Zweitens – es wird von Schicht zu Schicht immer flexibler
und plastischer – gewohnheitsbegründete oder gewohnheitsmäßige,
gewohnheits- oder gleichförmigkeitsbildende Schema- oder Musterinter-
pretationen (IS2), die durch Habitualisierung, durch gemeinsames Verhalten
vorsprachliche Begriffsbildung z. B. eingespielt worden sind, also durch Mus-
terstabilisierung oder – eingewöhnung. Drittens dann – für uns natürlich
148 Hans Lenk
besonders interessant – die konventionellen Begriffsbildungen, konventionelle

Schema-Interpretationen i. e. S. (IS3), die aufgrund von Verabredungen und
Normen oder Normierungen durch die Interpretationsgemeinschaft (Kultur,
Sprache, Gesellschaft) zustande kommen: Dazu zählen alle symbolischen
Scheam-Interpretationen ; denn Symbole sind ja konventionelle Zeichen. Hier
sind auch alle Verabredungen bzw. Interpretationen aufzuführen, also sozial
etablierte kulturell tradierte und aus der Interpretations- oder Sprachgemein-
schaft “übernommene” Begriffsbildungen. Hier muß man natürlich die Unter-
scheidung treffen zwischen solchen Begriffsbildungen, die aufgrund von
Gesten oder Verhaltensweisen konventioneller Art, aufgrund von Normen
oder Gewohnheiten, Sitten, Gebräuchen eingespielt worden sind (IS3a) – ohne
schon die Sprache explizit zu benutzen – und den sprachlich repräsentieren-
den, explizit sprachlichen Begriffsbildungen (IS3b).
Die nächste Stufe wäre, daß man das symbolische Repräsentieren auf
bestimmte Musteranwendungen und -einbettungen anwendet: die vierte Stufe
wären also die Unterordnungsschematisierungen und Einordnungsinter-
pretationen (IS4): Klassifikation, Subsumierung, Beschreibung, Artenbildung
und -einordnung, gezielte Begriffsbildungen alle diese Aktivitäten wären also
i. e. S. begriffliche, also verallgemeinernde schematisierende Einordnungsinter-
pretationen. Die nächsthöhere Stufe stellt dann die der Rechtfertigungs- oder
Begründungsinterpretationen (IS5) dar, die argumentative Interpretation,
wobei man eben erklärt, versteht, begründet, rechtfertigt.
Und die letzte Stufe (IS6) leistet die Thematisierung der interpretatorischen
Tätigkeit des Methodologen, des Interpretationstheoretikers, selber, nämlich
die Metainterpretation der Interpretationskonstruktmethode selber. Denn das,
was man als Methodologe beschreibt, ist ja ein Modell, das selbst auch dem
Muster des Interpretierens und dieses Schematisierens und somit dieser
Schichtung unterliegt. Insofern muß es auch so etwa wie eine höchste Stufe der
erkenntnistheoretischen oder methodologischen Metainterpretation der Inter-
pretationsmodelle und -methoden geben.1 Man sieht den Ansatz in dem wie-
dergegebenen Diagramm dargestellt, das einigermaßen selbstexplizierend ist.
Man kann durchaus sagen, daß es durch solche Stufungen und Schichtun-
gen der Schematisierungen und Interpretationen möglich ist, traditionelle
methodologische und philosophische Probleme unter neue Sichtweise zu
stellen und zu behandeln. Beispielsweise wird das Wahrheitsproblem sich
anders stellen, wenn man einen solchen interpretationstheoretischen Ansatz
1 Das Vorgehen kann man natürlich wiederholen, das ist eine kumulative Art, man kann natürlich
auch über philosophische Metainterpretationen wieder reflektieren – man müßte eigentlich eine
Metastufe höhergehen, aber das ist also in diesem Sinne kumulativ zu verstehen.
durchführt. Man kann etwa das Wahrheitsproblem dann im Sinne einer Bezie-
hung zwischen unterschiedlichen Interpretationsschichten deuten, z. B. als die
Beziehung zwischen konventionellen IS4- und IS5-Interpretationen und eben
den grundlegenden Urinterpretationen IS1 und IS2.
Eine generelle Schlußfolgerung ist, daß das Aktivieren und Etablieren von
Schemata auf der Grundlage von plastischen neuronalen Ensembles der gleiche
oder ein gleichartiger Prozeß ist wie das Reaktivieren. Das Wiedererkennen ist
ein ähnlicher Prozeß wie das Aktivieren von Schemata und das Ausbilden. Das
Bilden, Etablieren, Stabilisieren und Reaktivieren von Schemata nenne ich
Schemainterpretieren. Schemainterpretationen sind in diesem Sinne eben
repräsentierende Aktivitäten. Es ergibt sich hier eine Brücke zwischen dem
Konstruieren im engeren Sinne, dem Konstituieren von Gegenständen und dem
Rekonstruieren, dem Wiedererkennen, dem Wiederaktivieren von Schemata,
einfach deswegen, weil das Aktivieren das erste “Einspielen” und das wie-
derholte Aktivieren von entsprechenden Schemata, das “Wiederabspielen”,
jeweils eine Aktivierung von Neuronenassemblies und somit cum grano salis
derselbe Prozeß ist. Repräsentieren, Rekonstruieren ist gebunden an das
Wiederholen von Aktivierungen, deren Ablauf schon einmal eingespielt wor-
den ist. Das Metaschematisieren, das Metainterpretieren erfolgt auch nach
diesem Muster. “Erfassen” – und d. h. Erkennen und auch Handeln (“Fassen”
i. e. S.) – ist in diesem Sinne stets schemainterpretationsgebunden oder – etwa
beim Wahrnehmen – sogar an äußere Weltfaktoren und deren Einfluß gebun-
den (“Imprägnation”). Alle diese repräsentierenden Aktivitäten sind sche-
maimprägniert im expliziten Sinne. Wir können hier von einem Grundsatz
allen repräsentierenden “Erfassens” sprechen, dem Grundsatz der Schemain-
terpretationsgebundenheit und der Schemaimprägnierung – oder schlicht:
einem Grundsatz der Schemainterpretation und Schemaimprägnation. Wir
können nicht nicht schematisieren, wir können nicht nicht schemainterpretie-
ren. In der Interpretationstheorie heißt es ja eben herkömmlich – auch in der
hermeneutischen: wir können nicht nicht interpretieren. Das ist natürlich
durchaus richtig – und das läßt sich wie angedeutet auch verallgemeinern.
III.
Ich denke übrigens, daß alles zuletzt Gesagte auch über die herkömmlichen
hermeneutischen und sprachphilosophischen Disziplinen hinaus erweitert
werden kann, daß man sogar das Wittgensteinsche Modell der Sprachspiele
erweitern kann auf Schemaspiele (vgl. Lenk 1995), daß also in gewissem Sinne
verallgemeinert werden kann über das Verbalsprachliche, über das Sprachliche
hinaus und in Kontakt und Übereinstimmung gebracht werden kann mit der
150 Hans Lenk
Neurowissenschaft. Man kann direkt in Anschluß an und Analogie Kants Satz

aus der Kritik der reinen Vernunft “Gedanken ohne Inhalt sind leer, Anschau-
ungen ohne Begriffe sind blind” (KrV B 75) formulieren: Schemainterpretatio-
nen ohne Aktivierung, ohne Interaktionen und gar Interventionen sind leer
und Interaktionen und Interventionen ohne Schemadeutungen sind blind.
Interpretation, zumal Schemainterpretation, ist stets abhängig von Interaktion
und Intervention – und umgekehrt. Das heißt, es ergibt sich eine neue Zu-
sammenführung zwischen den traditionellen methodologischen Gesichts-
punkten der Handlungstheorie und der Erkenntnistheorie. Ich denke, daß hier
auch eine Brücke zwischen einem wissenschaftlichen Begründungsansatz und
auch dem Alltagserkennen und Alltagshandeln gefunden werden kann, al-
lerdings um den Preis einer abstrakteren, höheren Stufung. Das gilt übrigens
auch für die Überbrückung zwischen den unterschiedlichen Wissenschaften,
beispielsweise den Naturwissenschaften und den fast gänzlich bzw. teilweise
hermeneutischen Geistes- und Sozialwissenschaften. Es gibt im methodologi-
schen Zusammenhang eine Einheit der Erkenntnisformen auf höherer Stufe,
die die traditionelle methodologische Trennung zwischen den Disziplinen zu
überbrücken gestattet. (Der methodologische Separatismus bleibt (zu) vorder-
gründig.)
Das hier Entwickelte ist ein recht genereller, abstrakter, aber ein pragmati-
scher und umfassend anwendbarer, interdisziplinärer, ja, in gewissem Sinne
supradisziplinärer methodologischer, ja: metamethodologischer, Ansatz, der
die Theorie-, Begriffs-, Hypothesenbildungen der unterschiedlichen Diszipli-
nen übergreift.
Es gibt natürlich gewisse Unterschiede und Unvereinbarkeiten unterhalb
der abstrakten, höherstufigen Einheitsebene. Das will ich nicht bestreiten.
Beispielsweise konzentrieren sich ja die Geisteswissenschaften auf z. T. inter-
pretationsproduzierte, fiktive Gegenstände – und die Sozialwissenschaften
auch. Die Institutionen, der Staat oder was immer sind ja keine Dinge, sondern
Fiktionen, die nur dadurch sekundäre soziale Existenz erlangen, die aber
natürlich sehr wirkmächtig ist, dadurch, daß viele Menschen an sie glauben,
bzw. auch durch Regeln zu dieser fiktiven Existenz dressiert werden. Der
übergreifende gemeinsame Gesichtspunkt ist, daß Wissenschaft immer Men-
schenwerk ist. Man sieht, daß die Einheit der Wissenschaften auf einer hö-
heren methodologischen Abstraktionsebene unter diesem Primat eines inter-
pretationskonstruktionistischen Ansatzes modellhaft erkannt und analysiert
werden kann. Insofern läßt sich sagen, daß man im methodologischen Schema-
Interpretationismus eine höherstufige Brücke zwischen den zwei angeblich
völlig getrennten “Wissenschaftskulturen” wie zwischen Erkennen und Han-
deln wiederherzustellen vermag. Das ist die Hauptbotschaft und man kann das
natürlich auch (vgl. Lenk 1993, 1993a, Kap. 44) selbst an der Geschichte der
Hermeneutik und der hermeneutischen geisteswissenschaftlichen Disziplinen
exemplifizieren. Traditionell ist ja schon in der Geschichte der Hermeneutik
ein Perspektivismus hervorgehoben worden, z. B. bereits 1742 von Chladeni-
us, der die Erkenntnis nach Auszeichnung, Auswahl, Synthese, Komparation
immer als von einem Standpunkt aus – er spricht von einem “Sehepunkt”
(1969, S. 187) – auffassen wollte. Er sprach auch direkt von “Perspektiven”.
Und auf diesen “Sehepunkt” oder perspektivischen Ansatz ist jeder Interpret
in den Geisteswissenschaften in den Deutungswissenschaften, zumal in der
Exegese, angewiesen. Dabei war auch schon Chladenius klar, daß dieses Ein-
nehmen eines “Sehepunktes” und diese Interpretation unter diesem Gesichts-
punkt konstruktiv ist (ebd. z. B. S. 518), eine interpretatorische Tätigkeit, eine
Art von Kunst. Dieselbe Einsicht spielt dann natürlich bei Schleiermacher eine
große Rolle, der ja ausdrücklich von einer schöpferischen Synthese spricht,
von einem Begriff als einem intellektuellen Schema – in Anlehnung natürlich
an Kant – , durch das ein Subjekt sich auf seinen Gegenstand beziehen kann
und durch das auch bereits ein Gedanke als Ergebnis einer synthetischen und
symbolisierenden Leistung dargestellt werden kann. Diese konstruktiv her-
stellende, poietische oder gegenstandsformierende konstitutive Funktion der
Sprache ist also bei Schleiermacher in gewissem Sinne schon erkannt; er spricht
auch vom “Schematismus” (natürlich wiederum im Anschluß an Kant) und
von einer “Denkgemeinschaft” (Schleiermacher 1977, S. 443 ff., s. a. S. 29)
innerhalb eines Sprachkreises, die zu solchen Formierungen führt. Bei Dilthey
wird das Konstruktionselement noch deutlicher; er spricht von der “Nach-
konstruktion der gegebenen Rede” (Schleiermacher 1974, S. 83) oder von der
Nachbildung des ganzen Prozesses, durch den ein Werk “in Regelform”
entsteht. Er nimmt Schleiermachers Satz auf: “Ich verstehe nichts was ich nicht
als notwendig einsehe und konstruieren kann.” (Schleiermacher 1974, S. 31):
ein Satz, der übrigens von Vico und Hobbes schon vorweggenommen worden
ist. Dilthey (WW VII, S. 220) erweitert das, indem er sagt: “So entsteht der
Sinn, indem das Unbestimmte durch die Konstruktion bestimmt wird.” Er ist
sich also durchaus im klaren, daß die Konstitution konstruktiv ist, daß eine
Interpretation eine Tätigkeit des handelnden Subjekts darstellt; und er kriti-
siert m. E. völlig zu Recht an Kant, daß Kant das nur als Problem der reinen
Erkenntnis gesehen habe, als Kategorisierung im Sinne der Anwendung von
reinen Verstandesformen. Stattdessen meint er, völlig richtig, müßte man auch
die Handlungsregeln und die Handlungsformen, die Lebensregeln und die
Lebensformen zur fundamentalen Konstitution der entsprechenden Welt-
orientierung hinzunehmen. Und das entspricht genau dem Vereinen von
152 Hans Lenk
Erkenntnis und Handeln, wie es auch in der neueren Philosophie eine Rolle
spielt, nicht nur seit Peirce im Pragmatismus, sondern natürlich auch bei den
Ideen des späten Husserl, in dessen Begriff der “Lebenswelt”, bei Wittgenstein
in Gestalt der “Lebensformen” und im Sinne seiner Sprachspielkonzeption, die
ja eine Art von handlungsbasierter Fundierung eben auch der sprachlichen
Bedeutungen vornimmt. Die zentrale Idee des späten Wittgenstein ist ja, daß
man Bedeutungen zurückführt auf Handlungsgebräuche, auf eingeschliffene,
musterhafte, regelhafte Handlungsformen. Pragmatisierung der Semiotik und
Semantik ist das Stichwort. (Das kann hier nicht näher behandelt werden; vgl.
Lenk, 1998, Kap. 7.)
Wichtig im vorliegenden Zusammenhang ist auch jedenfalls, daß die Funk-
tion des verstehenden Erschließens von Gegenständen grundsätzlich und
generell – und zwar schon vor aller geisteswissenschaftlichen Spezialmethodik
– konstruktiv, entwerfend, schematisierend, interpretatorisch verfahrend ist
oder implizit so geprägt, jedenfalls in vielerlei Sinn aktiv(istisch) ist. Das
Verstehen ist also ein erschließendes Konstituieren und Rekonstruieren, wie es
z. B. auch Gadamer (1960, 1986) etwa am Beispiel der Konstitution eines
Kunstwerks bzw. eines Spiels durch deren Auffassung als Realisierung von
bestimmten jeweils geregelten Vollzugsmustern exemplifiziert.
IV.
Wir haben gesehen, daß der methodologische Ansatz der Interpretationskon-

strukte, des methodologischen Schemainterpretationismus, uns ein Werkzeug
an die Hand gibt, die interdisziplinären Beziehungen und Arbeitsweise der
symbolisch repräsentierenden Geisteswissenschaften unter einem einheitli-
chen, wenn auch höherstufigen Gesichtspunkt zu begreifen und eine neue,
aber abstrakte, methodologische Einheit der Wissenschaften zu vertreten –
und zudem Stukturierungsweisen (Schematisierungen, Schemainterpretation
und -imprägnation) des Erkennens und des Handelns einzubeziehen. Diese
Art von doppelter Supra- und Interdisziplinarität ist natürlich für die Geistes-
wissenschaften von höchster Relevanz.
Eingangs dieses Beitrages wurde kritisch-fragend das Problem der Zwei-
Kulturen-Trennung zwischen naturwissenschaftlichen und geisteswissen-
schaftlichen Ansätzen aufgeworfen und die expositorische Frage nach Über-
brückungsmöglichkeiten dieser inzwischen überholt erscheinenden Kluft
gestellt. Eine entsprechende Kluft scheint sich auch – vielleicht noch drasti-
scher als zwischen naturwissenschaftlicher und literarisch-geisteswissenschaft-
licher Intelligenz – zwischen der wissenschaftlich-technischen Theorien-
bildung und -anwendung sowie der hochkomplizierten computergestützten
Informationsverarbeitung einerseits und dem Alltagsverstehen bzw. Umgehen

mit umgangssprachlichen Konzepten andererseits aufzutun. Die Frage war, ob
die Erkenntnistheorie etwas zur Überwindung dieser Kulturentrennung
beitragen kann. Hierzu und zur künftigen Entwicklung der Epistemologie
unter diesem umgreifenden Gesichtspunkt ist noch in einem resümierenden
Ausblick Stellung zu nehmen.
Entgegen dem ersten Anschein einer Absolutheit und Unüberbrückbarkeit
der genannten “Kulturen”-Spaltung hat sich im Verlauf der differenzierten
Auseinandersetzung mit bisherigen Interpretationstheorien und interpretatio-
nistischen Ansätzen, mit wissenschaftstheoretischen Konzepten der Natur-
und Sozialwissenschaften sowie mit erkenntnistheoretischen Überlegungen
zur Alltagserkenntnis ergeben, daß es eine abstraktere übergeordnete
erkenntnistheoretische Ebene gibt, auf der diese so unüberbrückbar erschei-
nende Spaltung doch überwunden werden kann. Die “Brücke” dabei bildet der
Begriff der konstruktiven Interpretation bzw. der Schemainterpretationen und
der interpretatorisch-schematisierenden Aktivitäten sowie des Handelns und
Erkennens mit Symbolen und symbolanalogen inneren Repräsentationen. Das
Erkennen und Handeln sowohl im Alltag als auch in den Wissenschaften und
den geisteswissenschaftlichen wie auch den philosophischen Disziplinen ist
symbolvermittelt. Ernst Cassirers Einsicht (1944, dt. 1990), daß der Mensch
zwischen die Welt und sich eine “symbolische Zwischenwelt” spannt, die ihm
ein “symbolisches Universum”, seine symbolische Welt bedeutet, ist hier
einschlägig. Der Mensch als das “symbolische Wesen” ist auf die Ausbildung
eines “Symbolsystems”, “Symbolnetzes” angewiesen, das ihm erst die han-
delnden und erkennenden Zugänge zur Welt bzw. sogar deren Konstitution
als Gegenstandswelt ermöglicht und in differenzierter Weise zu strukturieren
erlaubt. Symbolverwendung und symbolische Darstellung sind nun für die
unterschiedlichen Umgangsweisen und Darstellungen sowie Handlungen auf
beiden Seiten der Kulturentrennung charakteristisch. Hier ist ein übergreifen-
der Gesichtspunkt gefunden, der zur Überbrückung der Spaltung auf höherer,
abstrakterer erkenntnisphilosophischer und methodologischer Ebene Ansatz-
punkte, sozusagen die Brückenköpfe für den Überbrückungsversuch, bietet.
Denn alle zentralen Begriffe der Erkenntnis und der Handlungsanalyse sowohl
im Alltag als auch in der Wissenschaft und in den geisteswissenschaftlich-
philosophischen Disziplinen sind auf Symbolbildung, -verwendung und -
deutung angewiesen, beruhen auf interpretatorisch-schematisierenden Aktivi-
täten, auf der Anwendung von teils evolutionär vorgegebenen und unbewuß-
ten, teils interaktiv mit der Umwelt entwickelten, teils konventionell aufgebau-
ten bzw. im engeren sozio-kulturellen Sinne erlernten Schemata. Deren Bil-
dung, Ausdifferenzierung, Weiterentwicklung und Anwendung ist in diesem
154 Hans Lenk
Beitrag als Interpretieren im weiteren Sinne aufgefaßt worden: genauer als

Schemainterpretieren. (Das traditionelle, von der Hermeneutik untersuchte
Verstehen von Texten wurde demgegenüber als Spezialfall, nämlich als Unter-
kategorie des Textinterpretierens verstanden.) Es zeigte sich, daß alle Arten der
Darstellung und des Zugriffs zur Welt, zum erkennenden und handelnden
Subjekt selbst und auch zur Situation der Person und des Menschen in seiner
Handlungs- und Weltsituation (Lebenswelt) zutiefst interpretatorisch geprägt,
von Schemainterpretationen strukturiert und durchdrungen, hinsichtlich ihrer
Darstellung und Erfaßbarkeit generell unvermeidbar interpretationsimprä-
gniert sind. Der Grundsatz des methodologischen Interpretationismus (Inter-
pretationskonstruktionismus), daß alles Erkennen und Handeln schemainter-
pretationsgebunden bzw. -imprägniert ist und wir nur in zutiefst interpreta-
tionsabhängiger Weise überhaupt etwas erfassen, konzipieren, meinen, ordnen
und komponentendifferenzierend (be)handeln können, steht jenseits aller
Bezweifelbarkeit. Er ist der Fundamentalsatz der methodologisch orientierten
Erkenntnistheorie der Interpretationen und Theoriebildungen. Selbst ein
irgendwie differenzierendes, also Komponenten herausstellendes und unter-
scheidendes, auf Erfassung hin ausgelegtes bzw. Einzelerfassung ermöglichen-
des Wahrnehmen ist schon durch eine wahrnehmungsbahnenspezifische, aber
erregungsleitungunspezifische Schemaanwendung in Gestalt einer Zerlegung
der ursprünglichen Reizinformation, einer Verarbeitung und späteren
Wiederzusammensetzung (cortikalen Synthese) gekennzeichnet. Schemainter-
pretation waltet hier überall, wie auch die neurobiologische neuere Gehirn-
forschung zweifelsfrei bestätigt hat, so wenig sie auch noch über die
Integrations- und Syntheseprozesse der höheren Stufen empirisch gesichert
aussagen kann.
Darüber hinaus sind zentrale Begriffe des symbolischen Erfassens natürlich
selber Teil eines interpretatorischen Ansatzes: Nicht nur ist das Modell des
Interpretationskonstruktionismus selbst ein erkenntnistheoretisches Inter-
pretationskonstrukt höherer Stufe, sondern auch die Grundbegriffe dieser
Methodologie und ihres Verständnisses (wie auch des Alltagsverständnisses
selbst) wie z. B. “Bedeutung”, “Information” usw. sind ihrerseits Interpreta-
tionskonstrukte – wie auch praktisch alle allgemeineren, unserer Welt- und
Selbststrukturierung zugrunde liegenden Alltagskonzepte: Selbst die Grund-
begriffe von “Ich” und “Welt”, der Unterscheidung von Subjekt und Objekt,
von Erkennen und Handeln, von Form oder Struktur und Gehalt sind
erkenntnistheoretisch-methodologische Begriffe von interpretativem Charak-
ter.
Zunächst ist der Interpretationskonstruktionismus als methodologischer
Ansatz konzipiert und entwickelt worden. Er kann jedoch auch in der Traditi-
on der traditionellen Erkenntnistheorie als ein quasi kantischer transzendenta-

ler Interpretationismus aufgefaßt und ausgebaut werden – gleichsam zu einer
Erkenntnistheorie des kulturellen und symbolischen Wesens, das der Mensch
ist.
Darüber hinaus – und auch über Cassirer hinaus weiterschreitend – müßte
dieser Ansatz zu einer Anthropologie des interpretierenden Wesens – genauer:
des meta-interpretierenden Wesens – ausgearbeitet werden. Der Mensch ist
nämlich nicht spezifisch genug durch die Charakterisierung als das symbole-
verwendende und -interpretierende Wesen gekennzeichnet, sondern gegen-
über den durchaus vielfältige Symbole beim Wahrnehmen, Erkennen verwen-
denden und auch in primitiven, etwa in Labyrinthsituationen und Entschei-
dungsdilemmata, “logisch” schließenden höheren Säugetieren, zumal Prima-
ten, ist der Mensch dadurch ausgezeichnet, daß er sein interpretierendes Er-
kennen und Handeln seinerseits wiederum zum Gegenstand einer höher-
stufigen Interpretation machen kann. Er unterscheidet, differenziert und
interpretiert nicht nur in einer spezifischen Schicht der Interpretationen – etwa
durch Konzeptualisierung unterschiedlicher Gegenstandsklassen, Eigen-
schaften, Relationen usw. – , sondern er kann das Interpretieren selbst wieder
zum Gegenstand des Interpretierens machen – in prinzipiell unabgeschlosse-
nen Übereinanderschichtungen von Interpretationsprozessen. Der Mensch ist
also ein in spezifischerem Sinne symbolisches und handelndes Wesen, nämlich
jenes, das seine Symbolverwendungen wiederum interpretierend zum Gegen-
stand machen kann, in Metastufen übersteigen, abstrahierend darstellen und
manipulieren kann: Er ist das Metastufenwesen, das metaschematisierende und
superinterpretierende Wesen, das symbolisch auch auf höheren Metastufen des
Erkennens und Handelns Symbole über Symbole, Interpretationen über
Interpretationen setzen, Interpretationskonstrukte von Interpretationskon-
strukten bilden kann. Es ist diese Möglichkeit der Emanzipation von den
Symbolverwendungen auf der Objektstufe bzw. in der Objektsprache, die den
Menschen als das metainterpretierende Wesen auszeichnet. (Die Ausarbeitung
einer derartigen Anthropologie des metainterpretierenden Wesens muß einer
weiteren Arbeit vorbehalten bleiben.)
Erkenntnistheoretisch gesprochen ist es jedenfalls deutlich, daß mit dem
Konzept der Schemainterpretation und der interpretatorisch-schematisieren-
den Aktivitäten ein umfassender Ansatz gefunden ist, der sowohl die hypo-
thetischen Theoriebildungen und -verwendungen der Naturwissenschaftler als
auch die Deutungs- und Verstehenskonzepte der Geisteswissenschaftler
(einschließlich der Philosophen und Methodologen selbst) wie auch der All-
tagskonzeptualisierungen umfaßt. Natürlich können damit differentielle
Unterscheidungen und Entgegensetzungen, ja, Unvereinbarkeiten im Sinne
156 Hans Lenk
einer jeweils wissenschafts- oder disziplinenspezifischen Erkenntnis- oder

Wissenschaftstheorie nicht geleugnet werden. (Diese zu analysieren wäre die
Aufgabe der speziellen Wissenschaftstheorien bzw. Hermeneutiken.) Insofern
als generell Arten des Interpretierens beim Erfassen und Verstehen von Welt-
und Selbst-Strukturen und -Verhältnissen unabdingbar sind, hat die traditio-
nelle Hermeneutik des Verstehens durchaus einen methodologisch richtigen
und unverzichtbaren Ansatz gewählt (wenigstens was die Methodologie der
geisteswissenschaftlichen Disziplinen im weitesten Sinne angeht, aber auch für
die praktische Verwendung von Hypothesen und Theorien sowie Ordnungs-
konzepten generell in den Naturwissenschaften – etwa der Auffassung her-
meutischer Methoden als “technologischer Kunstlehren” (Albert)). Darüber
hinaus hat die philosophische Hermeneutik – andeutungsweise schon bei
Dilthey beginnend, zumal aber bei Heidegger und Gadamer – die hermeneuti-
sche Geprägtheit der Welt- und Selbst-Vorentwürfe, der Konstitution und
ihrer methodischen Konzepte und methodologischen Vorbedingungen, durch-
aus gesehen, indem in dem intentionalen Gerichtetsein von etwas auf etwas
und dem Verstehen von etwas als etwas gewisse entwerfende Konstruktivitäts-
momente erkannt werden, die als Andeutungen eines methodologischen
Konstruktionsinterpretationismus aufgefaßt werden können. Jedoch verblieb
das hermeneutische Modell noch allzu sehr im Banne des traditionellen Text-
interpretationismus und des Leseparadigmas: Die Welt sollte sozusagen als
Text interpretiert werden, ja, Handlungen selbst nur als Texte aufgefaßt wer-
den. Die Universalhermeneutik blieb dem eingeengten Begriff der Textinter-
pretation verhaftet und konnte nur fallweise oder in Andeutungen zu einem
interpretationistischen Konstitutionalismus hinlenken. Insbesondere vermoch-
te es die traditionelle Hermeneutik wie auch die Universalhermeneutik nicht,
auch auf neuronaler und biologischer Grundlage die schematisierenden Aktivi-
täten, die die Vehikel aller symbolisch-interpretatorischen Tätigkeit des Orga-
nismus sind, einzubeziehen.
Demgegenüber hat die moderne Hirnforschung zweifelsfrei die Fruchtbar-
keit der Schemabildungen und Schemaverwendungen sowie der Primärinter-
pretationen und ihnen nahestehender Deutungen als Schemaaktivierungen und
-stabilisierungen zu erhärten vermocht. Unter dem Gesichtspunkt eines auf die
Schematisierungen (Schemabildungen, -aktivierungen und bewußte -verwen-
dungen bzw. -differenzierungen) ausgeweiteten Interpretationsbegriffs – also
eines Interpretationskonzepts im weiteren Sinne – gelingt es, die erkenntnis-
und handlungstragenden biologischen Strukturierungen mit jenen der symbo-
lischen und kulturellen unter einem begrifflich-theoretischen Dach zu vereinen
und in gleicher Weise auf das wissenschaftliche wie auch auf das alltägliche
Erkennen, Handeln und Konstituieren zu beziehen. Unter dem abstrakteren
methodologischen oder transzendentalen Gesichtspunkt des Interpretations-

konstruktionismus läßt sich also eine übergreifende, verbindende (wenn auch
mit einer gewissen Formalität und Abstraktheit erkaufte) Wiedervereinigung
der Erkenntnis- und Handlungsdisziplinen unter dem Gesichtspunkt einer
interpretationistisch-symbolistischen Anthropologie erreichen. Die Einheit
der Erkenntnis ist auf höherer Metastufe, jener bestimmter abstrakt-inter-
pretationistischer Formen, Regeln, Methoden, Erfordernisse und Ergebnisse
der Interpretationen, wieder hergestellt. Auch dies ist für die Geistes- und
Literaturwissenschaften wie für die Sozialwissenschaften von höchster Bedeut-
samkeit!
Literatur
Bunge, M.: Scientific Research : Bd. I. Berlin u. a. 1967.
Cassirer,E.: Versuch über den Menschen. Frankfurt a. M. 1990 (Orig. 1944).
Chladenius, J.N.: Einleitung zur richtigen Auslegung vernünftiger Reden und Schriften. Mit einer
Einleitung von L. Geldsetzer. Düsseldorf 1969 (Nachdruck der Ausgabe Leipzig 1742).
Dilthey, W.: Gesammelte Schriften. Stuttgart 1914 ff.
Gadamer, H.G.: Wahrheit und Methode. Berlin 19865.
Heckhausen, H.: ‘Interdisziplinäre Forschung’ zwischen Intra-, Multi- und Chimaren-Disziplina-
rität, in: Jürgen Kocka (Hg.) : Interdisziplinarität. Frankfurt a. M. 1987, S. 129-151.
Kant, I.: Kritik der reinen Vernunft. Auflage B 1787. Hamburg 1956.
Karli, P.: “L’experience du Centre de recherches transdisciplinaires sur les sciences et les techni-
ques (C. R. T. S. T.) de l’Université Louis Pasteur de Strasboug. Erfahrungen des Zentrums für
transdisziplinäre Forschung über Wissenschaft und Technik”, Universität Louis Pasteur, Straß-
burg, in: Arber, W. (Hg.): Inter- und Transdisciplinarität: warum – wie? Inter- et Transdisciplina-
rité pourquoi – comment? Bern 1993, S. 159-173.
Kocka, J. (Hg.): Interdisziplinarität. Frankfurt a. M. 1987.
Kreuzer, H.: Literarische und naturwissenschaftliche Intelligenz. Stuttgart 1969 (Tübingen,
München 19872).
Krüger, L.: “Einheit der Welt – Vielheit der Wissenschaft”, in: Kocka, J. (Hg.): Interdisziplinarität.
Frankfurt a. M. 1987, S. 106-125.
Lenk, H.: Werte als Interpretationskonstrukte, in: Hans Lenk : Zwischen Sozialpsychologie und
Sozialphilosophie. Frankfurt a. M. : Suhrkamp 1987, S. 227-237.
ders.: “Zu einem methodologischen Interpretationskonstruktivismus”, in: Zeitschrift für all-
gemeine Wissenschaftstheorie (Journal for general Philosophy of Science) 22 (1991), S. 283-302.
ders.: Interpretationskonstrukte : Zur Kritik der interpretatorischen Vernunft. Frankfurt a. M.
1993.
ders.: “Wissenschaftskulturentrennung und methodologische Wissenschaftseinheit im Blickwinkel
des Interpretationismus”, in: Herbert Mainusch/Richard Toellner (Hg.): Einheit der Wissenschaft:
Wider die Trennung von Natur- und Geist, Kunst und Wissenschaft. Opladen 1993, S. 195-225.
ders. (unter Mitarbeit von E. Kaleri): Philosophie und Interpretation. Vorlesungen zur Entwick-
lung konzeptionistischer Interpretationsansätze. Frankfurt a. M. 1993(a).
ders.: Schemaspiele : Über Schemainterpretationen und Interpretationskonstrukte. Frankfurt a. M.
1995.
158 Hans Lenk
ders.: Interpretation und Realität. Frankfurt a. M. : Suhrkamp 1995(b).

ders.: Interpretation – Intervention – Interaktion. Buch-Manuskript i. Dr.
von der Malsburg, Ch.: “Am I Thinking Assemblies?”, in:Palm, G./. Aertson, A. (Hg.): Brain
Theory. Heidelberg, New York. 1986, S. 161-176.
Roth, G.: “Kognition – die Entstehung von Bedeutung im Gehirn”, in: Krohn, W./Küppers, G.
(Hg.): Emergenz: Die Entstehung von Ordnung, Organisation und Bedeutung. Frankfurt a. M.
1992, S. 104-133.
D. E. Schleiermacher, F.: Hermeneutik. Herausgegeben und eingeleitet von H. Kimmerle. Heidel-
berg 19742.
ders.: Hermeneutik und Kritik. Hg. Frank, M. Frankfurt a. M. 1977.
Singer, W. (Hg. und Einführung): Gehirn und Kognition. Heidelberg 1990.
Snow, C. P.: Die zwei Kulturen. Stuttgart 1967.
Weingart, P.: “Das Dilemma: Die Organisation von Interdisziplinarität”, in: Wirtschaft und
Wissenschaft 1974, 3, S. 22-28.
ELLERY EELLS
Causal Decision Theory
Philosophers concerned with the nature of rational decision making have

distinguished two paradigms for rational deliberation. According to evidential
decision theory, one should, when faced with alternative courses of action,
perform an act whose performance would constitute the best possible evidence
that one will get the good outcomes and not get the bad outcomes. On this
paradigm, therefore, deliberation consists in trying to figure out which of the
available actions is such that performing it should cause one’s beliefs to be
altered in such a way that the good outcomes will be as subjectively probable
as possible and the bad outcomes as subjectively improbable as possible, given
constraints imposed by one’s prior beliefs. According to causal decision theo-
ry, on the other hand, one should be explicitly concerned with the causal
relations between the available acts and the possible outcomes. On this para-
digm, an act is evaluated according to the degree to which it can be expected to
cause (rather than merely provide evidence for) the good outcomes and pre-
vent (rather than merely provide evidence against) the bad outcomes. I will
give a precise and formal version of each of these two kinds of theories below.
Of course, one’s believing that an item A is a cause of an item B is not the
only reason why A might be evidence for B: A could be evidence in favor of B
for a person who believed that B causes A, or that there is a common cause of
both A and B. In decision problems in which the only reason why an act
would be, for the relevant agent, evidence for an outcome is that the agent
believes that the act is a cause of the outcome, evidential and causal decision
theory will agree about the appropriate course of action. Thus, many phi-
losophers concerned with the nature of rational decision have concentrated on
a class of decision problems, called ‘Newcomb problems’, in which acts provi-
de evidence for outcomes in other ways. For some such problems, it seems that
these two paradigms yield different prescriptions for action, the causal para-
digm giving the correct prescriptions and the evidential paradigm the wrong
ones. Some examples of this kind will be discussed below.
It has been argued – for example, by Kyburg (1980), Eells (1980, 1982,
1984b), and Jeffrey (1981, 1983) – however, that (roughly), if a decision maker
appropriately „monitors” the relevant aspects of his deliberation (e.g., he
knows what his beliefs and desires are), then evidential decision theory will
160 Ellery Eells
deliver the correct prescriptions, the same as those given by causal decision
theory. These arguments have come to be known as „metatickle” arguments;
a metatickle can be thought of, basically, as a piece of information about one’s
deliberation. I’ll describe three such approaches below, and note some of their
potential shortcomings. Given these shortcomings, and the intricacy of meta-
tickle analysis, it would seem that causal decision theory would have a great
advantage over evidential decision theory if, unlike evidential decision theory,
metatickle analysis were not necessary for causal decision theory’s delivering
the appropriate prescriptions. I’ll argue, however, that there are decision
problems in which even causal decision theory will give the wrong answers, if
metatickles are not taken account of. This will indicate that many of the short-
comings of metatickle analysis as employed in defense of evidential decision
theory are also relevant to the adequacy of causal decision theory. And it
suggests that it is just as much in the causal decision theorist’s interest as it is in
the interest of the evidential decision theorist to try to develop an adequate
account of the impact of metatickles on deliberation.1
1. The Two Theories and the Prima Facie Conflict
Evidential and causal decision theories give different ways of calculating the
value of an act in terms of the decision marker’s subjective probability as-
signment (symbolized here by ‘Pr’) and desirability assignment for the possi-
ble outcomes of his decision (symbolized here by ‘Des’). Evidential decision
theroy can be summarized as follows:
Value (A) = SumiPR (Oi/A)Des(Oi),
where A is any act and the outcomes Oi include (positively or negatively)
everything the agent cares about in his decision problem.2 According to the
subjective interpretation of probability, the sign and magnitude of the diffe-
rence between Pr(Oi/A and Pr(Oi) is supposed to reflect the degree to which
1 The thesis that metatickles actually have a legitimate and important role in deliberation (both
evidential and causal) is, incidentally, relevant to recent criticism by Frank Jackson and Robert
Pargetter (1983, 1985) of the „tickle” and „metatickle” defenses of evidential decision theory. See
Eells (1985) for a defense of evidential (and causal) decision theory against this criticism.
2 With the outcomes so understood, evidential and causal decision theory will agree about the
values of the outcomes. Hence the occurrence of ‘Des’ in both formulas for expected utility. If the
outcomes are not so specific, we should replace the function Des with the function Value, about
whose values the two theories may differ. I think the difference between the evidential and causal
paradigms is brought more clearly into focus when the outcomes are understood to be maximally
specific, so that the difference between the two paradigms can then be characterized in terms of
different ways of calculating the expected utility values of less specific things (such as acts) in
terms of the functions Des (whose domain is maximally specific outcomes) and Pr, about whose
values the two paradigms can find no room to disagree.
A would be evidence in favor of or against Oi. So the value of an act is high or

low to the degree to which the act would be, for the agent, evidence for the
good outcomes and against the bad one. The evidential value of an act A is also
sometimes called its ‘conditional expected utility’, CEU (A).1
Since an act can be evidence for or against an outcome it doesn’t cause or
prevent, and since causal decision theory wants to pay attention only to the
degrees to which acts cause outcomes, the probabilities of certain items must
enter into the calculation of causal expectation unconditional on the act. Here
is one version of causal decision theory.2 Let Kj’s be propositions that say
which of the Oi’s are outside the influence of which acts, which Oi ‘s are
within the influence of which acts, and what the degrees are of the relevant
acts’ influence on the relevant Oi’s. Another way of saying what the Kj’s say is
this: each Kj specifies, for each available act, one set of causal propensities for
the outcomes given that the act is performed.3 Then, according to causal
decision theory,
Value (A) = Sumi jPr(Kj)Pr(Oi/Kj&Des(Oi&Kj).
(Given the above understanding of the Oi’s, Des(Oi&Kj) = Des(Oi).) This is
the basic form of causal expectation. Different current versions of causal
decision theory differ only with respect to the analysis of causal propensity.
The most famous example of a decision problem which is supposed to
exhibit a conflict between evidential and causal decision theories is New-
comb’s problem. You are given a choice between taking just the opaque box
(act A1) and taking both the opaque and the transparent box (act2), where you
can see that the transparent box contains $ 1,000 and you know that the opa-
que box contains either $ 1,000,000 or nothing, according to whether a fan-
tastically accurate predictor predicted you would take just the one or you
would take them both. It is assumed that your beliefs about the predictor are
such that the probability of his predicting act Ai conditionally on your doing
Ai is high, for each i. If these conditional probabilities are each equal to 0.9,
and assuming monetary desirabilities, the evidential expectations of A1 and A2
are $ 900,000 and $ 101,000, respectively, so that evidential decision theory
prescribes taking just the opaque box. By the lights of causal decision theory,
however, this act is irrational, for taking just the opaque box is only sympto-
1 See Jeffrey (1965, 1983) for the full development of the theory.
2 This version of the theory is patterned after Skyrms’ statement of it in his (1982). See also
Gibbard and Harper (1978), Skyrms (1980, and Lewis (1981).
3 Whether the acts A, the outcomes Oi, and such things as the Kj’s are considered to be „proposi-
tions” (as in Jeffrey 1965, 1983) or „states of nature” is irrelevant to the philosophical points at
issue here.
162 Ellery Eells
matic of the $ 1,000,000’s being in the opaque box, and in no way causally
affects the box’s contents. Also, no matter what’s in the opaque box, you’ll get
$ 1,000 more by taking both boxes than you get by taking just the one. And
the causal formula for expectation reflects just this. Here, our credence is
divided between two Kj’s, corresponding to whether the $ 1,000,000 is stashed
or not: Kj’s, which says that the causal propensity for receiving $ 1,000,000
given A1 and the causal propensity for receiving $ 1,001,000 given A2 are both
equal to 1, and K2,which says that the causal propensity for receiving nothing
given A1 and the causal propensity for receiving $ 1,000 given A2 are both
equal to 1. What the exact values of A1 and A2 are, according to causal decision
theory, depends on how we divide our credence between K1 and K2, but no
matter how we do this, the causal value of A2 is $ 1,000 greater than the causal
value of A1.
Whatever one’s intuitions are about which act is correct in Newcomb’s
problem, there are other, closer to earth, problems for which the prescriptions
of causal decision theory are clearly correct, but where the evidential theory
seems to give the wrong answers. They have the following form. Suppose that
a prior state C tends to cause both act A and a very bad outcome O, where A
causes a slight good, O’. The Fisher smoking hypothesis example is of this
form, where A is smoking, O is lung cancer, O’ is the relatively small pleasure
one gets from smoking and C is a genetic condition. If O and O’ are all the
agent cares about in his decision problem, then it is clear that A is the rational
act. And the causal expectation of A is greater than that of ~A by the amount
of pleasure O’ brings. But if the causal relation between C, A, O, according to
one’s beliefs, is such that A would provide strong evidence for C, and thus for
O, then it seems that evidential decision theory will prescribe abstaining. Note,
incidentally, that Newcomb’s problem is of the general form described above,
if we assume that behind every good predictor is a common cause of both the
event predicted and the prediction of the event, such as cold fronts causing
both the rain and the falling barometer on which the weather forecaster bases
his prediction of rain. Such decision problems constitute not only prima facie
conflicts between evidential and causal decision theory, but also prima facie
counterexamples to the former.
2. Three Defenses of Evidential Decision Theory
I shall now describe three „metatickle” defenses of evidential decision theory,

to the effect that theory gives the same prescriptions as the causal theory in
Newcomb problems. Each defense in turn is supposed to accommodate poten-
tial shortcomings of its predecessor, the last thus being the most promising.
Before discussing the first real defense, note that, in the Fisher smoking case
for example, if the way the bad gene caused one to smoke was by producing a
certain kind of tickle in the taste buds, and one knew this and whether or not
one had the tickle, then this knowledge would break the evidential connection
between smoking and the gene, and evidential decision theory would then give
the right answer: smoke. Both conditional on having the tickle (and thus the
gene) and conditional on lacking the tickle (and thus the gene) and conditional
on lacking the tickle (and thus also the gene), smoking only produces evidence
for the small good, namely the relatively small pleasure one gets from smo-
king: in both cases, the vastly more important question of whether or not one
has the bad gene is settled by the known presence or absence of the tickle.
However, as Brian Skyrms has pointed out (1980), p. 131), there needn’t
always be such a „tickle”, so we can’t count these observations as constituting
a general defense of the evidential theory. However, the defenses described
below all operate in a formally similar way: some analogue of the tickle – a
„metatickle” – is sought that must always be present.
The first defense I shall discuss (Eells 1980, 1982) looks to a rational agent’s
beliefs and desires, i.e., subjective probabilities and desirabilities. If the gene,
say, causes smoking in rational, expectation-maximizing agents, then it must
affect either their subjective probabilities, ther subjective desirabilities, or their
choice of expectation formula. For it is just these three things which determine
the expectation-maximizer’s act. If we assume that the agent is rational and
will rationally deliberate, then it seems to be a safe assumption that the gene
will not affect the choice of expectation formula, for it is just this that models
rational deliberation in light of one’s beliefs and desires. So the gene just
affects one’s beliefs and desires, modeled decision theoretically by one’s sub-
jective probabilities and desirabilities. But the agent must know what his
subjective probabilities and desirabilities are if he is to apply any of the compe-
ting decision theories. Assuming that the agent, knowing some decision theo-
ry, fully believes all this, then his knowledge of his probabilities and desirabili-
ties should render the act evidentially irrelevant to the presence or absence of
the gene.
In general, if C causes A and the only way it does this is by causing agents
to have probabilities and desirabilities that rationally lead to A, then, if R is a
proposition that specifies a particular set of subjective probabilieties and
desirabilities, say our agent’s, we should expect our agent (assumed to know
some decision theory) to have subjective probabilities such that Pr (A/R&C)
= Pr (A/R&R ~C). The agent knows that the act will be based on the subjecti-
ve probabilities and desirabilities specified in R, and that, beyond whatever
role C or ~C had in producing these beliefs and desires, the prior state is now
164 Ellery Eells
completely inefficacious in affecting which act is chosen. So the agent’s subjec-

tive probability of his performing A should be based just on what R indicates
to him about the chances of his deciding on A, C now being both causally and
evidentially irrelevant to the act. By symmetry of probabilistic independence,1
it follows that Pr (C/R&A) = Pr (C/R& ~A), so that, conditional on one’s
knowledge of one’s beliefs and desires, the act is evidentially irrelevant to the
important prior state, and thus also to the outcome of which the latter is a
cause.2
One may object (Eells 1980, 1982, Horwich 1983), however, that although
the agent knows what his beliefs and desires are, he does not necessarily know
just what to make of the probability of A or of C in light of them. Consider Pr
(A/R&C) and Pr (A/R& ~C) for example. Considered as objective propensi-
ties, these two probabilities should be equal; but not necessarily equal when
construed as subjective probabilities. For there are crucial features of the
beliefs and desires described by R that the agent may not be aware of, but
which he may need to be aware of in order for R to „screen off” the evidential
relation between A and C. To the agent at an early stage of deliberation who
does not yet know which act is rational in light of the beliefs and desires
described by R, R&C might constitute better evidence than R& ~C that R
describes beliefs and desires of the kind that rationally lead to A. This would
be because R&C should be better evidence than R& ~C that R was influenced
by the presence (rather than the absence) of C, where C causes beliefs and
desires of the kind that rationally lead to act A. Thus, Pr(A/R&C) may very
well be greater than Pr (A/R& ~C), and by symmetry, Pr(C/R&A) greater
than Pr (A/R&P ~A). Perhaps we may still insist, however, that if the agent is
strongly rational and experienced in rational decision making, he will ignore
what is know to be no longer causally efficacious in controlling his decision
(i.e., the presence or absence of C) and concentrate only on his beliefs and
desires (i.e., on R) in gauging the probability that he will perform the sympto-
matic act, so that we in fact should expect that Pr(A/R&C) = Pr (A/R&P ~C),
and evidential irrelevance of the act to the prior state, given R.
1 Henry Kyburg has argued (1974), pp 287-290, and 1980, p. 153) that independence (and thus
also, by implication, positive and negative correlation) is not in general symmetric for the relevant
kind of subjective probability. But see my (1984a, pp. 83-86) for arguments to the effect that
Kyburg’s prima facie counterexamples to symmetry are not genuine.
2 If the causal „fork” from C to A and O is, subjectively, a Reichenbachian „conjunctive fork”
(Reichenbach 1956, p 159), then the evidential irrelevance of A to O follows, by the laws of
probability, from the evidential irrelevance of A to C. On the other hand, the fork might be
„interactive (Salmon, e.g., 1978). For a discussion of this distinction and the case of interactive
forks in decision theory, see Eells and Sober (1984), where the metatickle defense is extended and
applied to cases of this kind in defense of the evidential theory.
In any case, it is desirable for a decision rule to be applicable to less so-

phisticated agents, and it was perhaps considerations similar to those just given
that motivated Richard Jeffrey’s „ratificationst” approach (1981, 1983). On
this approach, the metatickle is not just one’s knowledge of one’s beliefs and
desires, but rather the opting (at least hypothetically, see below) for one act or
the other. Suppose that opting for an act confers an extremely high subjective
probability on the act, but short of 1 because of the realistic acknowledgment
that one may „slip” at the time of action. This will allow for post-choice
calculation of the expected utility of each act. Now consider Newcomb’s
problem again. And suppose for now that what the predicator predicts is not,
strictly speaking, what you will do, but rather what you will choose (more on
this assumption below). Now suppose you have opted for A1. Since this
confers a very high probability on A1, and because of the way in which you
are confident in the predictor’s accuracy, this confers very high probability on
the proposition, K1, that the predicator has predicted your choice of A1.
Because the predicator predicts what you opt for and not what you do, the
high probability of K1 should be the same conditional on your actually doing
as it is conditional on your actually doing A2 – all given your choice of A1. So,
when evidential expected utility is recalculated after you have made your
choice of A1, act A2 will turn out to have the greater expected irrelevance of
the acts themselves (as opposed to the opting for them) to the prediction. Also,
if you opt for A2, then A2’s post-choice expected utility will again be greater
than that our A1, again because we will have post-choice evidential irrelevance
of the acts themselves to the prediction.
This is what it means to say that a decision to perform a given act is ratifi-
able: the act has greatest expected utility relative to the post-choice probability
function which reflects the knowledge (or at least supposition, see below) that
one has opted for that act. A choice of A1 is not ratifiable since, as we have
seen, relative to the probability function one adopts after having opted for A1
(which takes account of one’s having so opted)j, it is A2 that has the highest
evidential expected utility. A2, however, is ratifiable. The maxim, then, is:
„Make ratifiable decisions.” (Jeffrey 1983, p. 16).1 And the metatickle is the
1 As pointed out by Jeffrey, there are examples of decision problems in which there is no ratifiable
action and problems in which there is more than one ratifiable action (1983, pp. 18-19). Jeffrey
councils that „in such cases you do well to reassess your beliefs and desires before choosing.” (P.
19) For cases in which there is exactly one ratifiable act, it’s not clear to me whether the intent of
Jeffrey’s maxim was that an act’s ratifiability is both necessary and sufficient for its being rational
to opt for it. However, an example of Brian Skyrms’ (1984, pp. 85-86) strongly suggests that an
act’s (unique) ratifiability cannot always be sufficient for its being rational to opt for it. Suppose
that there are three shells before you. You can have what’s under the one you choose. If the
predictor predicts you will take what’s under the first shell he puts 10 cents under it and nothing
166 Ellery Eells
knowledge (or the supposition) that one has made such and such a choice.
Note that this defense of the evidential theory is not vulnerable tho the objec-
tion to which the previous defense, above, seemed open to. For, given that the
agent has made a choice, the agent should no longer entertain an evidential
connection between (in the notation used in the previous defense) R&C and
the proposition that A is the rational act: presumably, when one has made a
choice, one has made the choice which one has already deemed to be the
rational choice in light of one’s beliefs and desires.
There are two at least potential difficulties with this approach. The first was
pointed out by Bas von Fraassen and reported by Jeffrey (1983, p. 20), in the
framework of the prisoners’ dilemma. The analogue of this problem for
Newcomb’s problem is this: suppose that , contrary to the assumption above,
the predictor predicts, properly speaking, the act and not the choice, where the
accuracy of the predicator holds up whether or not we succeed in performing
the chosen act. In that case, we do not get evidential irrelevance of the act to
the prediction of it, conditional merely on the choice of one of the two acts –
for now the predictor is able to predict which way a „slip” between an intenti-
on and the act will go. But it seems to me that this problem is not as serious as
the might seem, for the situation which the agent must envision for this kind
of problem to have an effect on the ratifiability of A2 is one which is, to say
the least , far from ideal for the application of standards of rational decision
anyway, as explained blow.
For i = 1, 2, let Ci be the opting for Ai. The following is a simple formulati-
on of the assumption that the predictor’s accuracy holds up conditional on
each choice (with the conditional probabilities going in the direction relevant
to the calculation of evidential expectation, as explained above):
Pr(K1/A1&C1) = Pr(K1/A1&C2),
>> Pr(K1/A2&C1) = Pr(K1/A2&C2).
Concentrating just on the leftmost probability on each side of the inequality
sign, we get, by symmetry of (positive, say) probabilistic correlation, that:
Pr(A1/K1&C1) > Pr(A1/K2&C1).
under the others. If he predicts you will take what’s under the second shell, he puts $10 under it,
$100 under the third, and nothing under the first. If he predicts you will take what’s under the
third shell, he puts $20 under it, $200 under the second shell, and nothing under the first. Assu-
ming that it’s extremely subjectively probable, conditional on your taking any given act, that the
predictor has predicted that act, then taking what’s under the first shell is the unique ratifiable act.
However, it seems clear (as both straight evidential decision theory and causal decision theory
agree) that this act is the worst.
And concentrating on the rightmost probabilities on the two sides of the

inequality sign, we get:
Pr(A1/K1&C2) > Pr(A1/K2&C2).
For simplicity, let’s suppose, for i = 1,2, that there is a perfect correlation
between Ki and the prior state of agents (or of whatever) which causes both
the act Ai and the predictor’s prediction of that act. Thus, we may think of Ki
as the prior state itself in the two inequalities just above. We see then that,
according to the agent’s beliefs, no matter what he chooses to do there is still
something about each prior state Ki that is causally positive for Ai: in case the
agent (rationally) chooses the act that isn’t correlated with the prior Ki that
actually obtains, that Ki causes (probabilistically) the agent to slip in favor of
it.1 But it seems that to the extent to which the agent believes that his act is
thus out of his control, he shouldn’t think it worthwhile to rationally delibera-
te anyway. For to the extent to which the agent believes this, he should believe
that the probability is high that his act will be in agreement with the Ki that
obtains no matter what the result of his rational deliberation (i.e., his rational
choice) is. So, to the extent to which the agent should consider his problem to
be appropriate for the application of standards of rational deliberation in the
first place, it seems that he should believe that „slip” between the intention to
perform a given act and the act itself should be „random” in direction, i.e., the
direction of the slip (towards A1 or towards A2) is uncorrelated with factors
that play other roles in his deliberation.2
If this first problem for ratificationism doesn’t seem so threatening under
the assumption of an ideal setting for the application of standards of rational
decision, the second problem isn’t so easily disposed of – if ratificationism is
taken to be the doctrine that ratifiability should be used as the criterion of
choice. Jeffrey explains that „we are to judge the ratifiability of each act by
estimating the desirabilities of all acts form the point of view the agent would
have upon choosing the one act whose ratifiability is in question.” (1983, p.
19). The problem is: How is the agent supposed to obtain that point of view?
Supposing that an act A is the single ratifiable act, there is no guarantee that
the agent will ever, at any point in his deliberation, actually choose it so that he
will then be in a position to assess its ratifiability. But perhaps the agent can
simply imagine that he chooses it and then assess its ratifiability from this
hypothetical perspective. But what exactly is the agent supposed to imagine
1 This, incidentally, is in accord with the „contextual unanimity” model of probabilistic causality,
as discussed by, for example, Cartwright (1979), Skyrms (1980), and Eells and Sober (1983).
2 For more details on this and another kind of „slippage”, and rational agents’ attitude about the
possibility of so slipping, see Eells and Sober (1984).
168 Ellery Eells
here? Is he to imagine that he has chosen the act on the basis of its having
maximal expected utility, or should he perhaps imagine that he has chosen it
on the basis of its ratifiability? Suppose he’s to imagine the first of these two
things, and that, on this supposition, the act has maximal expected utility, so
that his (hypothetical) decision to perform the act is ratifiable. This alone
would hardly seem to justify the act, for the act is ratifiable here only in the
sense that it has maximal expected utility when calculated relative to the sup-
position that it has maximal expected utility: i.e., were it to attain maximal
expected utility is recalculated relative to the knowledge that it has attained
maximality. This would only seem to justify the act if, at some point in the
agent’s deliberation, the act actually attained maximal expected utility, for it
could only then actually retain its maximality, having attained it, and perhaps
thereby be justified. But there is no guarantee that the act will ever initially
attain maximality.1
But perhaps the agent is supposed to imagine that he chooses the act on the
basis of its ratifiability. This is a natural suggestion if we suppose that the
agent is, in the end, going to use ratifiability as his criterion of decision, and
that he knows this. So now the question is whether the act is maximal on the
supposition that it is chosen on the basis of its ratifiability. The question
presents a vicious circle. For it is equivalent to: Is the act maximal on the
supposition that it is chosen on the basis of its maximality on the supposition
that it is chosen on the basis of its ratifiability, i.e., on the basis of its maximali-
ty on the supposition that it is chosen on the basis of its ratifiability, (and so
on)? Indeed, on this reading, the definition of ratifiability is circular.
It thus appears that, on any adequate notion of ratifiability, one cannot
really assess the ratifiability of a choice until and unless the choice has been
made, at least tentatively. But, again, there is nothing in ratifiability analysis
that guarantees that one will actually finally make a ratifiable choice. Indeed,
for all that has been said, it might require causal decision theory to deliver a
choice which, after it has been made, can (only) then be seen to be ratifiable.
What would seem to be necessary and sufficient for the rationality of an act is:
ratifiability together with its prior, independently principled choice.
What evidential decision theory needs is a plausible analysis of the pre-
choice dynamics of evidential deliberation; and if evidential decision theory is
to prove adequate, such an analysis must show that evidential deliberation will
eventually land the decision maker on an act that will turn out to be ratifiable.
I shall next turn to some suggestions along these lines (see also Eells 1984b). In
the examples we’ve looked at so far, the initial prescriptions of causal decision
1 Skyrms’ shell game (see Note 9 above) provides an example of a ratifiable, non-maximal act.
theory have all been ratifiable. The question arises of whether this is so in all
cases, or whether the initial prescriptions of causal decision theory may in
some cases be unratifiable, thus indicating that causal decision theory must
also develop an analysis of the pre-choice dynamics of deliberation which will
show that pre-choice causal deliberation will eventually land the decision
maker on a ratifiable decision. In the next section, I’ll argue, by way of an
example, that causal decision theory also stands in need of such an analysis.
Suppose our agent always wants to maximize his conditional expected
utility and he has to choose between two acts, B1 and B2, in some decision
problem. Let B0 be the act of choosing between the two in a random way –
plausibly the easiest way to decide. And let B1 be the act which, at this point in
the agent’s deliberation, maximizes his CEU, where the agent doesn’t necessa-
rily know at this point which act this is1 The agent does know, however, that
CEU(B1) CEU(B0). Since the inequality might be strict, and since the agent
wants to maximize his CEU, he should figure that it’s in his interest to calcula-
te the CEU’s of the acts. Now the agent knows which act B1 is.
Suppose the agent now comes into possession of some new information, e1.
This can change the evidential values of the available courses of action: the new
CEU of each act is equal to the old CEU of the conjunction of the act with e1,
and, where B2 denotes the act, the agent knows not which, that maximizes the
agent’s new CEU’s, the following inequality holds: CEU(B2&e1). And again,
since this inequality might, for all the agent knows, be strict, and since the
agent by hypothesis wants to maximize his CEU, he should want a recalculati-
on. In general, if, for all i less than or equal to some number n, the agent learns
ei after the i-th calculation of CEU, but before performing one of B1 and B2,
then the following inequality should hold (for any such n):
CEU(B n+1 &e1&en) CEU(Bn&e1&...&en).
Since, for all n, the agent should know that the inequality holds, and since he
desires to perform the CEU-maximal act, he should always desire a recalculati-
on of CEU after coming into possession of new information en. Assuming that
1 „Classical” Bayesian epistemology requires a rational agent to assign probability 1 to any

propositions known by him. This contradicts my assumption that our agent doesn’t know which
act is CEU-maximal, since I assume that he does know what his subjective probabilities and
desirabilities are. However, I believe that a „slightly more realistic” Bayesian epistemology is
possible (see, e.g., Hacking (1967) and Garber (1983) for two tries), one that does not require
(among other idealistic requirements of the classical theory) that the agent assign probability 1 to
(or even entertain) each logical consequence of known propositions.
170 Ellery Eells
the cost of calculation is low enough (or that the differences in utility between
the possible outcomes of his decision is great enough), he should recalculate.1
Note that the result of the very first calculation of CEU is itself new infor-
mation. If it weren’t new, the agent wouldn’t have had to calculate CEU in the
first place. Let us suppose that, in the course of the agent’s deliberation, he
obtains no new information „from the outside” – the only new information he
obtains in the course of his deliberation is the results of his various calculations
of expected utility. Thus, let, for each i, ei be the information obtained from
the i-th calculation of the CEU’s of the acts. It is thus plainly at least theoreti-
cally possible that even if the agent does not receive any new evidence „from
the outside” in the course of his deliberation, it is nevertheless possible that the
CEU-ranking of the acts should shift in the course of deliberation, for when
new evidence is absorbed (be it from the outside or not), the CEU’s of the
available acts can change.
Given the possibility of such shifts in the CEU ranking of the acts, we can
make sense of hypothetically supposing that such and such an act is CEU-
superior to the other, and vice versa: perhaps if the agent had deliberated
longer, making more calculations, or had earlier cut off deliberation, the
currently (actually) CEU-minimal act (the currently CEU-proscribed act)
would have been CEU-superior (currently CEU-prescribed). Given the intel-
ligibility of this idea, we can state a general ratificationism-like principle,
actually a coherence principle for decision rules D, as follows:
Principle of Ratification: IF (i) B1 and B2 are two of the acts among
which the agent must choose in his decision problem, (ii) at any time in
the course of deliberation, if decision rule D prescribed B1 (given of
course, the agent’s beliefs and desires) and if B1 were chosen at that time
on this basis, then post-choice, B2 would appear better than B1 relative
to decision rule D (and the agent’s beliefs and desires, of course), (iii) at
any time in the course of deliberation, if D prescribed B2 and if B2 were
chosen on this basis, then post-choice, B2 would continue to appear
better than B1 relative to rule D, THEN rule D must ultimately rank B2
above B1 for the agent; otherwise the rule is (at least for this reason)
inadequate for the agent in his decision problem.2
1 I’m not assuming, or arguing, that the decision to recalculate can be justified on evidential
grounds. Indeed, it seems t me that such decisions cannot always be so justified.
2 Note that Skyrms’ shell game example (see Note 9 above) does not tell against this principle. The
principle pertains to pairs of acts, and the principle applies to none of the three possible pairs of
acts taken from Skyrms’ problem. It’s possible, of course, that in a decision problem there will be
several pairs of acts to which the principle is applicable. But note that the principle doesn’t say that
either of the acts in any such pail should be chosen – only that one should be, in the final analysis,
I shall not argue for this principle, but merely assume it, since it seems clearly
correct. I shall now argue that, given plausible and realistic assumptions about
the effects of an agent’s multiple calculations of expected utility (another kind
of metatickle), evidential decision theory will prescribe the two-box act in
Newcamb’s problem – my working prima facie counterexample to evidential
decision theory – and that it thus satisfies the above ratification principle in
such situation. (In the next section, I’ll argue, by way of a different example,
that causal decision theory also requires accommodation of metatickles if it is
to satisfy principle.)
As noted above, I agree with Jeffrey that, post-choice (no matter what the
choice is), A2 should appear evidentially superior to A1 in Newcomb’s pro-
blem. Thus, (i), (iii) of the ratification principle above are satisfied. This means
that, if evidential decision theory is adequate for our agent in the sense of the
principle, then it must, eventually, after sufficiently many calculations of
CEU, prescribe the two-box act and continue to do so. Also, if it can be
shown that it does, then, I take it, it will have been shown that evidential
decision theory does deliver the rational prescription in Newcomb’s problem
– albeit possibly after a false start.
The reason why A2 looks beter than A1 post-choice, no matter what the
choice, is that the cohice, whatever it is, renders the acts evidentially irrelevant
to the Ki’s. As deliberation proceeds, we can expect that the agent’s subjective
probability of one of the acts will eventually get very close to 1. A plausible
pre-choice version of the idea of post-choice evidential irrelevance of the Ai’s
to the Ki’s is the following, where the subscript ‘t’ on ‘Prt’ indicates the time at
which the value is taken:
Assumption 1: As Prt(A1) approaches 1 (if it does), Prt(Ki/Ai) approa-
ches Prt(Ki/A2), for i = 1, 2; and the same goes for the approach to 1 of
Prt(A2).
For some agents, the two conditional probabilities may approach the initial
probability of the state on the act whose subjective probability is approaching
1.1
Assumption 1 is about the effects of changes in the probabilities of the acts.
The next assumption says what causes the probabilities of the acts to change:
ranked above the other. A simpler principle would apply only to decision problems in which there
are only two available acts. But in the next section, I’ll want to apply the principle to a three-act
decision problem.
1 Brian Skyrms’ model of „deliberation-probability dynamics” (1982) makes an assumption
similar to assumption 1, which incorpotates this latter idea – both assumptions based on a point of
Riffard Jeffrey’s (1981, p. 486).
172 Ellery Eells
Assumption 2: If CEUt(A1) > CEUt(A2), then Prt+1(A1) > Prt(A1); and

this also holds when the two inequalities are reversed.
This assumption merely states that the probability of the act that is CEU-
superior on the most recent calculation increases as a result of the calculation.1
The discovered current CEU-superiority of an act is evidence that it is the
rational act (to CEU-maximizers), and is thus evidence that it will be chosen,
so that this evidence should make the act subjectively more probable than
before.
Both assumptions are qualitative in the sense that they do not describe
exactly what probability values are determined by the antecedent probability
or CEU-values. This is desirable, I think, because it makes the analysis more
general than otherwise. However, in order to carry the defense through, it will
be necessary to be more quantitative about the idea in assumption 2.
In the model of „deliberation-probability dynamics” investigated by
Skyrms (1982), it is assumed that the rate of change of the probabilities of A1
and A2 is proportional to the difference between their expected utilities. This
has the consequence that if the probability of A1 is low enough at the outset of
deliberation, then evidential decision theory will eventually prescribe A2, and
continue to do so until the close of deliberation; but for all other initial proba-
bilities of A1, short of 1, evidential deliberation will leave the probability of A1
stuck somewhat sort 1 – deliberation thus not closing with a decision. (For
details, see the reference cited, or Eells 1984b.) This need not be taken to be
the fault of evidential reasoning itself, however. Instead, we may lay the blame
on the way in which the probabilities of the acts are taken to depend on their
CEU’s. Certainly if „mixed” or „ramdomized” acts (in which the agent is
allowed to choose chances for the acts) are not allowed, then we decision
theorists would not prescribe such deliberation-probability dynamics to a
friendly CEU-maximizer. And a sophisticated CEU-maximizer who unders-
tood the implications of these dynamics would not embrace them in a
Newcomb problem.
It seems to me that the model investigated by Skyrms makes the probabili-
ties of the acts sensitive to one thing to which they should not be sensitive, and
insensitive to another thing to which they should be sensitive. As to the first
point, it seems that the movement of the probabilities of the acts should not
depend on the magnitude of the difference between their CEU’s, but rather
only on the sign of the difference. For an agent should not perceive an act to
be more or less rational according to how much greater its expected utility is
perceived to be than its competitors’ utilities; rather, the agent should perceive
1 This assumption is also a part of Skyrms’ (1982) model.

an act to be more or less rational according to how clear his perception ist that,
in the final analysis, its CEU will be (simply) greater than its competitors’ –
where the perceived difference between the CEU’s should be irrelevant. And
the agent’s subjective probability of performing an act should be based on his
assessment of whether or not it will be evidentially superior in the final analy-
sis, so this also should be based on the sign, rather than magnitude, of the
differences between the relevant CEU’s.
As to the second point, it seems that a rational agent should be sensitive to
the fact that deliberation must eventually end and an act then chosen. That is,
the agent should be sensitive to the approach of the time of action, when
deliberation must end and action begin. This puts pressure on the agent to
make a choice, and it seems that a realistic and plausible way to model an
agent’s sensitivity to this factor is to constrain his subjective propbabilities of
the available acts to more and more extreme values as the moment of decision
approaches. Another rationale for such an assumption is that later calculations
of expected utility reflect more information (namely the results of more nume-
rous prior calculations) than do the earlier ones, so they should have a more
dramatic effect on the agent’s assessments of the relative merits of the available
acts, and hence also on his assessment of the probability that he will take them.
There is a whole family of such constraints, some descriptive of decision
makers who procrastinate taking a stand on which act is correct and others
descriptive of more impulsive decision makers. In the following assumption,
the constant v indexes this aspect of the decision mar’s personality, st is equal
to 1 or – 1 according to whether A1’s CEU is greater than or less than that of
A2’s at the time t-1,1 and T is the amount of time available for deliberation:
Assumption 3: Prt(A1) = 1/2 + (1/2)st(t/T)v.
The formula assumes that at the beginning of deliberation the agent is com-
pletely uncertain about which act he will eventually perform, so that Pr0(A1)
= 1/2.2
The basic idea is that max {Prt(A1) = Pr0(A2)} gradually grows as deliberati-
on proceeds. Because of the agent’s awareness of the fact that he must have a
decision when the time of action arrives, the quantity approaches 1 as the time
of action approaches. The value of v controls how fast the quantity grows at
1 A more general formula, for cases in which the subjective probability of A1 should continue to
move in the same direction and at the same rate as it would if a previous inequality had been
sustained. Or perhaps the subjective probability of A1 should move to 1/2, and then, inequality
later obtains, the probability of A1 should then obey the assumption.
2 A more general formula, for cases in which the subjective probability of A1 doesn’t begin at 1/2,
would be: Prt(A1) = r + qt(t/T)v, where qt = 1-r or – r according to whether CEUt-1(A1) is greater
or less than CEUt(A2), and r is the initial probability of A1.
174 Ellery Eells
the beginning of deliberation compared to at the end of deliberation. Assump-

tion 1 and 3 have the consequence that if the agent make sufficiently many
calculations of expected utility before the time of action, evidential decision
theory will, in the end, prescribe the two-box act and continue to do so for all
further calculations of expected utility. It’s worth noting that the exact form
which assumption 3 takes is irrelevant to the force of this metatickle defense:
as will become apparent below, all that is necessary is that the greater of the
probabilities of the two acts gradually grows and approaches 1 as the time of
action closely approaches.
Initially, A1 will be evidentially superior, but, as the probability of A1
increases, both acts become less and less evidentially relevant to the Ki’s. And
when the acts are sufficiently evidentially irrelevant to the Ki’s, A2 will be
evidentially superior and as a result take on a hig probability. If the acts beco-
me evidentially irrelevant to the Ki’s as fast when the probability of A2 gets
high, then, after the initial leap in the probability of A2, this probability will
not turn around: it will continue to get higher so that the acts become more
and more evidentially irrelevant to the Ki’s and evidential decision theory will
continue to prescribe A2. See Figure 1. If, for some reason, the acts don’t get
evidentially irrelevant to the Ki’s as fast when the probability of A2 gets high
as they do when the probability of A1 gets high, then the probability of A1
may bounce up and down several times – but in the end, when the probability
of A2 is high enough that the acts are sufficiently evidentially irrelevant to the
Ki’s for A2 to be evidentially superior when A2 has high probability, A2 will
remain evidentially superior as the probability of A2 continues to grow into
the time of action. See Figur 2.
I admit, of course, that there will be some agents for whom the assump-
tions of this defense of the evidential theory will not hold true, in particular,
the proviso that the agent recalculates expected utility sufficiently many times.
But it is arguable that this is not the fault of the theory: indeed, perhaps all we
can conclude is that evidential decision theory is simply not adequate for such
agents, be that the fault of the agent or the theory. In the next section, we’ll see
that causal decision theory stands in the same relation to similar agents.
3. Metatickles and Causal Decision Theory
We have just seen that the results of calculating the evidential expectations of
the available acts is itself evidence (a metatickle) that can sometimes require
reassessments of the evidential relevance of the acts to the outcomes. Indeed,
it seems that for some agents, evidential decision theory would prescribe the
wrong course of action were it not for this phenomenon. Turning to causal
decision theory, are there situations in which the results of calculating the
causal expectations of the available acts will be evidence (a metatickle) that will
require reassessments of the causal relevance of the acts to the outcomes? More
interestingly, are there situations in which causal decision theory will deliver
an inappropriate prescription if such reassessments are not carried out? In
what follows, I will describe a decision problem for which the results of causal
calculations do require reassessments of the causal significance of the acts for
the outcomes? More interestingly, are there situations in which causal decision
theory will deliver an inappropriate prescription if such reassessments are not
carried out? In what follows, I will describe a decision problem for which the
results of causal calculations do require reassessments of the causal significance
of the acts for the outcomes. Decision problems for which just this is true have
already been described – e.g., Gibbard and Harper’s „Death in Damascus”
problem (1978, pp 157-159), Skyrms’ „mean” and „nice demon” problems
(1982, pp. 705-706), and Jeffrey’s „green-eyed monster” and „triumph of the
will” problems (1983, pp. 18-19). But for these problems, it is clear, because of
a kind of symmetry in the problems, that there is no unique appropriate
action. The decision problem which I shall describe, however, is one for which
causal decision theory can give the wrong answer initially, and one which
brings the ratification principle of the previous section to bear on the causal
theory, thus suggesting the need for causal expectation maximizers to listen to
their metatickles.
The decision problem for which this is true is built up from two others.
The first – call it ‘decision problem I’ – is just Newcomb’s problem, described
above. The second problem, which I’ll call ‘decision problem II’, is one in
which the agent has control over the contents of the opaque box. In problem
II, the „predicator wits until after your choice before filling, or not filling, the
opaque box with the $ 1,000,000. In problem II, if you take just the opaque
box (act A1), then the predictor will put the $1,000,000 in, but if you take both
boxes (act A2), he leaves it out. Thus, the one-box act leads, you know, to your
176 Ellery Eells
getting $ 1,000,000, and the two-box act leads, you know, to your getting just
$1,000.
Finally, decision problem III is combined from problems I and II as fol-
lows.1 You have three available acts in decision problem III, A1 and A2 as
above, and A0: take no boxes, the invariable result of which is receiving $999.
Call the predictors of problems I and II ‘predictor I’ and ‘predictor II’, respec-
tively. In the third problem, there is a third predictor, predictor III, who
predicts whether or not you will do A0. If he predicts A0, then he turns you
over to predictor II (essentially putting you in decision problem II, except that
you have that third alternative), and if he predicts that you won’t do A0, he
turns you over to predictor I (essentially putting you into decision problem I,
Newcomb’s problem, except that again you have that third alternative). Your
subjective probabilities are such that it is extremely probable, conditional on
your choice, that predictor III has predicted that choice, and it is extremely
probable that predictor I has predicted your choice if it is one of A1 and A2;2
„predictor” II, you believe, is infallible.
Letting F0 and ~F0 be the prior states of predictor III’s predicting A0 and
~A0, respectively, and K1 and K2 the prior states of predictor 1’s predicting A1
and A2, respectively, the causal expectations of the three available acts are as
follows, assuming monetary utilities:
Uk(A0) = 999;
Uk(A1) = Pr(F0)1,000,000 + Pr( ~F0&K1)1,000,000;
Uk(A2) = Pr(F0)1,000 + Pr( ~F0&K1)1,001,000
+ Pr( ~F0&K2)1,000.
Let r = Pr(F0) and p = Pr(K1). And assume that F0 and K1 are independent of
each other. (Say predictor I makes a prediction no matter what predictor III
does, based on what the thinks you would do were you to choose one of A1
and A2. Or perhaps predictor I believes you are in decision problem I rather
than in decision problem III.3 Then
1 For a simpler example of this, see Eells (1985 b).

2 Actually, the agent’s believing predictor I to be accurate in the above sense, or even at all, is not
essential for my purposes. Indeed, this predictor, for the purposes for which I shall use decision
problem III, could decide in some random way which act to predict, as can be seen from the
calculations that follow. Decision problem III is complex enough as it is, however, and it is easier
to describe it (and perhaps also to grasp it) in terms of decision problem II and the already familiar
Newcomb’s problem, in which it is assumed that the predictor is accurate in the above sense.
3 As pointed out in Note 20, the causal ranking of the acts, given by the calculations below, would
be unaffected if predictor I decided what to predict in some random way, e.g., by flipping a coin,
in which case the plausibility of K1’s subjective probabilistic independence from F0 would be even
more plausible.
Uk(A0) = 999,
Uk(A1) = 1,000,000r + 1,000,000(1-r)p
= 1,000,000r + 1,000,000p – 1,000,000rp;
Uk(A2) = 1,000r + 1,001,000(1-r)p + 1,000(1-r) (1-p)
= 1,000,000p – 1,000,000rp + 1,000,
so that
Uk(A1) > Uk(A2) if and only if r > 1/1000.
And note that no matter what the values or r and p are, Uk(A2) is at least
1,000, so it is always greater than Uk(A0).
Thus, it is quite plausible that, on the initial calculation of causal expectati-
on, A1 will be the prescribed action. Suppose A1 is chosen on this basis. Be-
cause of the way in which we have assumed predictor III to be accurate accor-
ding to the agent’s beliefs, this should cause the agent’s subjective probability
of F0 to decrease. Suppose it gets lower than 1/1000. This means that, post-
choice, A2 has a causal expectation greater than that of A1. Thus, clause (ii) of
the antecedent of the ratification principle of the previous section is satisfied.
Now suppose that A2 is chosen on the basis of causal decision theory’s
prescribing it. This implies that at that point r > 1/1000. And it seems that the
effect of the choice of A2 on the probability of F0 can only be to lower that
probability, in which case, post-choice, causal decision theory will continue to
prescribe A2 – over both A1 and A0, the latter of which, as we have seen, has a
lower causal expectation than A2 no matter what the values of r and p are.
Thus, clause (iii) of the antecedent of the ratification principle is also satisfied,
as is of course, clause (i).
Thus, according to the principle of ratification of the previous section,
causal decision theory must, in the end, rank A2 over A1 – or else be inadequa-
te for the agent in this decision problem. Indeed, A2 is the only one of the
three available actions that one should like to carry out once one has decided
to do so, where if one were to have decided on any of the other actions, one
should then prefer to carry out A2 instead. (Actually, a choice of A0 should
make A1 optimal, though A2 would still be better than A0. Act A0 is evidence
that one is in problem II, in which A1 is best. But, as we have seen, if the agent
is then lowed to change his choice to A1, he would then regret is choice, wis-
hing he had chosen A2 instead.) On this basis, it seems that A2 is the unique
rational act in this decision problem.
How can causal decision theory avoid prescribing A1? We could build in a
ratification clause, which would allow the decision maker to change his choice
if it is not ratifiable. But adding such a ratification clause would, it seems,
result in the agent’s viewing his initial choice as only tentative. And if the
178 Ellery Eells
agent views his choice, of say A1 in decision problem III, as only tentative,
then he may not view the act as sufficiently probable to make F0 sufficiently
improbable to render A2 causally superior, whereas a choice viewed as final
would. (In the case of evidential decision theory, a tentative initial choice in a
Newcomb problem may not immediately render the acts sufficiently evidenti-
ally irrelevant to the outcomes to render the dominant action evidentially
superior, whereas a choice viewed as final should.) Also, as in the case of
evidential decision theory, there is nothing in ratifiablitiy analysis that in-
structs us how always to arrive at a ratifiable choice, one that we should view
as final. And, in addition, there is that viscious circularity problem, described
in the previous section, for ratifiability as itself a criterion of choice.
On the other hand, we could build into causal decision theory a sensitivity
on the part of the agent to metatickles of the kind discussed last in the previ-
ous section. Then, after the first calculation of expected utility, on which act
A1 has superior causal expectation, the probability of A1 goes up – perhaps
only a little though, the agent being wary (the value of v of the previous secti-
on’s deliberation-probability dynamics is greater than 1). The second calculati-
on still gives A1 the highest expectation, say, and the probability of A1 goes up
a bit more – as does that of ~F0. Eventually, though, the probability of A1 is
sufficiently high for the probability of F0 to be sufficiently low (so that it
becomes sufficiently probable that one is essentially in the original Newcomb
problem) for the causal expectation of A2 to become superior. At this point,
the agent’s subjective probability of A1 may have become quite high, though,
so that the probability of A2 is still low. Here, just as in the case of the third
metatickle defense of evidential decision theory, it might be necessary for the
agent’s sensitivity to the approach of the time of decision to result in a leap in
the probability of A1, from very close to 1 to very close to 0. For, in the first
place, decreases in the probability of A1 in small increments will not do if this
happens to the benefit of the probability A0 as well as that of A2: when the
probability of A0 is not too low, the probability of F0 will be high enough (
1/1000) so that A1 again becomes the causally superior act. And in the second
place, the agent might be running out of time for deliberation, so that, without
such a leap, the probability of the rational act, A2, might not get close enough
to 1 for its subjective probability to reflect a choice of that act.1
1 One might be inclined to suggest that a jump in the probability of A2

4. Conclusion
If evidential decision theory is to deliver the appropriate prescriptions in

Newcomb problems, then the agent must be sensitive to metatickles of some
kind. But the intricacy of metatickle analysis, and the fact that it is at a relati-
vely primitive stage of development (with respect to both formal rigor and
intuitive plausibility), does not constitute an advantage of causal decision
theory over the evidential theory. For we have seen that there is a class of
decision problems in which causal decision theory can also give wrong ans-
wers, unless metatickles of some kind are taken into account. And it seems that
the effects of metatickles must be very much the same in the application of the
causal theory as they must be in the application of the evidential theory if
either theory is to give the right answers – albeit in different kinds of decision
problems, and after false starts.
It thus appears that a final assessment of the relative merits of the causal
and evidential paradigms of rational deliberation must await further refine-
ments in the analysis of the role of „metatickles” in rational deliberation –
refinements both on the level of intuitive plausibility and on the level of for-
mal rigor (paying attention to both prescriptive and descriptive adequacy).
Metatickle analysis is essential to the „rescue” of both evidential and causal
decision theory, and it appears that the fate of each will turn on the develop-
ments and refinements of metatickle theory to come.
I thank the University of Wisconsin-Madison Graduate School for financial support, and Elliott
Sober for useful comments on an earlier draft. Rep. From PSA 1994, Vol. 2, pp. 177-200, With
permission of the Philosophy of Science Association.
References
Campbell, R. And Sowden, L. (eds.) (1985): Paradoxes of Rationality and Cooperation: Prisoner’s
Dilemma and Newcomb’s Problem, University of Britisch Columbia Press, Yancouver.
Cartwright, N. (1979): „Causal Laws and Effective Strategies,” Nous 13, 419-437.
Eells, E. (1980): Newcomb’s Paradox and the Principle of Maximizing Conditional Expected
Utility, Unpublished Ph.D. Dissertation, University of California, Berkeley, University Micro-
films Publication Number ADG80-29383.
—— (1982): Rational Decision and Causaltiy, Cambridge University press, Cambridge, Entland,
and New York.
—— and Sober, E. (1983): „Probabilistic Causality and the Question of Transitivity,” Philosophy
of Science 50, 35-57.
—— (1996): „Common Causes and Decision Theory,” Philosophy of Science 53, 223-245.
—— (1984a): „Newcomb’s Many Solutions,” Theory and Decision 16, 59-105.
—— (1984b): „Metatickles and the Dynamics of Deliberation,” Theory and Decision 17, 71-95.
—— (1985a): „Reply to Jackson and Pargetter,” in Cambell and Sowden (1985) 219-223.
180 Ellery Eells
—— (1985b): „Weirich on Decision Instability,” Australasian Journal of Philosophy 63, 473-478.

Garber, D. (1983): „Old Evidence and Logical Omniscience in Bayesian Confirmation Theory,”
in J. Earman (eds.): Testing Scientific Theories, University of Minnesota Press, Minneapolis, 99-
131.
Gibbard, A. And Harper, W. (1978): „Counterfactuals and Two Kinds of Expected Utility,” in
Foundations and Applications of Decision Theory, Vol. 1 (C.A. Hooker, J.JU. Leach, and E.F.
McClennen (eds.): The University of Western Ontario Series in Philosophy of Science, Vol. 13.
Reidel, Dordrecht, 125-162. As reprinted in Ifs. W.Harper, R. Stalnaker, and G. Pearce
(eds.)(1981): The University of Western Ontario Series in Philosophy of Science, Vol. 15, Reidel,
Dordrecht, 153-190. Also rep. in Campbell and Sowden (1985), 133-158.)
Hacking, I. (1967): „Slightly More Realistic Personal Probability,” Philosophy of Science 34, 311-
325.
Horwich, P. (1983): „Decision Theory in the Light of Newcomb’s Problem,” unpublished
manuscript, MIT.
Jackson, F. And Pargetter, R. (1983): „Where the Tickle Defense Goes Wrong,” Australasian
Journal of Philosophy 61, 295-299.
Authors 181
AUTHORS
Prof. J. Anthony Blair, Department of History, Philosophy & Political Scien-

ce, University of Windsor, Windsor, Ontario, Canada N9B3P4.
Prof. Ellery Eells, University of Wisconsin - Madison, Department of Phi-
losophy, Madison, WI 53706, USA.
Prof. Leo Groarke, Wilfrid Laurier University, Department of Philosophy,
Ontario, Canada.
Prof. Henry W. Johnstone, Jr. Department of Philosophy, The Pennsylvania
State University, 2450 Sparks Building, University Park, PA 16802, USA.
Prof. Ralph H. Johnson, Philosophy Program, Windsor, Ontario, Canada
N9B3P4.
Prof. Dr. Hans Lenk, Universität Karlsruhe, Fak. für Geistes- und Sozial-
wissenschaften, Kaiserstraße 12, D-76128 Karlsruhe, Germany.
Prof. Manfred Kienpointer, Karl-Franzens-Universität, 8910 Graz, Austria.
Prof. Dr. Dr. Dieter Mans, J.W. Goethe-Universität, FB: Gesellschaftswissen-
schaften, 60054 Frankfurt am Main, Germany.
Prof. Robert C. Pinto, Philosophy, University of Windsor, Windsor, Ontario,
Canada N9B 3P4.
Dr. habil. Gerhard Preyer, Privatdozent, J.W. Goethe-Universität, FB: Gesell-
schaftswissenschaften, 60054 Frankfurt am Main, Germany.
Prof. Chris Tindale, Department of Philosophy, Catharine Parr, Traill Colle-
ge, Trent University 310 London Street, Peterborough, Ontario, Canada/K9H
7P4.
Prof. Douglas Walton, Department of Philosophy, University of Winnipeg,
Winnipeg R3B 2E9, Canada.
Prof. John Woods, Department of Philosophy, University of Lethbridge, 4401
University Drive, Lethbridge, Alberta Canada T1K 3M4.
182 Imprint
IMPRINT
PROTO
SOCIOLOGY
An International Journal of Interdisciplinary Research ISSN 1434-4319
Editor: Gerhard Preyer
Johann Wolfgang Goethe-University Frankfurt am Main,
Department of Social Sciences
Editorial Staff: Georg Peter
Editorial of the Vol. 13 1999: Georg Peter
Project: Frank Siebelt, Dieter Mans, Mathias Boes
Layout und Druckvorstufe: Georg Peter
Editorial Office: PROTOSOCIOLOGY, Stephan-Heise-Str. 56, 60488 Frankfurt am
Main, RFA, Phone: 069-769461,
Email: preyer@em.uni-frankfurt.de, peter@protosociology.de
Bank: Dresdner Bank AG, Frankfurt am Main, Kto: 44 121 168 01, BLZ: 500 800 00
Wir danken debis, concept and products für ihre freundliche Unterstützung.

Die Zeitschrift soll 1/2jährlich erscheinen. Die Anzahl der jährlich erscheinenden Hefte und
Sonderhefte bleibt jedoch vorbehalten.
Copyright: Die in dieser Zeitschrift veröffentlichten Beiträge sind urheberrechtlich geschützt. Alle
Rechte sind vorbehalten. Übersetzungen, Nachdruck, Vervielfältigung auf fotomechanischem
oder ähnlichem Weg oder im Magnettonverfahren, Wiedergabe durch Vortrag, Funk- und
Fernsehsendungen sowie Speicherung in Datenverarbeitungsanlagen, auch auszugsweise, sind nur
mit Genehmigung des Herausgebers möglich. Für eingereichte Beiträge wird keine Haftung über-
nommen. Weitere Publikationsrechte von Artikeln bleiben vorbehalten. Zweitpublikationen
seitens des Autors werden eingeräumt. Bei einer Zweitpublikation ist das Heft (Nummer, Titel,
Erscheinungsjahr) der PROTOSOCIOLOGY zu zitieren. Für unaufgefordert eingesandte
Manuskripte wird keine Haftung übernommen. Gerichtsstand ist Frankfurt am Main.
Copyright: All rights reserved. This publication may not be reproduced, stored or transmitted in
any form or by any means without the prior permission in writing of the publisher. Additional
publications of the articles are reserved. The authors retain the personal right to re-use their own
articles. Authorization to photocopy items for internal or personal use, or the internal or personal
use of specific clients is garanted by PROTOSOCIOLOGY, provided that the base fee is paid
directly to VG Wort, Goethestr. 49, 80336 München RFA.The publisher accepts no responsibility
for submitted manuscripts.
ProtoSociology im Internet:
http://www.protosociology.de
http://www.rz.uni-frankfurt.de/protosociology
On ProtoSociology 183
ON PROTOSOCIOLOGY
“Protosociology occupies an important position in the European intellectual

scene, bridging philosophy, economics, sociology and related disciplines. Its
volumes on rationality bring together concerns in all these topics, and present
an important challenge to the cognitive sciences.”
Donald Davidson, Berkeley (USA)
“Protosociology publishes original papers of great interest that deal with funda-
mental issues in the human and social science. No academic library is complete
without it.”
Nicholas Rescher, Pittsburgh (USA)
“Protosociology has been remarkably successful in publishing interesting work

from different tradition and different disciplines and, as the title signals, in
giving that work a new, eye-catching slant.”
Philipp Pettit, Canberra, Australia
“Protosociology is a truly premier interdisciplinary journal that publishes arti-

cles and reviews on timely topics written by and for a wide range of interna-
tional scholars. The recent volumes on rationality are remarkable for their
breadth and depth. Protosociology would be a great addition to any library.”
Roger Gibson, St. Louis (USA)
184 Volumes
PROTOSOCIOLOGY
VOL. 12, 1998 — Special Edition
AFTER THE RECEIVED VIEW — Developments in the Theory of Science

Gerhard Preyer, Georg Peter, Alexander Ulfig (Eds.)
in memoriam
Wolfgang Stegmüller
Content
Introduction: Developments in the Theory of Science

Gerhard Preyer, Georg Peter, Alexander Ulfig
LOGICAL OPERATIONALISM — SIGNIFICANCE AND MEANING
Wilhelm K. Essler
Truth and Knowledge. Some Considerations concerning
the Task of Philosophy of Science
Gerhard Preyer
The Received View, Incommensurability and
Comparison of Theories — Beliefs as the Basis of Theorizing
Robert Schwartz
Reflections on Projection
Jeffrey E. Foss
The Logical and Sociological Structure of Science
STRUCTURALISM — MEANINGFUL MEASUREMENT —

THE CONCEPTION OF PHYSICAL LAW
C. Ulises Moulines
Structuralism vs. Operationalism
Nicholas Rescher
Meaningless Numbers
Volumes 185
R. I. G. Hughes
Laws of Nature, Laws of Physics, and the Representational
Account of Theories
James R. Brown
Einstein’s Principle Theory
INDUCTIVE INFERENCES — INTERPRETATION OF PROBABILITY — GAME THEORY
Kevin T. Kelly, Cory Juhl

Transcendental Deductions and Universal Architectures for Inductive
Inferences
Howard H. Harriott
R.A. Fisher and the Interpretation of Probability
Brian Skyrms
Evolution of an Anomaly
PROPERTIES — UNDERDETERMINATION — SCIENTIFIC REALISM
George N. Schlesinger
Degrees of Characterizations
Carl A. Matheson
Observational Adequacy as distinct from the Truth about Observables
Thomas R. Grimes
Scientific Realism and the Problem of Underdetermination
Paul C. L. Tang
On Paul Churchland’s Treatment of the Argument from Introspection
and Scientific Realism
RATIONALITY — METAPHORS — VALUES IN SCIENCE
David Resnik
Scientific Rationality and Epistemic Goals
Aldo Montesano
Rationality in Economics: A General Framework
Joseph Agassi
Science Real and Ideal: Popper and the Dogmatic Scientist
Michael Bradie
Models and Metaphors in Science
David Gruender
Values and the Philosophy of Science
186 Volumes
ProtoSociology
VOL. 11, 1998

COGNITIVE SEMANTICS II — EXTERNALISM IN DEBATE
Content
RADICAL INTERPRETATION, ONTOLOGY AND THE STRUCTURE OF LANGUAGE
Gerhard Preyer, Michael Roth

On Donald Davidson’s Philosophy: An Outline
Richard Manning
All Facts Great and Small
Barbara Fultner
Of Parts and Wholes: The Molecularist Critique of Semantic Holism
Louis Goble
Re-Evaluating Supervaluations
David Simpson
Interpretation and Skill: On Passing Theory
Wulf Kellerwessel
Katz on Semantics and Pragmatics
EXTERNALISM AND THE INDIVIDUATION OF CONTENT
Ron Wilburn
Knowledge, Content, and the Wellstrings of Objectivity
Anthony Brueckner
Content Externalism and A Priori Knowledge
Consuelo Preti
The Irrelevance of Supervenience
Michael Liston
Externalist Determinants of Reference
Arnold Silverberg
Semantic Externalism. A Response to Chomsky
Gerhard Preyer
Interpretation and Rationality:
Steps from Radical Interpretation to the Externalism of Triangulation
Volumes 187
ProtoSociology
VOL. 10, 1997

COGNITIVE SEMANTICS I — CONCEPTIONS OF MEANING
Content
MEANING, TRUTH AND THE USE OF LANGUAGE
Herman Cappelen, Ernie LePore Jeffrey King

Semantic Theory and Indirect Speech The Source(s) of Necessity
Kirk Ludwig Filip Buekens
The Truth about Moods The Genesis of Meaning (a Myth)
Louise Röska-Hardy
Language, Use and Action
REFERENCE, INDEXICALS AND SPEAKER MEANING
Robert Hanna Gerhard Preyer

Extending Direct Reference Verstehen, Referenz, Wahrheit. Zur Phi-
Peter Ludlow losophie Hilary Putnams
Semantics, Tense, and Time: a Note on Reinaldo Elugardo
Tenseless Truth-Conditions for Token- Descriptions, Indexicals and Speaker
Reflexive Tensed Sentences Meaning
RATIONAL DIALETISM OF ANTINOMIES
Mark Sainsbury
Can Rational Dialetheism Be Refuted
By Considerations about Negation and Denial?
CONTEMPORARY DEBATES IN THE PHILOSOPHY OF MIND
Klaus Sachs-Hombach
Philosophy of Mind: Die Simulationstheorie
IN RETROSPECT
Joseph Agassi Kent A. Peacock, Richard Feist

Wittgenstein – The End of a Myth The Einstein-DeSitter Controversy
188 Volumes
ProtoSociology
VOL. 8/9, 1996
RATIONALITY II & III
INTERPRETATION, INTENTIONS AND PROPOSITIONAL ATTITUDES

David K. Henderson: Epistemic Rationality, Epistemic Motivation, and Interpretive
Charity; Roger F. Gibson: Stich on Intentionality and Rationality; Alfred Mele: Ratio-
nal Intentions and the Toxin Puzzle; John Heil: The Propositional Attitudes; Volkmar
Taube: Exemplifikatorische Darstellung. Zu den Grundlagen einer kognitiven Ästhetik
NATURALIZED EPISTEMOLOGY, RATIONALITY AND NORMATIVITY

Paul K. Moser, David Yandell: Against Naturalizing Rationality; Harvey Siegel:
Naturalism, Instrumental Rationality and the Normativity of Epistemology; Ralf
Naumann: Internal Realism, Rationality and Dynamic Semantics; Axel Wüstehube:
Noch einmal: Rationalität und Normativität
RATIONAL EXPLANATION, REASONING AND JUSTIFICATION

Philip Pettit: Three Aspects of Rational Explanation; Keith Lehrer: Rationality and
Trustworthiness; Alexander Ulfig: Stufen der Rechtfertigung
CONCEPTS OF PRACTICAL RATIONALITY

Peter A. French: Rationality and Ethics; Pierre Kerszberg: Feeling and Coercion: Kant
and the Deduction of Right; Raymound Boudon: Une éthique est-elle possible en
l'absence de croyances dogmatiques?; Raimo Tuomela: Rational Cooperation and
Collective Goals; Wolfgang Welsch: Vernunft heute
REVIEWS
STATE OF THE ART
Marcus Birke: Externalismus in der Philosophie des Geistes
SCIENCE AND HISTORY

Joseph Agassi: Die gegenwärtige Rolle des Technik- und Wissenschaftshistorikers
'RXEOHYROXPHSDJHVSULFH'0RULQFOXGLQJFRXUWDJH
GLJLWDOYHUVLRQSULFH'0RU
http://www.rz.uni-frankfurt.de/protosociology
Volumes 189
Forthcoming
ProtoSociology
PARSONS-SPECIAL EDITION
CLASSICAL SOURCES AND CURRENT USES OF TALCOTT PARSONS’S

THOUGHT
edited and with an introduction by Bernard Barber and Uta Gerhard
This volume wishes to cut across the vast array of misunderstandings, misreadings, and
misinterpretations which have marred the proper recognition of Parsons’s work for the
last thirty years. If the question is asked why such recognition would be important, the
answer is twofold. For one, Parsons’s work is a classic that reaches from the l930s into
the l970s of the twentieth century while sociology’s now arguably uncontroversial
classics - like Max Weber, or Emile Durkheim - were only able to analyze society until
about the end of World War I. Therefore, Parsons’s oeuvre has a special modernity in
that his work takes into account the New Deal, National Socialism, McCarthyism, and
the Student Rebellion, to name but a few timely events which are part and parcel of
Parsons’s sociological analysis. Furthermore, his conception of social order and anomie,
especially since they seriously discuss the problematic viability of utilitarian economics,
are of particular importance today.
Introductionary Essay:
The Parsons Agenda
Bernard Barber, Uta Gerhardt
PART 1: Revisiting The Structure of Social Action
Mark Gould: Neoclassical Economics, Social Order, and “The Structure of Social
Action”
Uta Gerhardt: National Socialism and the Politics of “The Structure of Social Action”
Richard Münch: The Problem of Order: Sixty Years After “The Structure of Social
Action”
PART 2: Understanding Parsons’s Thought

Bryan S. Turner: Sacricity and Solidarity in Talcott Parsons’s Economic Sociology
Edward A.Tiryakian: An Emergent French Connection: Revisiting Parsons’s
Durkheim
Bernard Barber: The Nature and Functions of Religion and Beyond
PART III: Evaluating the Scope of Sociology

Daniel Bell: The Limits of General Theory in the Social Sciences
Hans Jürgen Puhle: (Title to be announced)
190 Bookpublications
%RRNSXEOLFDWLRQV
PROTOSOZIOLOGIE IM KONTEXT
„Lebenswelt“ und „System“ in Philosophie und Soziologie
Gerhard Preyer, Georg Peter, Alexander Ulfig (Hrsg.)

in memoriam Paul Lorenzen
Einleitung: „Lebenswelt“ und „System“ in Philosophie und Soziologie, Gerhard
Preyer, Georg Peter, Alexander Ulfig
ZUM BEGRIFF DER LEBENSWELT:
Ernst W. Orth (Trier, Deutschland): Lebenswelt als unvermeidliche Illusion? Husserls
Lebensweltbegriff und seine kulturpolitischen Weiterungen, Walter Biemel (Aachen,
Deutschland): Gedanken zur Genesis der Lebenswelt, Alexander Ulfig (Frankfurt/
Main, Deutschland): Lebenswelt und Reflexion. Anhang: Lebenswelt als Fundament
der Wissenschaft, Gerhard Preyer (Frankfurt/Main, Deutschland): Hintergrundwissen:
Kritik eines Begriffs, Hubert A. Knoblauch (Konstanz, Deutschland): Soziologie als
strenge Wissenschaft? Phänomenologie, kommunikative Lebenswelt und soziologische
Methodologie
LEBENSWELT — BEGRÜNDUNG — WISSENSCHAFT
Jürgen Mittelstraß (Konstanz, Deutschland): Das lebensweltliche Apriori, Peter
Janich (Marburg, Deutschland): Die Rationalität der Naturwissenschaften, Jürgen
Mittelstraß (Konstanz, Deutschland): Rationalität und Reproduzierbarkeit, Elisabeth
Ströker (Köln, Deutschland):Lebenswelt durch Wissenschaft, Paul Janssen (Köln,
Deutschland): Lebenswelt, Wissen und Wissenschaft, Richard T. Murphy (Concord,
USA): E. Husserl's Phenomenology of Reason
LEBENSWELT/LEBENSFORM — SPRACHE
Pierre Kerszberg (Pennsylvania, USA): Lifeworld and Language, John F.M. Hunter
(Toronto, Canada): The Motley Forms of Life in the Later Wittgenstein, Peter A.
French (San Antonio, USA): Why did Wittgenstein read Tagore to the Vienna Circle?
Georg Peter (Frankfurt/Main, Deutschland): Die Nebenbeschäftigung der Symbole.
Zu Wahrheit und Funktion der Metapher
SYSTEM — SOZIALSYSTEM — GESELLSCHAFT
Niklas Luhmann (Bielefeld, Deutschland): Die Lebenswelt nach Rücksprache mit
Phänomenologen, Niklas Luhmann (Bielefeld, Deutschland): Observing Re-entries,
Gerhard Preyer (Frankfurt/Main, Deutschland): System-, Medien- und Evolutions-
theorie. Zu Niklas Luhmanns Ansatz, Richard Münch (Düsseldorf, Deutschland):
Autopoesis per Definition, Hans Zitko (Frankfurt/Main, Deutschland): Ästhetik als
Soziologie der Kunst, James Bohman (St. Louis, USA): The Completeness of Macro-
Sociological Explanations: System and Lifeworld, Göran Ahrne (Stockholm,
Schweden): Outline of an Organisational Theory of Society, Anhang: Karl Otto
Hondrich (Frankfurt/Main, Deutschland): Zu Göran Ahrnes Ansatz
392 Seiten, 68,- DM
Verlag Königshausen & Neumann, Würzburg 1996
STRUKTURELLE EVOLUTION UND DAS WELTSYSTEM

Theorien, Sozialstruktur und evolutionäre Entwicklungen
Gerhard Preyer (Hrsg.)
Einleitung:
Gerhard Preyer: Strukturelle Evolution und das Weltsystem: Theorien,

Sozialstruktur und evolutionäre Entwicklungen
ZENTUM UND PERIPHERIE — INSTITUTIONELLE ENTWICKLUNG — ASKRIPTIVE
SOLIDARITÄT
Shmuel Noah Eisenstadt: Social Division of Labor, Construction of Centers and

Institutional Dynamics: A Reassessment of the Structural—Evolutionary Perspective,
Bernhard Giesen, Kay Junge: Strukturelle Evolution, Gerhard Preyer: Soziale
Gesetze und askriptive Solidarität. Eine Skizze zu den Grundlagen der
Gesellschaftstheorie, Anhang: Die modernen Gesellschaften verstehen. Richard Münchs
Entwicklungstheorie moderner Gesellschaften, Erwin Rogler, Gerhard Preyer:
Relationslogische Darstellung der sozialen Gesetze, Gerhard Preyer: Die modernen
Gesellschaften „verstehen“. Zu Richard Münchs Entwicklungstheorie moderner
Gesellschaften, Dieter Claessens: Bemerkungen zur Entstehung der modernen
Ökonomie: Das Organistionsproblem, Richard Pieper: Strukturelle Emotionen,
elementare Strukturbildung und strukturelle Evolution
DIE EVOLUTION POLITISCHER ORDNUNGEN
Rainer C. Baum: Parsons on Evolution of Democracy, Mathias Bös: Zur Evolution
nationalstaatlich verfaßter Gesellschaften, Konrad Thomas: Das Ethnische und das
Staatliche, Volker Bornschier: Die westeuropäische Integration als Gesellschaftsmodell
im Zentrumswettbewerb
ZUR SOZIOLOGIE DES WELTSYSTEMS
Roland Robertson: Globalization and reflexive Modernization, Immanuel
Wallerstein: Evolution of the Modern World-System, Christopher Chase-Dunn,
Thomas D. Hall: The Historical Evolution of World-Systems: Iterations and Trans-
formations, Albert Bergesen: Postmodernism: A World System Explanation, Richard
Münch: Modernity and Irrationality: Paradoxes of Moral Modernization, Walter L.
Bühl: Transformation oder strukturelle Evolution? Zum Problem der Steuerbarkeit
von sozialen Systemen
STATE OF THE ART: Michael Schmid: Soziologische Evolutionstheorien
Suhrkamp Verlag, Frankfurt am Main 1998, Suhrkamp Taschenbuch

Wissenschaft
LEBENSWELT — SYSTEM — GESELLSCHAFT

Konstruktionsprobleme der „Theorie des kommunikativen
Handelns“ von Jürgen Habermas
*HUKDUG3UH\HU
I DIE ENTWICKLUNGSLOGIK VON WELTBILDERN

1. Die Rationalisierung von Weltbildern
2. Strategien der Analyse von Weltbildern
3. Folgeprobleme und Kritik
II GESELLSCHAFT ALS LEBENSWELT UND SYSTEM
1. Die Strukturen der Lebenswelt
2. Allgemeine Bezugsprobleme der soziologischen Evolutionstheorie
3. Die Verständigungsformen
1. Weltbilder und soziale Integration
2. Zur Durkheim-Interpretation
4. Kommunikationsmedien und generalisierte Kommunikationsweisen
5. Folgeprobleme und Kritik
III DIE FORMAL-PRAGMATISCHE BEDEUTUNGSTHEORIE
1. Die sprechakttheoretische Grundlegung
2. Interpersonal geltende Bedeutungskonventionen
3. Regelbewußtsein und Handlungskompetenz
4. Die Expansion des semantischen Gehalts
5. Verständigung und die Herstellung interpersonaler Beziehungen
IV DER ERWERB DES MORALISCHEN BEWUßTSEINS
1. Die sozial-kognitive Grundausstattung
2. Diskurs und moralisches Bewußtsein
3. Folgeprobleme
V KONSTRUKTIONSPROBLEME UND KRITIK
1. Rekonstruktionshypothesen
2. Zu den Konstruktionsproblemen
ANHANG:
Max Webers Religionssoziologie als eine Typologie des Rationalismus
290 Seiten, DM 28,– , auch als CD lieferbar
Digital bookpublication
Humanities-Online, Frankfurt am Main
http://www.humanities-online.de
Email: editor@humanities-online.de
Die globale Herausforderung

Wie Deutschland an die Weltspitze zurückkehren kann
Gerhard Preyer
Der Begriff Globalisierung hat bereits Schlagwortcharakter bekommen und ist in aller
Munde. Globale Märkte haben die Grundsituation wirtschaftlichen und politischen
Handelns grundsätzlich verändert. Das heute entstehende globale Weltsystem und der
Medienverbund auf dem es beruht, fordert uns alle heraus. Was bedeutet Globali-
sierung? Worin bestehen die globalen Herausforderungen, die zu bewältigen sind?
Darauf wird eine Antwort gegeben.
NEUE MEDIEN, WISSENSCHAFT UND TECHNIK IM ZEITALTER DER GLOBALISIERUNG

Was heißt Globalisierung? – Die neuen Medien: Eine Kopernikanische Wende – Der
Überlebensimperativ: Technologieentwicklung – Die Zukunft der deutsche Universi-
täten
VERÄNDERTE KONSTELLATIONEN
Globale Wirtschaft und globale Ordnung – Die Globalisierung der Finanzmärkte –
Auf dem Weg zur virtuellen Organisation – Das Ende des klassischen Arbeitsmarktes
– Zu einer neuen Wirtschaftspolitik
EUROPA IM ZEITALTER DER GLOBALISIERUNG
Zur Ausgangssituation im Zentrumswettbewerb – Frankreichs Zentralismus: Grenzen
der Marktwirtschaft – Das Überlebenssystem Italien – Groß-britannien zwischen
Tradition und Modernisierung – Deutsche Stärken im Umbruch – Neu gemischte
Karten – Die Wiedergewinnung des Standorts Deutschland
DIE EVOLUTION DES MITGLIEDSCHAFTSCODES
Zur Soziologie der Grenzziehungen – Gesellschaft, Organisation und Interaktion –
Gesellschaftsinterne Globalisierung und Entwicklungstrends – Die globalisierte
Gesellschaft
290 Seiten, DM 72,–
Frankfurter Allgemeine Zeitung/Gabler Edition, Frankfurt am Main

1998
INTEGRIERTES MANAGEMENT
Was kommt nach der Lean-Production?
Gerhard Preyer/Jakob Schissler
Es gibt heute im Weltmaßstab kein Unternehmen, das seine Umstrukturierung nicht an
den Bestandteilen des japanischen Modells der schlanken Produktion und des schlanken
Managements, der Lean-Modelle, orientiert. Die Lean-Modelle sind zu einem inte-
grierten Management fortzuentwickeln: dem über Vernetzung integrierten
segmentierten Unternehmen, mit durchlässigen, ständig wechselnden Grenzen zwischen
Unternehmen, Lieferanten und Kunden.
Integriertes Management ist das Unternehmensmodell, das eine erhöhte Qualifizierung
der Mitarbeiter und die Fähigkeit, in sich selbst steuernden Teams mitzuarbeiten,
erfordert. Das bedeutet aber auch die Entwicklung eines neuen Führungsstils. Mit inte-
griertem Management kann jedes Unternehmen sofort anfangen. Dazu wird in diesem
Buch ein Weg gezeigt. Es wird eine wichtige Funktion als Impuls für weitere Entwick-
lungen haben.
Inhalt
DIE NEUEN HERAUSFORDERUNGEN
1. Vom japanischen und amerikanischen Management lernen?
2. Information, Wissen und Entscheidungen
KREATIVITÄT — ORGANISATION — FÜHRUNG

1. Managementfähigkeiten
2. Innovative Organisation
3. Management und Führungsstile
STRATEGIE — MARKT — KULTUR

1. Unternehmensstrategie
2. Controlling
3. Marketing
4. Zur Unternehmenskultur
STICHWORT
Was kommt nach der Lean Production?
164 Seiten, 32,- DM
Frankfurter Allgemeine Zeitung, Verlagsbereich Wirtschaftsbücher,

Blickbuch-Wirtschaft, Frankfurt am Main 1996
ÜBER DAS KÄMPFEN

Zum Phänomen des Kampfes in Sport und Gesellschaft
Axel Binhack
Kämpfen ist stets als ein Beziehungsphänomen, gleichsam als „Fortsetzung der Kommu-
nikation mit anderen Mitteln“ zu verstehen. Anhand von sechs Strukturmerkmalen
wird eine „Prototypik des Kampfes“ entwickelt. Ihre zusammenhängende Dynamik ist
gesellschafts- und kulturunabhängig in vielfältigen Konfliktformen zu erkennen. In
dem vorgelegten interdisziplinären Anschnitt wird die Struktur der Kampfbeziehung
von vergleichbaren Sozialphänomenen abgegrenzt und am Beispiel des Kampfsports in
Form des „Zweikampfes“ analysiert.
0HUNPDOHHLQHU)RUPDOVWUXNWXUGHV.DPSIHV
Der duale Beziehungsaspekt: Antagonismus und Ambivalenz — Der duale Träger-
aspekt: Entscheidungsorientiertheit und tendenzielle energetische Totalität — Der
duale Inhaltsaspekt: Zweckgerichtetheit und riskante Offenheit
'LH6WUXNWXUPHUNPDOHGHV.DPSIHVLP9HUJOHLFKPLW$UEHLW6SLHOXQGGHQ
%H]LHKXQJVSKlQRPHQHQ7DXVFK)OLUWXQG0DFKW
Der „antagonistische“ Unterschied — Freiheit und Zweckgerichtetheit — „Zeitentho-
bene“ innere Unendlichkeit und „zeitbeschleunigende“ Entscheidungsorientiertheit —
Die formalstrukturelle „Folgenhaftigkeit“ des Kampfes
=XP.DPSIVSRUW
Kampfsport als Radikalisierung des sportlichen Kampfes — Der Kampfsport als kul-
turelles Phänomen — Der Kampfdiskurs als Symbol männlichen Agierens — Ästhe-
tische Perspektiven
'LH.DPSIVSRUWDUW.DUDWH'RDOV%HLVSLHO
Historische Entwicklungslinien des Karate-Do — Der Schwertkampf und die „Leere
Hand“: Zum karatespezifischen Aspekt der Waffenlosigkeit — Die Kultivierung des
Kampfphänomens im Karate-Do — Der Kampf als Lehrmeister im Sinne des Zen
Campus Verlag, Campus Forschung Bd. 768, Frankfurt am Main 1998

System der Rechte, demokratischer Rechtsstaat und Diskurstheorie

des Rechts nach Jürgen Habermas
hrsg. von Werner Krawietz / Gerhard Preyer
Editorial: Zivilgesellschaftliche Assoziationen und spätmoderner

Rechtssstaat (Werner Krawietz)
I JURISTISCHE ENTSCHEIDUNG, BEGRÜNDUNG UND RECHTFERTIGUNG DES RECHTS IN DER

PERSPEKTIVE DISKURSIVER REFLEXIONSTHEORIE
Enrique P. Huba: Standortbestimmung in der zeitgenössischen Rechtstheorie - Rawls,
Dworkin, Habermas, Alexy und andere Mitglieder der modernen ‚Heiligen (Rede-)
Familie‘, Thomas McCarthy: Legitimacy and Diversity: Dialectical Reflexions on Analytical
Distinctions, Gerhard Preyer: Rechtsgeltung — Argumentation — Entscheidung
II ORDNUNG, RECHT UND RATIONALITÄT IM JURISTISCHEN DISKURS

Karl-Heinz Ladeur: Rechtliche Ordnungsbildung unter Ungewißheitsbedingungen und
intersubjektive Rationalität, Ulfrid Neumann: Zur Interpretation des forensischen Diskur-
ses in der Rechtsphilosophie von Jürgen Habermas, Ota Weinberger: Diskursive
Demokratie ohne Diskurstheorie
III KONSENSUSTHEORIE VERSUS SUBJEKTPHILOSOPHIE? VORAUSSETZUNGEN UND FOLGEN DES

JURISTISCHEN DISKURSES IM DEMOKRATISCHEN RECHTSTAAT
Michael Pawlik: Die Verdrängung des Subjekts und ihre Folgen. Begründungsdefizite in
Habermas’ „System der Rechte“, Uwe Steinhoff: Probleme der Legitimation des
demokratischen Rechtsstaats, William Rehg: The Place of Consensus in Democratic
Legitimation: A Recommendation
Rechtstheorie
Zeitschrift für Logik, Methodenlehre, Normentheorie und Soziologie des Rechts 3
1996, Habermas-Sonderheft (erschienen 1998)
Duncker und Humblot, Berlin
INTENTION — BEDEUTUNG — KOMMUNIKATION

Kognitive und handlungstheoretische Grundlagen der Sprachtheorie
Gerhard Preyer, Maria Ulkan, Alexander Ulfig (Hrsg.)
Einleitung: Zu kognitiven und handlungstheoretischen Grundlagen der Sprach-

theorie, Gerhard Preyer, Maria Ulkan, Alexander Ulfig
I INTENTIONEN UND KOMMUNIKATIVE HANDLUNGEN

Maria Ulkan: Kommunikative und illokutionäre Akte; Georg Meggle/Maria Ulkan:
Grices Doppelfehler. Ein Nachtrag zum Griceschen Grundmodell; Jan Nuyts:
Intentionalität und Sprachfunktionen
II INTERPRETATION UND BEDEUTUNG

Gerhard Preyer: Kognitive Semantik, Anhang Sprechaktsemantik: J.L. Austin, J.R.
Searle, H.P. Grice, P.F. Strawson; Louise Röska-Hardy: Sprechen, Sprache, Handeln;
Frank Siebelt: Zweierlei Holismus. Überlegungen zur Interpretationstheorie Donald
Davidsons; Peter Rothermel: Semantische Implikaturen; Volkmar Taube: Referenz
und Interpretation. Zur Theorie nichtsprachlicher Symbolisierung; Georg Peter: Zu
Richtigkeit und Interpretation der Metapher: Kognitive Funktion und rekonstruktive
Schemainterpretation
III KLASSIFIKATION VON SPRECHAKTEN

Maria Ulkan: Informations- und Aufforderungshandlungen; Dirk Hartmann: Kon-
struktive Sprechakttheorie; Volkmar Taube: Bildliche Sprechakte
IV KOMMUNIKATIVES HANDELN UND INTERSUBJEKTIVE GÜLTIGKEIT

Jürgen Habermas: Sprechakttheoretische Erläuterungen zum Begriff der kommunika-
tiven Rationalität; Karl-Otto Apel: Illokutionäre Bedeutung und normative Gül-
tigkeit. Die transzendentalpragmatische Begründung der uneingeschränkten kommuni-
kativen Verständigung; Peter-Paul König: Kommunikatives und strategisches
Handeln. Kritische Bemerkungen zu zwei zentralen Begriffen der „Theorie des
kommunikativen Handelns“ von Jürgen Habermas; Alexander Ulfig: Präsuppositionen
und Hintergrundwissen. Eine Kritik am formalpragmatischen Präsuppositionsbegriff
V DIALOGSTRUKTUR UND ARGUMENTATION

Wilhelm Franke: Konzepte linguistischer Dialogforschung; Franz Hundsnurscher:
Streitspezifische Sprechakte: Vorwerfen, Insistieren, Beschimpfen; Dieter Mans: Argu-
mentation im Kontext, Exkurs: Zu Christoph Lumers „Praktische Argumentations-
theorie“
408 Seiten, 58.- DM
Westdeutscher Verlag, Opladen 1997

Wittgensteins Spätphilosophie
Analysen und Probleme
Wulf Kellerwessel, Thomas Peuker (Hrsg.)
Die Spätphilosophie Ludwig Wittgensteins entfaltet nicht nur in der gegenwärtigen

Sprachphilosophie ihre Wirkung, sondern auch zunehmend über den Bereich der
Sprachphilosophie hinaus. Zwar macht dieser einen Schwerpunkt des Wittgensteinschen
Denkens aus, aber dennoch verdanken die Philosophie des Geistes, die Erkenntnistheo-
rie und die Philosophie der Religion der Wittgensteinschen Sprachphilosophie wichtige
Impulse. Diese Schwerpunktsetzung zugunsten der Sprachphilosophie wie auch ihre
Zusammenhänge mit den genanten anderen Disziplinen der Philosophie zeichnet der
vorliegende Band aus.
Inhalt
W. Kellerwessel, T. Peuker: Einleitung, A. Ofsti: Methodischer Solipsismus,
Metasprachen(problem) und Dexis. Einige Überlegungen zum Privatsprachenproblem,
T. Peuker: Die faktische Öffentlichkeit der Sprache. Zu Wittgensteins Privatsprachen-
argument, P. Niesen: Gemeinschaft, Normativität, Praxis: Zur Debatte über Wittgen-
steins Regelbegriff, A. Berndzen: Einer Regel entsprechen - einer Regel folgen. Zu einer
Kontroverse in der Interpretation von Wittgensteins ‘Philosophischen Untersuchungen’,
G. Preyer: Sprachbedeutung ohne Regelbefolgung, C. Stetter: Sprache und Schrift bei
Wittgenstein, H.J. Schneider: Mentale Zustände als metaphorische Schöpfungen, W.
Kellerwessel: Zum Begriff der Gewißheit in Wittgensteins ‘Über Gewißheit’. Ein
Kommentar, M. Kroß: „Glaube Du! Es schadet nicht“. Ludwig Wittgensteins
Vermischte Schriften zur Religion - Anhang.
Verlag Königshausen & Neumann, Würzburg

Donald Davidsons Philosophie

Von der radikalen Interpretation zum radikalen Kontextualismus
Gerhard Preyer
Für Bruce Aune

EINLEITUNG
I RADIKALE INTERPRETATION, LOGISCHE FORM UND EREIGNISSE

1. Donald Davidsons Philosophie: Ein Überblick
2. Wahrheit, Bedeutung und radikale Interpretation
2.1 Radikale Interpretation als ein radikaler Externalismus
(1) Der Zirkel zwischen Überzeugung und Bedeutung:
Die Asymmetrie von RI
(2) Die Strukur von Inhaltssätzen: sagen, daß ...
(3) Der Grundsatz der Nachsicht
(4) Die Autonomie der Bedeutung
(5) Radikaler Externalismus: Der innertheoretische Schritt
(6) Die Ontologie von RI
2.2 Von der Idiolekttheorie zum dritten Dogma des Empirismus
(1) Ausgangs- und Übergangstheorien der Interpretation
(2) Die Demontage eines Mythos
(3) Die epistemischen Beschränkungen von Verstehen
3. Die logische Form von Handlungssätzen und die singuläre Kausalaussage
3.1 Logische Form und adverbiale Modifikation
3.2 Kausale Beziehungen
4. Wahrheit und Überzeugung
4.1 Zum Haupteinwand
4.2 Zur empirischen Offensichtlichkeit von Einstellungszuschreibung
4.3 Rationalität als normativer Begriff
II KÖRPERBEWEGUNGEN UND HANDLUNGEN

1. Das logische Verknüpfungsargument
2. Primäre Gründe und die Identitätsthese:

Die synkategormatische Fassung von Absichten
3. Basisakte
4. Flucht vor den Körperbewegungen
4.1 Thalbergs Handlungstheorie und der „Akkordeon-Effekt“
4.2 A.I. Goldmans Kritik an der Identitätsthese
4.3 H.L.A. Hart: Zuschreibungen
4.4 Körperbewegungen als Bestandteil von Handlungen
III EINE RADIKALE THEORIE DES HANDELNS

1. Der Begriff der Einstellungsrationalität
1.1 Die Homogenität der Interpretation
1.2 Bewertende Einstellungen
2. Handlungsbeschreibungen und Handlungsverursachung
3. Eine Handlungserklärung
3.1 Überzeugungen, Absichten, Situationen
3.2 Praktische Gedanken
3.3 Handlungsgründe
3.4 Zu Hempels Ansatz
4. Praktische Schlüsse
4.1 Zur Gültigkeit praktischen Schließens
4.2 Entscheidungen und die Ausführung von Absichten
4.3 Überzeugungen und willentliche Handlungen
5. Radikaler Kontextualismus
6. Individualismus versus Holismus
7. K.R. Poppers Kritik am Historizismus
Literatur

Digital Publication:
Order from: http://www.rz.uni-frankfurt.de/protosociology
Language, Mind and Epistemology

On Donald Davidson’s Philosophy
Gerhard Preyer, Frank Siebelt, Alexander Ulfig (eds.)
Introduction: On Donald Davidson's Philosophy (Gerhard Preyer, Frank Siebelt,

Alexander Ulfig)
PHILOSOPHY OF LANGUAGE
Jerry Fodor, Ernie Lepore (New Brunswick, USA): Meaning, Holism, and the Problem of
Extensionality; Olav Gjelsvik (Oslo, Norway): Davidson's Use of Truth in Accounting for
Meaning; Wilhelm K. Essler (Frankfurt/Main, Germany): Was ist Wahrheit? Arend
Kulenkampff (Frankfurt/Main, Germany): Eigennamen und Kennzeichnungen
EPISTEMOLOGY
Roger F. Gibson (St. Louis, USA): Quine and Davidson: Two Naturalized Epistemologists;
Eva Picardi (Bolongna, Italy): Davidson and Quine on Observation Sentences; Ralf
Naumann (Düsseldorf, Germany): Events and Externalism; Dorit Bar-On (Chapel Hill,
USA): Conceptual Relativism and Translation; David K. Henderson (Memphis, USA):
Conceptual Schemes after Davidson; Frank Siebelt (Frankfurt/Main, Germany): Singular
Causal Sentences and two Relational Views
PHILOSOPHY OF MIND AND THEORY OF ACTION

Louise M. Antony (Raleigh, USA): The Inadequacy of Anomalous Monism as a Realist
Theory of Mind, Louise Röska-Hardy (Darmstadt, Germany): Internalism, Externalism and
Davidsons Conception of the Mental, Marcia Cavell (Berkeley, USA): Dividing the Self,
Ralf Stoecker (Bielefeld, Germany): Willensschwäche - Wie ist das nur möglich? Klaus
Puhl (Graz, Austria): Davidson on intentional Content and Self-Knowledge, Johannes
Brandl (Salzburg, Austria): Sharing Beliefs and the Myth of the Subjective, Kirk Ludwig
(Gainesville, USA): First Person Knowledge and Authority, Gerhard Preyer
(Frankfurt/Main, Germany): Rationalität: Absichten, Primärgründe und praktisches
Denken.
Donald Davidson: Dialectic and Dialogue
445 Seiten
Kluwer Academic Publishers, Synthese Library, Dordrecht
Reality and Humean Supervenience

Essays on the Philosophy of David Lewis
Gerhard Preyer, Frank Siebelt (eds.)
Preface
Gerhard Preyer, Frank Siebelt

Reality and Humean Supervenience - Some Reflections on David Lewis’ Philosophy
MODAL REALISM
Phillip Bricker: Island Universes and the Analysis of Modality, John Bigelow: Time
Travel Fictions, Peter Forrest: Counting the Cost of Modal Realism, Paul Teller:
Lewis's Defence of Counterpart Theory, Harold W. Noonan: The Case for Per-
durance,
PHYSICALISM, CAUSATION AND CONDITIONALS

Daniel Bonevac: Naturalism for the Faint of Heart, D. M. Armstrong: Going throug
the Open Door again: Counterfactual vs. Singularists Theories of Causation, Jonathan
Bennett: On Forward and Backward Counterfactual Conditionals
REDUCTION OF MIND
Terence Horgan: Multiple Reference, Multiple Realization and the Reduction of Mind,
Michael Tye: Knowing what it is like: The Ability Hypothesis and the Knowledge
Argument
Rowman and Littlefield Publishers, Lanham, USA

LOGICAL FORM, LANGUAGE AND ONTOLOGY

On Contemporary Philosophy of Language and Linguistics
Gerhard Preyer, Georg Peter (eds.)
One of the central topics analytical philosophy and especially the theory of language is
concerned with is the concept of logical form. As typically understood the concept of
logical form not only covers investigations into universal logical features underlying
languages. Conceived that way the logical form of, say, arguments, sentences, and other
parts of language may be the focus of linguists and semanticists. However, from Frege
and Russell on logical form analysis were not confined to such a narrow linguistic
perspective. For them investigating the logical form of languages always followed the
wider philosophical perspective of trying to understand the structure of language as our
principle means for representing the world. From Russell’s theory of definite descriptions
up to Davidson’s truth-theoretical analysis of adverbial modification, citation and
reported speech, to lay open the logical structure underlying language was always seen
as a means to reveal the structure and features of the thereby represented world.
Following such a broader philosophical perspective, the book contains several new essays
which discuss for example: Russell’s understanding of logical form analysis, the
relational structure of belief-sentences, the descriptivist view of that-clause-structures in
languages, the logical sources of intensionality, the logical content of de se attitudes, the
relation between ontological questions and questions regrading logical form. The
collection brings together work by philosophers from diverse points of view, and as such,
it illuminates the lively and ongoing debate the concept of logical form still arouses
within contemporary philosophy.
Contributors:
Mark Richard (Medford,USA), Barry Schein (Los Angeles, USA), Bernard Linsky,
James Higginbotham (Oxford, Great Britain), Norbert Hornstein (Maryland, USA),
Marga Reimers, (Tucson, Arizona), Ernest Lepore, Kirk Ludwig (New Brunswick,
Gainesville, USA,), Richard Larson, (Stony Brook, New York), James Tomberlin
(Northridge, USA), Robert May/R.F. Fiengo, (Irvine, USA), Robert J. Stainton,
(Ottowa, Canada), Ray Elugardo (Oklahoma, USA), Peter Ludlow, (Stony Brook,
USA), Stephan Neale (Berkeley, USA ), Roger Schwarzschild, (New Brunswick, USA),
Thomas Baldwin, (York, Great Britain), Robert Mattews, (New Brunswick, USA),
Mark Crimmins, (Ann Arbor, USA), Jeffrey King, (Davis,USA)
Forthcoming
For further Information see: http://www.rz.uni-frankfurt.de/protosociology

204 Subscription
PROTOSOCIOLOGY ISSN 0940-4147

Subscription
I subscribe PROTOSOCIOLOGY 30,- DM (15,- Euro) each issue, abroad: $ 30

(including courtage); payable by cheque. No price fixing for Double-Volumes.
Date, signature ...................
Separate order
I order PROTOSOCIOLOGY volume Nr. .......
Single issue: 35,- DM (17,50 Euro), abroad: $ 35 (including courtage); payable by

cheque.
All prices plus postage and packing
Date, signature ............
Your address:
Name .................... Address ....................
Send the order to:
PROTOSOCIOLOGY ISSN 0940-4147
Editorial office: Stephan-Heise-Str. 56, D-60488 Frankfurt am Main 90, E-Mail:

preyer@em.uni-frankfurt.de, Tel. 069/769461
Can be ordered directly from: Bockenheimer Bücherwarte

Bockenheimer Landstr. 127
60325 Frankfurt am Main, RFA
Tel.: 069-771088
PROTOSOCIOLOGY participates in Poiesis Online Service, Philosopical

Dokumentation Center, Bowling Green State University, Bowling Green, OH
43403-0189, http://www. bgsu.edu/pdc/
Editor: Gerhard Preyer, Johann Wolfgang Goethe-Universität Frankfurt am Main,

Department of Social Sciences. Editorial staff: Georg Peter.
http://www.protosociology.de
Digital Publications – Special Offer 205
Digital Publications – Special Offer:

ProtoSociology publishs digital versions of its volumes.
Vol. 11 is a free download from our homepage under “Special Service”.
Vol 13 – Reasoning and Argumentation

Vol 12 – After the Received View – Developments in the Theory of Science
Vol.11 – Cognitive Semantics II – Externalism in Debate (free download!)
Vol 10 – Cognitive Semantics I – Conceptions of Meaning
Vol 8/9 – Rationality II &III (double volume)
Single digital version (Journals only): 7,50 Euro or $ 15.-

Special offer:
Three digital Journals for the price of two:
15.- Euro or $ 30. (including courtage)
Payable by cheque. E-Mail posting!
Date, signature ............
Your address:
Name .................... Address ....................
Send the order to:

PROTOSOCIOLOGY, editorial office: Stephan-Heise-Str. 56, D-60488 Frankfurt am
Main 90, Tel. 069/769461
Order via E-Mail: preyer@em.uni-frankfurt.de or peter@protosociology.de
Editor: Gerhard Preyer, Johann Wolfgang Goethe-Universität Frankfurt am Main

FB 3: Department of Social Sciences. Editorial staff: Georg Peter. Project: Frank
Siebelt, Dieter Mans, Mathias Bös
KWWSZZZU]XQLIUDQNIXUWGHSURWRVRFLRORJ\
206
Was ist DENKBAR?
Ziel der DENKBAR ist die Förderung

von Philosophie, Kunst und
Wissenschaft. Sie bietet zum Beispiel
Experten wie auch Studenten die
Möglichkeit, in Ihren Räumen
wissenschaftliche Vorträge zu halten.In
diesem Sinne versteht sich die
DENKBAR als ‘philosophische
Werkstatt’ und Denkraum für Kunst
und Kultur. Wenn Sie an den Zielen
der DENKBAR interessiert sind, laden
wir Sie sehr herzlich ein, sich mit uns
Alles ist DENKBAR! auf das “Abenteuer der Vernunft”
(Kant) einzulassen
Denkbar
Schillerstraße 26
60313 Frankfurt am Main
Öffnungszeiten:
täglich – außer Sonntag – von 18.00 bis 23.00 Uhr
207
.,1'(5'25)
gibt Kindern eine Zukunft
3(58
Schenken Sie mit uns
Straßenkindern eine Zukunft
Meine Vision
Das Kinderdorf Peru ist meine Antwort auf ein soziales Problem in Lima/Peru. Nachdem
ich selbst, vom Schicksal begünstigt, 30 Jahre Erfolg in meinem Frankfurter Modege-
schäften erleben konnte, möchte ich dem Leben etwas von dem Glück und Überfluss
zurückgeben, mit dem es mich bedachte.
Die Not und das Elend der Straßenkinder in Peru, die ich auf einer Reise in Südamerika
kennenlernte, bewegten mich zu dem Entschluss, helfen zu wollen und Not zu lindern.
Weil Helfen Freude macht, entschloss ich mich Ende 1995, meine Firma zu verkaufen,
das Kinderdorf Peru zu planen und hierfür einen Förderverein zu gründen.
Inzwischen verbringe ich mehrere Monate des Jahres in Lima, um dort vor Ort den
schwächsten Mitgliedern der Gesellschaft, den Kindern, zu helfen.
Ein Teil meines privaten Vermögens habe ich bereits in die ersten Projekte investiert.
Zur Erstellung und Unterhaltung des Kinderdorfes jedoch bitte ich um Ihre Mithilfe, da
dies meine finanziellen Möglichkeiten übersteigt.
Ich würde mich freuen, wenn Sie meine Vision teilen könnten und Ihre Hilfe den
Kindern geben würden.
Robert Lemli
Mehr über das

KINDERDORF PERU
können Sie auf unserer
Hompage erfahren: http://www.kinderdorf-peru.de
Email: Kinderdorf.Peru@t-online.de
Kinderdorf Peru e.V., Tannenwaldstr. 6,
D-61389 Schmitten, Tel.: 06082-930273
ProtoSociology
Vol.13 - 1999
Reasoning and Argumentation
P Editors of the volume: Gerhard Preyer and Dieter Mans

P Editorial of the Vol. 13 1999: Georg Peter
P Layout and Technical Conception: Georg Peter
P Editorial Office: ProtoSociology, Stephan-Heise-Str. 56, 60488
Frankfurt am Main, RFA, Phone: 069-769461, E-Mail:
preyer@em.uni-frankfurt.de

(Dieter Mans, Gerhard Preyer (Eds.) ) Proto Sociolo (BookFi) PDF

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

(Dieter Mans, Gerhard Preyer (Eds.) ) Proto Sociolo (BookFi) PDF

Încărcat de

Drepturi de autor:

Formate disponibile

Proto

VOL. 13, 1999

REASONING AND ARGUMENTATION

Henry W. Johnstone, Jr.

Digital Publications – Special Offer 205

How shall we distinguish between argumentation and reasoning? What is

This is a perfect argument in the following sense. Since the conclusion is a

Blair gives an answer by looking at an very important class of arguments.

1 These critical questions belong to an argumentation scheme. Walton discusses 25 of these

1 This term has been coined by S. Toulmin (1958).

1 D. Walton, op.cit. 1996.

John Woods’ contribution is about abduction or abductive inferences.

description points to three ingredients in any theory of abduction. The first

Dieter Mans, Gerhard Preyer

Reasoning, Argumentation and The Network Problem

1 See Lambert and Ulrich (1980) and Govier (1987:182).

(S3) 99% of ravens are black.

II. The Network Problem

Once we begin to focus our attention on reasoning, we open up another a host

III. The Nature of Reasoning: Some Positions

1 Finocchiaro is here rejecting an approach to conceptualizing reasoning that approximates the

when the reasoner produces what is called a non-sequitur? Is such a specimen

C. Walton on reasoning and argument (1990)

1 The criteria given by Johnson-Laird are as follows (65-66):

Without wishing to claim completeness for this inventory of how phi-

IV. Reasoning and Argument

3. When I make a controversial assertion and back it up with reasons, I

species. That would be a mistake comparable to identifying fruits with apples.

Therefore Boston is in the United States.

The Fox and the Hedgehog: On Logic, Argument, and Argumentation

In many ways, argumentation theory is a reaction against the concept of

Argument Traditionally Conceived

We might describe the “traditional” concept of argument simply. It defines an

Though they are often misconstrued in argumentation texts (see Groarke

than true or certain premises. This is an important change to the traditional

1 I am indebted to Lorne Falkenstein for this point.

Argument in Argumentation Theory

Though many argumentation theorists have criticized the notion of argument

traditional way, while inductive arguments are understood and assessed in

Another important instance of argumentation theory’s willingness to

Kinds of Argument and the Deductive/Inductive Distinction

Argumentation theory’s attempt to distinguish different kinds of argument is

implicitly, make claims about what sort of connection is supposed to hold

argumentation which are standardly assumed to be most removed from argu-

Convincing and Persuading Reconsidered

reasoning. An impassioned plea for an end to genocide in Kosovo will inevita-

Expanding the Realm of Argument

kidnapper presents his argument in a non-verbal way. A famous example

—— (1998). Argument Structure: A Pragmatic Theory. Toronto: University of Toronto Press.

Presumptive Reasoning/Argument: An Overlooked Class1

The aim of this paper is to draw attention to an important, distinctive, relative-

2. Reasoning and argument, and their norms

Reasoning, whatever else it may entail, is uncontroversially at least the mental

act conventions, or by failing to respond to the interlocutor’s challenges as

premises and their relation to their conclusion, it is rational to accept that

case A in the salient respects; so a token of action/judgement-type X is correct

(d) All “balance of considerations” or “pro and con” reasoning or arguments.

4. The challenge of deductivism

Whether these variants of presumptive reasoning/arguments represent a dis-

Argumentation 1986, Vol. 3, Argumentation: Analyses and Practices, 34-41. Dordrecht-

Argument Schemes and the Evaluation of Presumptive Reasoning:

In “Presumptive Reasoning/Argument: An Overlooked Class,” Professor

1) In what sense does presumptive reasoning/argument constitute a sui