Documente Academic
Documente Profesional
Documente Cultură
TO UNOBSERVABLES
IN SCIENCE AND
PHILOSOPHY .
Richard J. Connell
Connell, Richard J.
From observables to unobservables in science and philosophy I
Richard 1. Connell.
p. em.
Includes bibliographical references and index.
I. Knowledge, Theory of. 2. Science-Philosophy. 3.
Empiricism. 4. Inference. 5. Concepts. I. Title.
BDI81.C667 2000 121--dc21 00-023498 CIP
Q'"'
'V The paper used in this publication meets the minimum
requirements of American National Standard for Information
Sciences-Permanence of Paper for Printed Library Materials,
' ANSI 239.48-1984
Contents
Preface
Bibliography 227
Index 231
Preface
A while ago Richard Rorty said to philosophers 1 that they serve only
a very limited function in society. Because they have failed to live up to
their claims to solve systematically the problems that are concerned with
the various kinds of human knowledge (regarded, since Descartes, as
philosophy's proper domain), they are now left with the improvement of
human conversation as their sole function, which they exercise in part by
acting as gadflies, stimulating discussion to see whether various "theories"
"hang together." No one, Rorty says, ought to think that philosophy
actually understands anything about the real world, the world "out there."
Whether or not we agree with Rorty, we certainly must admit that
philosophy is experiencing a malaise, a sign of which is the way
philosophers speak. Somewhere Thomas More remarks that lawyers are
able to take a proposition with a very plain sense and after talking about
it for a time make it so obscure that no one knows any longer what it says.
And it would seem that lawyers are not a singular species in this regard,
for too many philosophers appear to share the same "talent," in evidence
of which one need only look at the journals, filled as they are with the
obscurity that comes from complex and prolix language, from the
multiplication of jargons, and from the use of words in senses removed
from those of ordinary speech. However, many problems arise not merely
from the obscure use of language but instead from philosophers such as
the deconstructionists who claim that words do not put us in touch with
an objective world, claiming that the only reality we have is that which we
form from language. This is indeed a sign of an unhappy condition of the
modem intelligence ..
When we read newspapers, magazines, and many books being
published today we can see, upon examination, that the authors do not
know what the mind is supposed to do and how it is to be governed. The
problem has many facets, not the least of which is the post-modernist-as
it is called-abandonment of objective sources and standards for the
mind's judgments and opinions. Of course there are serious philosophers
From Observables to Unobservables
who do not maintain that the only reality we know is the one we create
and who seek to discover how things really are. Some of them are in the
field of the philosophy of science, and a number of the topics we shall
take up here are of interest to them. Still, we cannot undertake to translate
their language and to expound upon their many expressions; nor can we
discuss even those topics they treat which stand in some relation to our
own. Were we to do so we would at the very least obscure our work. And
so we wish to say that we shall presuppose a realist position in philosophy
and then go on to consider in detail where the intelligence gets its starting
points in observation and how it moves from them. More specifically, our
basic issue is to show how the mind moves from sensible characteristics
to realities beyond sensation; that is, to show how the mind moves from
observables to unobservables not only in its basic operations but also how
in particular ways this movement is realized in systematic disciplines, that
is, in science and philosophy. We shall not take up here the realist anti-
realist debate, but instead we shall attempt to cast some light on realism
by the procedures we mentioned just above. As far as we know this has
not until now been done; so now let us continue.
The contrast between intellectual and sensory knowledge caused some
of the pre-Socratics to deny the reality of sensible properties. Later
Descartes put philosophy inside the mind when he started from his cogito,
in contrast to Hume who confined philosophy to observables-impre-
ssions of the senses-when he claimed that ideas, which presumably
belong to the intelligence, are but fainter copies of impressions,
precluding thereby the mind's coming to understand anything unob-
servable. Since that time philosophy has not been able to return to the
objective world, a world that has been abandoned to the natural sciences
and to a lesser extent to those that are social. As a consequence realist
philosophy, despite the statements of many who claim that the mind must
start from observation, does not seem to know where and how to begin;
and so as we said we propose to show how the mind comes to understand
unobservables from observables: that is the principal issue. Once we have
completed that task, and one we have talked about the sciences, we shall
say a little about what philosophy is and attempt to argue that it is first of
all a systematic discipline that bears on the real world and so depends
upon ordinary experience and observation both for its starting points and
for its measure. We then can describe philosophy in a general way as
systematically considered common experience, and it must of course
originate in observation.
Preface
It was for many centuries a central belief that the whole world,
from the motion of falling bodies to the existence of God, could
ultimately be understood by pure thought alone. Thus philosophers
and theologians labored mightily to construct "proofs" of the
existence of God; thus thinkers such as Aristotle and Plato declared
that nature worked according to certain principles, and sought to
deduce from those principles the phenomena of the natural world,
relegating to a distant second place empirical tests of whether the
world behaved as their theories required it to. 2
Apparently Lindley thinks that before the days ofNewton academics paid
little attention to observation, to sensory data as the foundation for its
conceptions of nature. But few minds in the history of the human species
have been more empirical than Aristotle, however difficult that may be to
see. And despite some of his theories, many of Plato's dialogues have
solid anchorings in sensible data. But let us look at the Introduction in a
second source where the authors say the following:
Our chapters have, in the main, been constructed in such a way that
they can be read independently, and the notes and references are
collected together accordingly. Scientists with no interest in the
history of ideas can just skip the chapters in which they are
discussed. Likewise, non-scientists can avoid mathematics
altogether if they wish. One last word: the authors are cosmolo-
gists, not philosophers. This has one very important consequence
which the average reader should bear in mind. Whereas philoso-
phers and theologians appear to possess an emotional attachment
to their theories and ideas which requires them to believe them,
scientists tend to regard their ideas differently. They are interested
in formulating many logically consistent possibilities, leaving any
judgment regarding their truth to observation. 3
Notes
II
Chapter 1
The Problem
A. Introduction
For a realist observables are the foundation of all that we know about
unobservables, and our dependence on them is manifest in the external
expressions of the intelligence. Not surprisingly, then, observables and
their cognitive relation to unobservables have been topics of discussion
among philosophers. The problem that arises from this relation has been
posed in several ways: one can ask how we relate observables to
theoretical entities, or he can ask about the ground for the distinction
between observation and inference, or he can ask without qualification
how we know unobservables. But these different questions indicate that
the issues are not viewed in exactly the same way by all, a state of affairs
that requires us to present our own case and, we hope, with a minimum of
entanglements. As we said in our Preface, a treatment of every aspect of
the problem is impossible, and obtaining a commonly accepted set of
philosophical starting points for its resolution is at the very least difficult.
No one should find it strange, then, if we first delineate the starting points
that will guide our discussion, and to set the stage for that delineation let
us first contrast a philosophic position with the position of the ordinary
man. The contrast will be instructive because of the light it will cast on
our starting points.
B. A contrast
To the ordinary man Quine's position appears strange, for the ordinary
man regards physical objects as real and inseparably related to experience:
physical objects are realities that are known through direct contact. For
him, to think of physical objects as conceptually postulated entities
differing only in degree from Homer's gods is equivalent to denying what
is most obvious, namely, the reality of the world "out there." But as the
reader knows, that is essentially what many philosophers question, if not
actually deny. They tend to spend their professional lives inside their
heads, lamenting their inability to escape.
Currently the word "empirical," which one might take to imply a
contact with extrinsic realities, is often employed to mean something that
is only psychological, something that appears internally but cannot be
regarded as putting us in contact with realities of any sort. Such a position
is either skeptical or solipsistic, and so it would seem that we may point
to two statements that represent the content of this position: ( 1) we know
sensory impressions, not things as they are "out there" (2) observables
(impressions) do not lead the mind to unobservable realities. The reader
will recognize Hume and Kant in these propositions, we shall,
nonetheless, present thumbnail sketches of what they say about
observables and unobservables.
C. Historical antecedents
Hume told us that ultimately all our ideas originate from impressions
of the senses, and Locke before him had held a similar view. Ideas, Hume
said, are but fainter copies of impressions, which means that our
understanding is limited by the capacity of things to produce sensory
The Problem 3
It does not matter whether the view one holds is precisely that
described by W artofsky or whether it is a variant that shifts the ground
from language to concepts, making the business of being theory-laden
conceptual rather than linguistic. In either case, what counts as an
observable is a matter of choice, assuming that no claim is made about a
theory imposing itself necessarily upon the intelligence. Such positions
stand in plain contrast to those that regard "the boundary between what
can be observed and what must be inferred [to be] largely determined by
fixed, architectural features of an organism's sensory perceptual
psychology," as Jerry Fodor claims. 4 Moreover, the reader will recognize
that what we have said so far bids fair to introduce many philosophic
problems that have something to do with the issues related to foundatio-
nalism and the primitive "givens" of philosophical discourse. For the
moment, however, we shall pass them by, since our aim here is to find a
starting point that the reader will accept. By so doing, we shall be able to
leave many issues aside for a later time.
The starting point we think acceptable is the following: the recognized,
systematic disciplines regularly appeal to facts both to generate problems
and to measure proposed solutions. In other words,facts are commonly
admitted to have priority over both the interpretations made of them and
the inferences drawn from them, and it matters not whether the systematic
discipline is a natural science, a social science, or listing it separately,
history. Therefore we are justified in starting from the proposition
presented just above.
In the chapters of Part I we shall discuss the following main issues: (1)
how we know unobservables through observables (chapters 2-5); (2) what
some of the properties are of the intelligence that are the consequence of
its empirical character (chapters 6-7); (3) the independence of observation
on theory, that is, it is not theory-laden (chapter 8); (4) the impossibility
of innate ideas (chapter 9). In Part II, we discuss ( 1) the relation of science
to its principles; (2) the relation of philosophy to its principles and to
experience and observation. We need not say that a number of ancillary
issues will be discussed along with the main problems.
6 From Observables to Unobservables
Notes
I. From a Logical Point ofView (Cambridge, Massachusetts: Harvard University
Press, 1953), p. 44. We wish to note here that Quine uses "myth" in the sense of
"fiction," and as we shall see later that is a distortion of its proper sense.
2. Prolegomenoa to Any Future Metaphysics The Library of Liberal Arts (New
York:The Bobbs-Merrill Company, Inc., I 950).
3.Marx W. Wartofsky, Conceptual Foundations of Scientific Thought (New York:
The Macmillan Company, 1968), p. I I 7.
1.
:I
Chapter 2
Observables-What they are
A. Introduction
That the factual has cognitive priority over what is not factual is
admitted in every systematic discipline (except perhaps mathematics,
because of the latter's special character), and that provides us with an
appropriate starting point for our considerations on observables. It is
generally admitted that the "raw material" of scientific activity consists
of regularities that provide the facts relevant to the discipline's
investigative considerations. Ultimately the aim of the investigations is
either to interpret the facts or to infer something from them, an aim that
plainly grounds the observation/inference distinction; or as we prefer to
say, it grounds the observable/unobservable distinction. Stating the point
another way, every scientific, systematic discipline takes pains both to
discover its problems in the facts and to measure its solutions by them.
Thus it seems appropriate to begin our considerations by reviewing what
is meant by the word "fact."
through facts. Electrical currents are known through their heating and
mechanical movements; magnetic fields are recognized through
deflections of compass needles and other similar motions; a disturbance
in a planet's orbital revolution tells us of an unseen source of gravity; a
fast pulse indicates excitement or perhaps an infection; insulting speech
a hostile attitude or prejudice, and so it goes. The point, then, seems
evident enough: observables provide evidence for unobservables, and that
is what we mean when we say that unobservables are known through their
observable manifestations.
Only for the sake of completeness need we mention that the natural
and social sciences are not the only enterprises to use observables for
manifesting unobservables. Imaginative literature uses metaphors and
symbols made from them to tell about human plots and emotions; myths
employ images to represent unsystematized ideas about the origins of
things; parables tell about life beyond the domain of experience. In sum,
the human mind universally erects its unobservable edifices on observable
foundations.
But not all talk about unobservables is regarded as respectable.
Philosophy, it seems, tends to specialize in unobservables, sometimes not
making much sense, so that early in this century some philosophers
denounced a large part of the philosophic enterprise as fictitious.
Metaphysics, they told us, was nonsense; and they were inspired largely
by the unfounded, unverifiable statements that some practitioners of
philosophy made. And though at present this view is not widely accepted,
it does signal a difficulty that comes from philosophy's preoccupation
with unobservables, especially those removed from observable
manifestations. It behooves philosophers, therefore, to pay very careful
attention to their empirical starting points, as well as to the procedures by
which they argue from them. More particularly, the empirical starting
points of philosophy are obtained from words that signify either
observable properties or things endowed with such properties. But our
point needs expansion:
'I I
I, It was not an empty remark RudolfCamap made some years ago when
he wrote the following words:
ll1
I
In its mythological use the word [God] has a clear meaning. It, or parallel
words in other languages, is sometimes used to denote physical beings
which are enthroned on Mount Olympus, in Heaven or in Hades, and
which are endowed with power, wisdom, goodness and happiness to a
greater or lesser extent. Sometimes the word also refers to spiritual beings
Observables-What They Are 13
We do not intend to bring back all that logical positivism endorsed, but
it is important to stress that the lines cited above do make a fundamental
point. No attempt will be made here to give an exegesis of Camap's
positions; instead we shall note only that what he says above is
legitimately interpreted to correspond with the regularity which we have
been at pains to describe, namely, that reflection shows us to know
unobservables through observables. Given that regularity, one has reason
to say that the name "God" is truly meaningless if it is divested of every
and any reference however indirect to observable manifestations. His
name cannot refer to something that is in all ways beyond experience; but
this is not to say that observation of physical manifestations is always
adequate to showing the full character of an unobservable entity, even of
many physical causes, much less God. Indeed our knowledge of
unobservables suffers limitations, a point about which more will be said
later.
We may now consider our point to be adequately established, namely,
that we know unobservables through the medium of observables, as a
result of which we may ask a very fundamental question: what is an
observable?
E. A definition of observable
. II'
II
I
or taste, etc., since the senses do not give rise within themselves to
movements that go from one property to another; or to put it another way,
they do not form propositions. There is, then, nothing in the least
"problematic" about the distinction between observable and unobservable,
and the failure to keep it in mind cannot but produce confusion. (The
principle of distinction employed above obviously starts from the
proposition that different passivities require different agencies to actuate
or actualize them. An ability to be magnetized must be actualized by an
actually existing magnetic field, which is to say that ability and agency
must be proportioned to one another. 10)
F. Asummary
Notes
7. For those who may not be familiar with this point, see Sir Arthur Eddington,
New Pathways in Science, (Cambridge: At the University Press, 194 7), c. I,
section III.
8. Hereafter we shall call that which is attained by any operational power its
"object"-that will be the meaning of the term.
9. It is important to note that an observable property-color, say-must not be
confused with an emotional stimulus. In one person the olor red may recall the
very pleasing scent of roses, whereas in another it may recall the horror of are
different and hence not the result of the color alone.
10. This principal blood from a serious wound. Clearly the emotional responses
e will be discussed at length in a later chapter.
1'
--~------------------------------------------------------ ---
I"
.\.
j
;.
Chapter 3
The Fundamental Cognitive Relation
A. The precise formulation of the problem
We may not say that stones are blind or that trees are unskilled, although
we may say that stones are non-sighted and trees are non-skilled. Thus the
negative term in privative opposition can be said only of those things that
ordinarily are endowed with that which the affirmative term signifies at
a time when the latter can be expected to be present. We do not consider
newly born puppies to be blind, for they are not mature enough to
exercise the operation of seeing. Here again, then, differences in reality
are responsible for the differences of signification between contradictory
and privative opposition.
There is another kind of opposition in which the terms are opposed as
extremes within a category of positive realities. For instance, to succeed
and to fail divide a common category as extremes, for both success and
failure are kinds of activity; one cannot fail if he does not try. Kind and
cruel are opposites of this sort too, since a cruel act is not the same as no
act. Such opposites, because both are within the same genus of reality, are
related as the perfect to the imperfect; that is, one always lacks something
of that which the other possesses, which means that the differences that
define each are related as affirmative and negative. One difference will
remove what the other posits; one opposite will explicitly deny a part of
what is contained in the notion of the other. Hard and soft are examples
of this sort of opposition, for a soft body differs from a hard one in
offering minimal resistance to a penetrating agent yet the soft body does
not lack all resistance to such an agent. Opposition of this sort is called
"contrary opposition," and like the others, it, too, reflects differences in
realities.
The last kind of opposition involves pairs of terms of which one refers
to the other in its understanding or definition and both of which are
positive in character; for instance, teacher implies student and father or
mother implies child. Similarly, whole implies part and ruler the ruled.
Now although both of these terms are positive, pairs of terms of this
description are not said of the same thing at the same time or in the same
respect and for that reason are included under the notion of opposition;
but it is important to emphasize that neither term is negative, neither
implies the absence of the other reality but instead requires its
Eimultaneous existence, each in its own entity. 2
Now the point of importance in this discussion of opposites is that
knowing one of the two precedes knowing the other. To illustrate: the
mind's knowing non-red, non-spherical, darkness, and silence depends
upon its first knowing red, spherical, light, and sound. Even when one of
24 From Observables to Unobservables
B. An unexpressed premiss
equilibrium, we could not, as with the magnet, tell which was cause and
which effect. Thus, to know one reality as cause and another as effect
requires that the mind go beyond observation.
The relations involved in location, bodily position, and time are
perhaps less clearly unobservable, and in the preceding chapter we did
classify them secondarily as common sensibles. But when we locate a
body, we do so by comparing it to another body that serves as a point of
reference. For instance, when we say that the chair is next to the table, the
tree is north of the house, the lamp is on the desk, etc., we describe one
body by comparing it to another to which we give some temporary or
permanent priority; and the reference of the one to the other is the
relation. Furthermore, because the relation requires certain real (and in
this case observable) properties and proportions in the things themselves,
it clearly is not a fiction, though it clearly is a "minimal reality."
Bodily position is like location in regard to the relations it involves. If
we were to say that a man or a horse or some other organism is sitting or
standing or lying down and then were to describe one or the other of these
positions, we could not escape referring some parts of the organism's
body to others. (A perfectly homogeneous body cannot have the
characteristic of bodily position as do organisms with clearly
differentiated appendages.) The sizes and shapes of the parts are
observable, of course, and they are the foundations of the relations of
position. Nonetheless, the relations themselves do not variegate the
sensory response. Similar remarks can be made about time, which because
it involves measurement, implies a relation of measure to measured. But
some arguments will help make our case.
The reader will probably concede that relations are indivisible, for we
cannot take away from them some part or fraction; either the "whole"
relation exists or nothing. For instance, we cannot take away a part of the
relation next to or north of Yet if we consider a visual field made up of
a chair next to a table, the visual field can be reduced by a part. If the
chair is removed, only a part of the field is gone, but in removing the chair
the entire relation of next to has been done away with. Certainly we
cannot explain the relation's removal by claiming that the whole of it
belongs to the chair; for that would mean the table contributed nothing,
which is clearly not the case. For the moment, however, let us assume that
the relations mentioned above are observable, and if they are, they can be
expected to be visible. The sense of sight attains all the common
sensibles, and that is what such a relation must be assumed to be. 5 It then
The Fundamental Cognitive Relation 27
follows that when two objects are placed next to one another, the relation
they acquire is a new observable property which must affect the senses as
long as the objects are next to one another; and when they are not next to
one another, then the observable property ceases to be. Stated more
precisely, since the relation of next to belongs to both objects and not to
one alone, each related object must affect the sense differently than it does
when one is not next to the other.
Let us now assume that the objects possessing the next to relation are
a table and a chair and that a screen is put in front of the chair so that we
no longer see it. The screen does not destroy the relation; therefore it
continues to exist. But if someone unseen by us removes the chair, the
relation will be destroyed; and thereafter the table should affect our visual
sense differently than it did when the chair was next to it, assuming, as we
did, that the relation is observable. But plainly the table does not affect
our sense differently, and observation will not tell us whether the chair
has been removed or not; in this respect the relation is not different from
that of cause-effect. So however observable the shapes and dimensions
upon which spatial relations are founded, we have to say that the relations
themselves are not. Again, a visual or other sensory field may contain a
multiplicity of observables that are foundations for relations, but the
relations themselves do not affect the senses. On the other hand, location,
position, and time are relations that are so close to observation by reason
of the sensible properties upon which they are founded that we often
speak of them as ifthey were observable. So it would seem that although
the linguistic convention ought to be recognized, such things as time,
position, and location are not observable in the strict sense of the term.
We have argued at length in another essay6 that, contrary to Hume,
substance is real and that it does indeed function as a substratum for
properties, though it is certainly not an "unknowable somewhat." We must
concede, however, that substances are unobservable, which means that
they too must be known through observables. Words such as "gas,"
"mineral," "fluid," "water," "gold," "electron," etc., signify stuffs or
substances; and when we distinguish one from the other, we do so through
their observable manifestations, that is, through their observable
properties or effects. From a distance a jar may appear to contain water,
and that might be what the stuff is; yet upon coming closer and smelling
its odor- and tasting its flavor-which also is not the stuff, we discover
an odor and a flavor that we do not associate with water but with alcohol.
Hence, if water is one stuff and alcohol another, we see very readily that
28 From Observables to Unobservables
In the strictest sense of the term only certain properties are observable;
yet a broader sense is commonly employed, and we must be careful to
take note of it. Ordinarily we say that salt and iron and other stuffs are
observable, and we speak that way because the stuffs have observable
properties. Similarly, we say that cats and dogs are observable things,
though strictly speaking only their properties are sensed. We do not truly
see a cat or a dog; so there is, then, an extended use of the term
"observable" or "sensible" according to which any thing or stuff that has
observable properties is itself said to be observable, and the extended
sense provides an economical way of speaking. Only a pedant would
insist on the strict sense of the term in ordinary conversation. Nonetheless
it is imperative to stress that this usage does indeed involve an extension
of the terms "observable" and "sensible" that must be kept in mind when
one discusses substances in a technical way.
There is still another way in which we use "observable." We say, for
instance, that we have observed the growth of an animal or plant, or that
we saw someone sneer, smirk, or do something similar; and these are
cases in which "observed" or "saw" truly signifies an observable property
(increased dimensions in the case of growth, shape of the mouth in the
cases of sneer and smirk). Yet in addition to the observable properties, the
words "sneer," "smirk," etc., signify relations which the properties have
to unobservable, interior causes). For that reason the meanings of these
words and others like them are complex in the sense that the words
signify both an observable and a relation of the observable to its
unobservable cause. Terms of this sort are many, and that leads us to
30 From Observables to Unobservables
another point.
When we speak of apprehending sneers, smirks, and the like, we often
use the word "perceive, "7 which commonly signifies something grasped
that is over and beyond the observable itself. Whenever we consider an
observable and something known through an observable we have to
employ a number of cognitive capacities, and we have to order what we
apprehend through them. Thus we must take care not to confuse a union
that comes from an order with a unity that is absolute and implies the
activity of only one operational power. Such a confusion can lead to our
assigning to sensory powers that which does not truly belong to them,
which seems to us to be the main reason for the difficulties about
unobservables.
E. Recapitulation
Notes
!.Of course the particles themselves were not observed, but measurable effects
were.
2. Currently "opposed" is used improperly to mean "distinct from," as in "He
watched the game as opposed to working on his taxes," rather than "He watched
the game instead of working on his taxes." It is true that the two activities cannot
occur simultaneously and so in that respect they embody the principle of non-
contradiction. But they do lack the determinate relation that opposition taken more
The Fundamental Cognitive Relation 31
B. Analogies
When we are close to someone striking a nail, our seeing and hearing
the blow appear simultaneous to us and do not reveal any difference in the
time at which we see the impact and hear its sound. But when we stand at
some distance we first see the hammer strike the nail and only several
seconds later do we hear its impact. On such occasions we are aware of
a difference in the time of our apprehension of the two sensible properties,
a difference that prompts us to infer that sound takes time to travel. We
also infer (having excluded other alternatives) that the propagation of the
sound depends upon an unobserved movement of air which is between us
and the object struck.
But a question arises about the sort of movement that is responsible for
sound transmission; and because the physicist cannot infer the nature of
the movement directly from the difference in time or from the character
of the sound, he is obliged to construct an hypothesis which postulates the
nature of the propagation. Furthermore, to form his hypothesis, he must
make use of sensory impressions he already possesses, as well as
whatever intellectual understanding he has already acquired from them.
Using these starting points, he then proposes that sound is a wave motion.
If we now ask from what the notion of wave is first obtained, we
obviously must say from the waves of a lake or the sea, which are the
waves most familiar to us and which are part of everyone's ordinary
experience. (Of course, other observable waves might provide starting
points, for example, the oscillation of a plucked string.) Then the physicist
compares this familiar motion to the transmission of sound, with a view
to modifying or adapting his original conception of wave to this quite
different phenomenon. In doing so he first notes the characteristics of
observable waves that the unobservable sound propagation must be
assumed to resemble, namely, the medium is disturbed, the disturbance
has a periodicity, a speed of propagation, etc. Having done that, he
The Similarity-Dissimilarity Relation 35
postulates the differences that must exist in order to account for what is
special to sound, for example, that its waves are not propagated in a plane
but in three directions from the point of origin. In short, the characteristics
which the wave motion of a body of water has that sound propagation
cannot have must be denied of sound, and so negations play an essential
role in understanding the nature of these waves.
Modem physics first conceived light, too, to be propagated by a wave
motion; and the relation of our understanding of both sound and light
waves to ordinary experience is brought out well by Christian Huygens
in the passage below: 2
We know that sound is sent out in all directions through the medium of
the air, a substance invisible and impalpable, by means of a motion that is
communicated successively from one part of the air to the next; and as this
movement has the same speed in all directions, it must form spherical
surfaces that keep enlarging until at last they strike the ear. Now there can
be no doubt that light likewise reaches us from a luminous substance
through some motion caused in the matter lying in the intervening
space-for we have seen above that this cannot take place through
transmission of matter from one place to another.
If, moreover, light requires time for its passage-a matter we shall discuss
in a moment-it will then follow that this movement is caused in the
substance gradually, and therefore is transmitted, like sound, by surfaces
and spherical waves. I call these "waves" because of their likeness to those
formed when one throws a pebble into water, which are examples of
gradual propagation in circles, although from a different cause and on a
plane surface. 3
From this passage we see that the points we have already made about
how we go from an observable to a description or definition of an
unobservable are confirmed by the actual historical record as outlined by
Huygens. That record provides us an illustration ofhow we spontaneously
use observables to come to an understanding of what a certain unobser-
vable is; and in this case there is no cause-effect relation between the
observable water waves and the unobservables sound and light waves that
are defined through them.
That light is more difficult to come to understand than sound is easily
36 From Observables to Unobservables
Unaided sensation reveals two kinds ofbodies, the first of which are those
that are lighted and are visible by themselves. Such bodies are said to be
"luminescent," and ordinary language simply calls them "lights." Physical
theory then extends the word "light" to name that which travels, that
which is propagated from the luminescent body, and plainly that is an
extended sense of the term.
The second kind ofbodies, those that are not visible by themselves, are
not luminescent, and in order to be visible they must be in the more or less
proximate presence of a light. Instead of luminescence, such bodies have
a colored (or colorless) surface by means of which they are made visible.
Just how the lighted body makes the colored body visible the physicist
must explain, and it is here that we see how little data he has. Ordinary
observation shows us that a body which is not visible by itself can be seen
only if it is in the presence of another body that is luminescent. Inside a
windowless closet, no non-luminescent body is visible; but introduce a
light, and it can be seen, from which we infer (instantaneously, of course)
that the luminescent body somehow makes the non-luminescent body
visible by acting on it. And to our inference that light is active we must
add some others:
Now experience shows that these things lighted may stand to each other
in varied relations; in attempting to reduce these relations to order and
understandability we make a certain invention. This is prompted by
several cardinal experimental facts: in the first place, things lighted have
a simple geometrical relation to each other, in that screens placed on
-
.........
....- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
straight lines between the lighted objects may suppress the illumination of
one or the other and themselves become illuminated. This leads to the
concept of rectilinear beams of light, which is no more than a description
of the geometrical relation between lighted objects. Then we have the
experimental fact of the asymmetrical relation of the lighted objects,
described in terms of sources and sinks. Finally, we have the discovery
made at a much later stage, and not possible until physical measurements
had reached a high refinement, that light has properties analogous to the
velocity of material things. This was first discovered in connection with
astronomical phenomena in the shift of the time of eclipse of Jupiter's
satellites and in aberration, but was later found to hold for purely
terrestrial phenomena, in that a beam of light reflected from a distant
mirror does not return to the source until after the lapse of a time interval
that can be measured with means sufficiently refined. 5
When it comes to saying that light travels, the physicist depends upon a
correlation of observables, on the basis of which he must fashion a
satisfactory account of the nature of light propagation, making use of
ordinary experience; and it is not hard to see that the analogies he
employs are founded on a similarity-dissimilarity relationship between an
observable and an unobservable that are not related as cause and effect.
Additional testimony to the value of analogies is provided by the
physicist's use of fluids flowing in pipes to manifest something about
electrical currents, and here, too, the observable is not the cause of the
unobservable:
Regarding the text cited, let us note that the order which the authors
followed in manifesting their point is an order the mind must employ.
That is to say, the mind starts by noting the similarity between an
unobservable and an observable and then goes from there to the
differences between the two. Given the origin of human knowledge from
sensible traits, the mind can do nothing else.
Analogies, Nagel tells us, are widely employed. 9 If, however, someone
argues that analogies are helpful but not necessary, we wish to say 1) that
because no one can escape comparing like to unlike, and 2) because
everyone must know unobservables through observables, it follows that
analogies are not only helpful but are indispensable. To read the nature of
an unobservable from an observable, we have no alternative but to use the
similarity/dissimilarity relation.
There is reason to note that the similarity relation is also used on those
occasions when we can infer from an observable effect something about
the nature of its unobservable cause. For instance, if we notice that the
moon is eclipsed, we infer from the shadow-the effect-that an opaque,
spherical body has been interposed between the moon and the sun.
However, in other cases that are unlike the eclipse, a likeness can be so
remote, so extended, so tenuous that we speak of the more manifest as
only "suggesting" the nature of another. One ought not always to think
image or copy when he hears "like" or "similar," especially in a systematic
context, for the unobservables of nature ordinarily are not images or
The Similarity-Dissimilarity Relation 39
The reason is that our minds start from sense experience, and therefore
sensible things are the first to be intellectually apprehended. But all
sensible things have magnitude, hence the point and the unit can only be
negatively defined. For the same reason whatever transcends these
sensibles can be known by us only negatively ....
The same is true of things that are known through their opposites, for
example evil or black, both of which are related to their opposites as
privations. One of two contraries is always imperfect or privative in
relation to the other. He [Aristotle] then adds, as though by a reply, that
the intelligence in a certain way knows each of these through its opposite,
namely evil through good and black through white. 11
D. A recapitulation
Notes
1. Principle is more generic than cause; for every cause is a principle, but not
every principle is a cause. Principle has the notion of a beginning; and a journey,
for example, has a beginning, but its beginning is not its cause. Similarly, a point
is a beginning of a line, but it is not a cause of the line.
2. We may ignore for our purposes here the particle-like properties light has been
discovered to have.
3. The text cited is from "The Wave theory of Light," in Classics of Modern
Science, ed. William S. Knickerbocker (Boston: Beacon Press 1962), p. 54.
4. The Logic ofModern Physics, (New York: The Macmillan Company, 1927),
p. 150.
5. Percy Bridgman, ibid., p. 151.
6. Eric M. Rogers, Physics for the Inquiring Mind (New Jersey: Princeton
University Press, 1960), p. 150.
7. Arthur J. Vander, James K. Sherman, Dorothy S. Luciano, Human Physiology,
(New York: McGraw-Hill Book Company, 1970), p. 103.
The Similarity-Dissimilarity Relation 41
8. Ernest Nagel, The Structure ofScience, (New York: Harcourt, Brace & World,
Inc. 1961), p. I08.
9. As the reader knows, "model" is another name either for an observable or for
a prototype invented by the imagination with the aid ofthe mind and put together
from the sensory data of ordinary experience.
IO. When "sign" is taken strictly, it excludes "image," "symbol," and "likeness."
II. Commentarium in Aristotelis Librum de Anima (Rome: Marietti, 1948), Lib.
III, teet. II, nn. 757-9.
Chapter 5
More on Analogies and Metaphors
A. Our knowledge of internal states
Mont.[the numbers that follow each name tell the age] (7;0 months): "You
know what it means to think? _ Yes _ Then think of your house.
What do you think with? _ The mouth. _ Can you think with the
mouth shut? _ No. _ With the eyes shut? _ Yes _ With the
ears stopped up? _ Yes. _ Now shut your mouth and think of your
house. Are you thinking?_ Yes _ What did you think with? _
The mouth.
... we have traced three distinct stages, the first of which is easily
distinguishable from the other two and appears to contain a purely
spontaneous element. During this stage children believe that thinking is
"with the mouth." Thought is identified with the voice. Nothing takes
place either in the head or in the body. Naturally, thought is confused with
the things themselves, in the sense that the word is a part of the thing.
There is nothing subjective in the act of thinking. The average age for
children of this stage is 6. 1
The responses reproduced above are not isolated, Piaget tells us, but
are indeed representative of the young, which must not be taken to
suggest that we can all obtain the same sort of responses. On the contrary,
Piaget maintains that considerable skill is required to make sure the child
is responding to the questions according to his own mind and not
according to what he has picked up from his environment, nor according
to what he thinks the interrogator might like. But when investigations are
carefully made, the results do indicate that children fail to recognize that
their internal states are distinct from the objects known through those
states. Children do not understand internal states as internal states; their
minds require time to develop sufficiently to make distinctions. So given
the priority of observables over unobservables, we are hardly astonished
to find that children do not at first distinguish thought itself from thought's
sensible manifestation-speech-and the mind from the mouth's
organ-the mouth.
That in our understanding of observation we should find the object to
have priority over the activity itself becomes more intelligible when we
consider the principle at issue, a principle illustrated by how we
distinguish one artistic activity from another. We would instruct anyone
not knowing the difference between painting, say, and sculpturing by
pointing out to him the works each art produces. As a painting itself
differs from a piece of sculpture, so the art of painting (the mental quality)
differs from the art of sculpturing. Thus we distinguish activities from one
More on Analogies and Metaphors 45
another through their objects, and the same is true of the operations that
are passive in relation to their stimulus-objects; for were we in doubt
about whether an organism had a sense of hearing or a sense of sight, we
would settle the question by exposing the organism to the object-stimuli.
In sum, experience of our own behavior shows us that we know the
external realities which are the objects of our activities before we know
the activities themselves, which is a fundamental trait of an empirical
intelligence. Though we are a bit ahead of ourselves, we wish to say that
ultimately everything the mind knows, even about itself, it knows through
objects that the senses apprehend.
Focusing now on the results ofPiaget's investigations of children, we
may say that only after a time do we come to see that our interior states
or activities are distinct from the objects they attain; yet knowing what the
internal states are is even more dependent upon external realities, a point
we consider central and one we shall now attempt to show.
what the word "activity" signifies. Were the sun the only body in
existence, it would be hot, yet it would do no heating. The sun would also
possess a gravitational mass, but it would do no attracting. In sum,
convention imposes the words "activity" and "action" upon the physical
modification in one body that is brought about by a quality or motion in
another; that is the primary meaning of these words.
In contrast, our activity oflooking at the moon, for example, does not
modify the moon nor any of its properties; nor does anything happen to
the piano when we listen to it. Even when we feel warmth or wetness or
softness or heaviness, the touching sensation itself does not modify the
object that is touched. Though a physical contact (and hence a physical
modification) is necessary for tactile sensations to occur, the contact is not
the sensation or operation, but the condition of the operation's exercise.
We could repeat our descriptions in the paragraphs above using "state"
in place of "action," "activity," or "function"; and in so doing, we would
begin with a term that is more general as well as equivocal, which would
require that we separate those states that are correlated with physical
modifications from those states that are not so correlated. Thereafter,
however, the procedure would be similar. And though "action," "activity,"
and "operation" appear the better terms, we shall not make an issue of
conventions. Whether one uses "state" or "activity," the notions signified
are the same. However, as Aristotle pointed out long ago, the activity we
call "sensation" is not the same as the kinds we have used as illustrations
above. With our external sensory apparatus we cannot detect other
people's (or our own) sensations; for plainly the movements of eyeballs,
fingers, etc., although necessary, are not the sensations themselves, which
are "psychic activities"; and the fact that sensory operations bear such a
name suggests that they differ from physical actions in important ways.
Still, physical and psychic activity do resemble one another; for we know
from eliciting and controlling our sensory operations that they require
both a "doing," a "sensing," as well as an object toward which the doing
or sensing is directed.
We may now come to the point: understanding sensation as an activity
obliges us to describe it as lacking something inherent in physical action:
psychic actions do not modifY the objects toward which they are directed.
Thus we must introduce a negative element into the definition of such
sensory states or actions because they differ from the actions that we first
know: those that have observable effects. To recapitulate: we define
sensation first by naming it an "activity" because of its similarity to
More on Analogies and Metaphors 47
Words are human artifacts, meaningless save as our associating them with
experience endows them with meaning. The word "swarm" is initially
meaningful to us through association with such experiences as that of a
hovering swarm of gnats, or a swarm of dust motes in a shaft of sunlight.
When we extend the word to desks and the like, we are engaged in
drawing an analogy between swarms ordinarily so-called, on the one hand,
and desks, and so forth, on the other. The word "molecule" is then given
meaning derivatively: having conceived of desks analogically as swarms,
we imagine molecules as the things the desks are swarms of. 3
In his own words Nagel confirms the central points we have made above,
but we must add two qualifications. First, words that are no longer used
metaphorically do not "lose" their original meanings. On the contrary, the
original meanings must be retained if the secondary meanings are to be
clearly understood without a substitute meaning being supplied at the
moment. We have argued that what actually happens is that the word
ceases to be a metaphor through acquiring another literal meaning.
Second, the use of metaphors is not only a "pervasive talent," it is also a
necessary one with which we are not able to dispense. Let us elaborate.
Because we must know unobservables through the medium of
observables, our initial transference of the name of an observable property
or entity to something that is unobservable has to be a metaphor. For
instance, the first step the physicist takes towards providing a category
term for sound propagation is to call it a "wave"; yet as long as he rests
there, as long as he goes no further in his description (his understanding),
his predication of wave is metaphorical; for up to this point he
understands only a likeness that has not yet been circumscribed.
Moreover, as the reader knows, metaphor has often been defined as a
compressed simile, which indicates its character very well; and the first
More on Analogies and Metaphors 49
step the physicist takes amounts to saying that sound propagation is like
a wave. That there may be no noticeable time lapse before he moves from
understanding sound as a likeness to understanding it as distinct in certain
respects does not affect the issue; for until he (mentally) sees that sound
propagation differs in certain determinate respects, he has not gone from
the metaphor to the analogy.
Whenever we use a metaphor, we assign a name without assigning the
definition to that to which the name is joined. For instance, the statemenJ
"John is a moose," attaches the name but not the defmition or description
of moose as a four-legged, antlered mammal, etc., to John. Consequently,
the plain intent of the metaphor is to tell us that John is like a moose.
Similarly, when we first call sound a "wave," we are saying that it is like
a wave; and in both the case of the moose and the case of the wave, we do
not consider the named entities with respect to what is proper to them. (Of
course, "moose," unlike many metaphors, will probably not acquire a
literal meaning in relation to men. Furthermore, metaphors can be used
when there is no attempt to delineate the proper nature of the thing
described, as is usual in imaginative literature.) So in summary, whenever
an unobservable receives its first description from an observable, whether
internal or external, we have done no more than present an uncircum-
scribed likeness. Not until we have assigned a distinguishing characteristic
that separates the unobservable from the observable does the name receive
an extended literal sense. Metaphors, then, are an essential element in
human cognition. Stated another way, the human intelligence must begin
with what many call the "concrete."
The use of such mechanical models, very far from facilitating the
understanding of a theory by a French reader, requires him, in many cases,
to make a serious effort to grasp the operation of what is often a very
complicated apparatus, as described to him by the English author. Quite
an effort is required in order to recognize the analogies between the
properties of this apparatus and the propositions ofthe theory that is being
illustrated. This effort is often much greater than the one the Frenchman
needs to make in order to understand in its purity the abstract theory
which it is claimed the model embodies. 5
The connection between theory and reality was formerly always conceived
of as though the symbols occurring in the laws of nature represented
simple magnitudes, or quantities, which could either be immediately
perceived, or could at least be regarded as being of the same nature as such
magnitudes or quantities (e.g., a length of Ill OOm). Thus, in Newtonian
mechanics, the fundamental concepts represented by lines in space, time,
and mass, were three terms of which the meaning seemed to be derived
immediately from sensory imagery. All three are combined in the concept
of motion, which is equivalent to temporal change in the spatial position
of a mass. Motion is that process in which the basic requirement for
knowledge appears to be fulfilled in a pictorial manner -namely, as the
perception of the constant element in change. That which is moved-the
52 From Observables to Unobservables
The point to be made from the citation above is that the mechanical view,
the representation of unobservables wholly in terms of observables, is an
inclination we must be careful to discipline, otherwise we seriously distort
the intellectual enterprise.
E. A recapitulation
... expresses a revolt against an approach to philosophy that takes its point
I I of departure from crystallized beliefs and theories handed down by a
tradition which only too often perpetuates preconceptions and prejudg-
II ments. This negative part, the identification and deliberate elimination of
!
I I theoretical constructs and symbolisms in favor ofthe return to the unadul-
'.
terated phenomena, is by no means a simple and easy affair; it takes a
determined effort to undo the effect of habitual patterns of thought and to
return to the pristine innocence of first seeing.
One has to agree that the goal phenomenology sets for itself is the
right one, and the foregoing words do appear to say that phenomenology
has abandoned the enterprise of examining to the exclusion of everything
else, what is within the cognitive powers. There are, however, serious
doubts as to whether it has; and since it claims to find its origins in real
things, it will be beneficial to examine its claims for the purpose of
throwing more light on what is genuinely an empirical origin of
knowledge.
Spiegelberg tells us that the phenomenological method has seven
major steps, each of which has to be considered if one is to understand the
nature of the method. The first of the steps is called "investigating
particular phenomena," and the first sub-consideration under it is
"phenomenological intuiting." After some descriptive comments and an
apology for the term "intuiting," Spiegelberg turns to the example of
force, which he uses throughout his discussion to illustrate the points at
issue. Having paid attention to some confusions to be avoided and some
restrictions on the use of the term "force," Spiegelberg then says the
following:
The focus of his text is clear in the sense that he claims that
phenomenology is concerned with force (and other such realities) insofar
as they issue from or affect us and so are the object of our internal
awareness. In other words, phenomenology does not address phenomena
that are external realities taken as such; instead it addresses phenomena
according to their mental or emotional appearances. Stated another way,
it addresses phenomena that are "experiences" insofar as the term signifies
an awareness rather than the reality on which the awareness bears. Thus,
More on Analogies and Metaphors 55
The point we wish to make from what is said above is that we possess
a number of cognitive powers that are distinct and often difficult to
distinguish from one another, which means that unless we are careful we
can make erroneous statements about what pertains to external
observation. Often without realizing what we do, we assign what internal
cognitive powers know to the external object, and in so doing we go
beyond the immediate external observables, which is a confusing state of
affairs, to say the least. (Later, in a chapter on theory-laden observation,
we shall have more to say about this sort of thing.) There is, however, one
more point we wish to make about phenomenology, a point about its aim
of bracketing existence, of abstracting from existence and leaving it in
abeyance.
According to Spiegelberg the "minimum meaning" of the "phenome-
nological reduction" can be stated as follows:
By way of comment let us say frrst that all of our ordinary knowledge
as well as that which is systematic begins with an existing world. We start
with entities known to exist, and we examine them; hence it is impossible
to abstract from their existence except in a qualified sense of the word.
Could anyone truly think that a newly born child's first concepts are of
possible realities? It is true that to the extent our systematic investigations
consider regularities, to the extent they consider classes and what is
common to many individuals, to that extent our investigations abstract
from the existence of the individuals themselves but not from the general
existence of the species. When a biologist investigates sharks, for
More on Analogies and Metaphors 57
Notes
I. The Child's Conception of the World, trans. Joan and Andrew Tomlinson
(Totowa, New Jersey: Littlefield, Adams & Co., 1965), pp. 39-40.
2. Although our description of sensation is accurate as far as it goes, we do not
pretend that the definition we have outlined above is complete. Enough has been
given, however, to make our point.
3. W. V. 0. Quine, "Posits and Reality," reprinted in Theories and Observation
in Science, ed. Richard E. Grandy (Englewood Cliffs, N.J., Prentice-Hall, Inc.,
1973 ), section II.
4. Op. cit., p. 107.
5. The text that follows is from The Aim and Structure ofPhysical Theory (New
York: Atheneum, 1962), pp. 66-71.
6. From Philosophy of Nature, trans. Amethe von Zepplin (New York:
Philosophical Library, 1949.), p. 27.
7. The Phenomenological Movement (The Hague: Martinus Nijhoff, 1960.) The
quotations we shall present here are all from Volume II, Part 5, Section XIV, The
"Essentials of the Phenomenological Method."
8. Ibid., pp. 656-8.
9. Ibid., p. 662.
10. Ibid., p. 691.
Chapter 6
Signs, Symbols and Words
A. Introduction
Having discussed the two relations on the basis of which the mind
moves from an observable to an unobservable, we must talk now about
signs and symbols as properties and as manifestations of this mode of
knowing, for we all recognize that the function of signs and symbols is to
direct the mind to something not directly and immediately apprehended
by it. To begin properly, however, we must first point to some
representative instances of signs and symbols.
Many things that we observe function as signs: black clouds tell of a
coming storm, leaves turning color forecast an approaching winter, the
smell of smoke warns of a fire, a sharp tone of voice declares the boss's
irritation. Physicians note the symptoms (a synonym for "signs") of their
patients and diagnose their diseases. Psychologists observe the abnormal
behavior of their clients and read their emotional disturbances, and so
both physicists and psychologists make use of signs. More prosaic affairs
depend on signs too, for all of us in the course of performing our daily
tasks read multitudes of signs. Yet not everything one observes is a sign.
The grass in the lawn, the bush next to the house, the squirrel in the yard,
the automobile parked by the curb-all these ordinarily have no signifying
function. Now all of this is familiar, for in essence we are repeating what
we said earlier when we gave evidence to show that we know
unobservables through observables. The repetition is desirable, however,
because we shall say something about how artefactual signs are made. 1
First, then, let us note that the genus or category according to which
sign must be defined is observable. Moreover, we must emphasize that the
process of making artefactual signs (and symbols) will have to imitate, if
the device produced is to perform properly, the mode according to which
the intelligence knows unobservables through observables. Symbols of
60 From Observables to Unobservables
course have something in common with signs, and the two names are
often applied to the same objects. We hear man described as a "symbolic
animal" or a "symbol-using animal," more often perhaps than we hear him
described as a "sign-using animal." But we ought not to be blind to the
differences between "sign" and "symbol" when the things to which they
are applied differ. "Symbol," for instance, is applied to artefacts but rarely
to things in nature; so one important use of"symbol," is to indicate certain
signs of human manufacture.
Other words, too, are employed as more or less synonymous with
"sign," the more common of them being "symptom," "signal," "token,"
and "indicator." Such a plurality of nearly synonymous names should tell
the reader that although the word "sign" may be used in preference to the
others, the definition of sign, not its name, must be our primary concern.
As we have already seen, signs, in the strict sense of the word, are
observable realities; they are not images in the sense or concepts in the
mind, as some of the textbooks say. Furthermore, not everything, not even
all organisms, make use of signs-only men and certain animals, a point
that need not be labored. It is also clear that signs are starting points,
never end points, of a cognitive process from which the mind moves to
the knowledge of other things; and so only an extended use of the word
would permit one to apply "sign" to an unobservable, for an unobservable
can never be an absolutely first starting point. Once again, then, a sign is
by nature an observable reality through which the mind comes to know
something else, something not present to the senses, whether that which
is not present is in principle unobservable, or whether it is unobservable
only because of some circumstance.
Photographs, paintings, statues, and the like are of course not called
signs, though they certainly direct the mind to unobservable realities; and
the reason they are not signs is that they are all images. Convention does
not permit us to call a likeness a "sign" of the absent object it resembles.
As soon as one compares the things that convention does call "signs" to
things that are likenesses, he sees the difference between a sign and what
is not a sign, even though the non-sign may have the function of directing
the mind's attention to something that is unobservable. Briefly: a sign is
"qualitatively" distinct from that which it signifies, whereas a likeness, as
the name tells us, requires some measure, however remote, of qualitative
"! Signs, Symbols and Words 61
~'
t..
:i
similarity, and when the likeness directs the mind to something beyond
the reach of sensation, it often goes by the name of "symbol." Thus not
only does "symbol" suggest human manufacture, it also-even though it
can substitute for "sign"-suggests that in symbols some measure of
likeness is ordinarily the source oflight for the mind. 3 Literary images and
metaphorical devices are said to have "symbolic meaning," of which
parables and myths are perhaps the primary instances. 4 Let us add, too,
that a word such as "signal" usually signifies some practical activity to be
performed, a use to which "symbol" is not often put. 5 Having said those
things, we are now in a position to see that sign is to be defined according
to the following formula: a sign is a sensible reality that by reason of the
impression it makes on a sensory power draws the mind's attention to
something different from the sensed object, something that is not itself
able to produce a sensory impression (at least at the moment). As so
defined, a sign seems to be any observable that points to an unobservable,
but that is not quite right, for we ordinarily do not use "sign" and
"observable" coextensively. We do not say, for instance, that a moving
compass needle is a sign of a magnetic field, nor do we say that a shadow
on the moon is a sign of an eclipse. The reason of course is that these are
per se effects and they point directly to the cause. Signs, however, are
usually incidental-not per se-effects. In a somewhat similar fashion, an
image such as a photograph or painting leads the mind to an unobservable
(at the moment at least) reality, but it is not called a symbol because the
likeness is an image. Thus just as a per se effect is not a sign, so an image
is not a symbol. But speaking more generally, symbols, to the extent that
they are more remote likenesses of that to which they point, obviously
differ from signs in not being incidental effects of the unobservable to
which they point. Insofar as a symbol is a likeness but not an image it
does not have a cause-effect relation of the sort typical of signs.
From what has been said we now know that signs can be classified
according to the two different sources from which their signifying
function is derived. Some are of human origin, having been brought about
by the human intelligence and will, and others are independent of human
causal activity and function as signs whether we want them to or not. The
former may be called "arbitrary" signs to denote their human origin,
whereas the latter are obviously "natural." Black clouds, smoke, animal
tracks, glacier marks, etc., are natural signs, while traffic signals, many
trademarks, and most of all language, are "arbitrary."
The distinction between the two kinds can be approached, however,
62 From Observables to Unobservables
C. Language as a sign
not be mistaken; for our aim here is not to question whether purely
arbitrary signs can be made or whether they are in any way useful. On the
contrary, our intention is merely to note that if in making a sign one starts
with something that has a natural function or established signification
already proportioned in some measure to the new signification to be
assigned, he will achieve superior results. In other words, an act of sign
making that modifies a regularity already possessing a meaning makes use
of the natural mode of the mind. On those grounds some artefactual signs
can be judged to be better or worse than others.
In this connection let us note that according to some scripture scholars,
the account of the origin of the universe in the Book of Genesis consists
of stories that are found in already existing myths about creation. If that
is true, it certainly would illustrate the principle we defended just above,
for the account in Genesis does not merely repeat these other stories but
modifies them according to their new intention, their new signification.
The advantage of this sort of origin of Genesis lies in its starting with
something familiar and commonly known and then undertakes to modify
it, thereby telling its audience that it has a new signification. 8
A more general illustration of the use of observables that occurs in
Scripture, this time an illustration of knowing a cause through its effects,
was given in a lecture by Dominique Barthelemy at the University of
Freiburg in Switzerland. Commenting on the Hebrew name of God that
is translated as "I am who I am," he remarked that although this is a very
legitimate rendering, there is another that is more fundamental and that
also renders the Hebrew accurately and is translated as "I intervene as I
intervene." (Plainly this refers to events that were not to be explained as
natural phenomena and stem from natural laws; on the contrary, their
special character as observables pointed to something beyond the
regularities of nature. In general that is the function of miracles: they are
sensible manifestations of events or phenomena that transcend the
ordinary processes of nature.) This understanding of the Hebrew
expression certainly accords with God's calling himself The God of
Abraham, Isaac, and Jacob, for he intervened in their lives, as the
Hebrews knew, and so he identified himself through those interventions.
That was the "name" he gave them.
The point of these illustrations is that in both of them the ordinary
mode of the human intelligence is observed. Genesis, if composed as
described, takes accounts of creation already familiar to many peoples,
including the Hebrews, and modifies them to signify something new; and
tl.
D. Imitative art
The reader undoubtedly knows that the critics of the imitation theory
of art understand the theory to endorse the view that works of (fine) art
have as their function the copying of nature. An imitation, they tacitly
assume, is a copy, which means that the theory implies the artistic ideal
to be a photograph of the reality portrayed. As a consequence, those who
support, for example, an emotive theory do so by attacking the obvious
weaknesses of such a view; but they misunderstand the position they
decry.
Once again we must draw attention to our lack of expertise in regard
to things artistic, yet we must say something about the imitative theory of
art precisely because its proper (and traditional) understanding entails the
empirical mode of the intelligence. When the Greeks said that the function
of the fine arts is to imitate nature, they did not mean that art should
attempt to produce copies. If that were its function, the artistic
representation would be as a source of light inferior to nature itself. But
even one who is only casually familiar with such things knows that art is
in some way an "improvement" on nature, at least insofar as art addresses
the human intelligence, though it also makes use of the human affective
response. This is not to say that art will strike the mind or affect the
emotions with more vehemence than nature; on the contrary, that often
may not be the case and is not to be expected. Traditionally, art was held
to imitate nature when the artist modified that which nature presented in
order to make some aspect of the object more manifest and so more
intelligible to the viewer. For instance, the function of a portrait painter
would be to show some principal quality or qualities of the character of
the person represented, thereby making them more manifest than they are
in the person himself or herself. To be effective the artist would have to
subdue some observable traits and highlight others; he would "interpret"
nature, portraying something of it in such a way that the viewer could see
and understand nature's works better through the artistic object than he
66 From Observables to Unobservables
Repeating what has been said above, signs are observables that point
to unobservables; and when the intelligence is itself a cause of signs, it
does well to follow the natural order of the intelligence by modifying
something that already has a public signification or by modifying some
regularity; that is the proper way to introduce the new signification. And
our point that signs ought not to be purely arbitrary is what Plato had said
before us in the "Cratylus," where he argues that signs are not well made
when they are purely arbitrary And this same point is implicit in the
words of a philosopher who was a logical positivist and who died some
years ago. We are speaking of Rudolf Camap, who claimed that
metaphysics is nonsense because it comes from what we might call the
"misuse" of words. Camap says that "A word which ... has a meaning ..
.is said to designate a concept," whereas words that are peculiar to
"metaphysics" designate pseudo-concepts, which is but another way of
saying that they are meaningless. 9 To illustrate his point he quotes from
an author who was his contemporary, but instead of using Camap's
illustration, I would like to provide one of my own choosing but not of
my manufacture, an illustration that in fact comes from an abstract of a
paper given at a philosophical convention:
Nonsense and pseudo-concepts: that is the right description, and one finds
it difficult to understand how anyone could offer those words as a
description of what he planned to say. Now pseudo-concepts, Camap
claims, are not concepts at all but words that at one time had a meaning
which later they lost. Originally they signified either an observable or
something related to an observable, but they lost that connection when
they fell into the hands of "metaphysicians," continuing nevertheless to
exist as familiar conventional sounds. That is why they have only the
appearance of meaning and are pseudo-concepts. If words have no
connection to concepts, then they also have no connection to reality and
hence have no meaning. To be sure, the same word need not be a
pseudo-concept in all contexts, for its user may have found the word's
original sense and extended it in a very legitimate way.
Were the human intelligence to function in a Kantian, a priori fashion,
then we would be hard pressed to say how a pseudo-concept could come
about; yet once we see how the meanings of words originate in
observation, we readily understand that Camap is correct. Many words
have indeed become meaningless; and much jargon-especially
bureaucratic jargon-is, despite its often mundane subject matter,
"metaphysics" in Camap's sense. Only because it pretends to deal with
familiar affairs through familiar sounds does such jargon differ from the
more pretentious utterances of some academicians; and too often the
bureaucrats do not have clear ideas of what they are saying. But to make
our point more evident let us look again at the several meanings of words
that are related to one primary sense, which has to be the source of any
clarity secondary meanings have.
directly, then they must signify realities indirectly, through the concepts.
Thus Camap correctly used word as his category-term for pseudo-concept
because the word was all that remained of what once was a concept. Yet
when Camap used the term "metaphysics," though he had certain people
in mind, 10 he did declare that not everything done under the title of
"metaphysics" throughout the history of philosophy warranted being
called nonsense. Also we wish to note that the word "metaphysics" is no
longer pejorative in sense as it was when the positivists held sway. But
leaving the term aside for the moment, let us illustrate what we have said
above by looking at the several meanings of words and their relations to
one another. And though there will be some repetition of what we have
said, we think it useful for our considerations on the order of naming.
We need only consult the dictionary to see that most words have
several meanings that ordinarily are related to one another. "Shape," for
instance, means first of all the contour, the configuration, the figure of a
body; and then it is extended from this first sense to mean more generally
a visible form or semblance of a type. In a still further extension, "shape"
comes to stand for any representation, even a phantom, after which it is
once more extended to mean a form of embodiment, as in words. And
there are yet more extensions: form of thought, mode of existence of
something ("the first shape of the writing"); yet even these do not bring
us to the end of the listings.
Another example is the word "form," which in the original Latin had
significations much like those of"shape." (In English "form" has actually
retained its first Latin sense, which is shape or likeness.) Now here, too,
the extensions are many, an important one being the ideal or intrinsic
character of anything, following which come manner or method, conduct
regulated by custom, orderly arrangement, and most generally, mode of
existence. The words "shape" and "form" remind us of "posture," which
first signifies a bodily configuration, and then is extended to mean a
military policy, as in "our defense posture." In each of these illustrations
the extensions are made on the basis of a similarity the second reality
signified has to the first, even though the similarity may in some instances
be remote.
In chapter 4 we talked about "health" and "healthy" as well as
"cheerful," for they are words that have original impositions that differ
from the examples above because both of them refer in their first sense to
something we detect directly by internal experience. When we are ill or
indisposed in some way, we are usually aware of that indisposition-our
Signs, Symbols and Words 69
lack of health-by how we feel, despite our not knowing exactly what
health is. The word "health" is then extended in the forms of "healthy" or
"healthful" when it is applied to food, medicine, and complexion. "Cheer"
and "cheerful" are like "health" and "healthy," signifying, as they do, a
mood we are in or not in that is known through internal observation. Later
these terms, too, are extended to mean cause of health or cheer, as well as
sign or expression of health or cheer, which can be read, for instance,
urine in the case of health and in the face or voice in the case of cheer. In
these examples the many meanings are related to one that is primary on
the basis of a cause or principle or sign relation. Of course not all words
with several meanings have senses that are related but instead are purely
equivocal; for example, "pen," which signifies a writing instrument and
a small enclosure. That, however, does not concern us here.
The reason for the relation among meanings is not hard to see. If
words signify concepts directly and realities only through concepts, then
the order according to which words signify realities is not the absolute
order the realities have in the world, but the order they have as they are
known by us. If we derive our knowledge of unobservables from
observables, then the order of our concepts is not a one to one mapping
on the real world made according to the world's own relations. We cannot
claim that the mode of presentation in the intelligence is a copy of the
mode of occurrence in the world. So to repeat: the order according to
which we acquire concepts through observables determines the order
according to which words acquire their many related meanings. 11 Thus
realities are knowable by us according to their proximity to or distance
from observation. Furthermore, it follows from what we have said that a
reality which has no direct or indirect relation whatsoever to observable
traits cannot be known by us, as Carnap correctly maintained. We have no
warrant to speak about such "realities," and they present a grand occasion
for self-deception. (In passing we might add that what we have just said
indicates why probative arguments for God's existence present such
difficulties. The intelligence is obliged to start from observables and then
move to the object that is at an infinite distance from observation.)
Earlier we noted that Plato tells us in the "Cratylus" that "names ought
to be given according to a natural process ... " and as we have seen, the
natural process begins with an imposition of the name upon an
observable, after which the extensions follow the order in which concepts
are derived. If this order is followed, the first meaning of the word throws
light on those that are extended, and it does so either on the basis of a
cause-effect relation (in a broad sense of "cause") or on the basis of
similarities and dissimilarities. But as Plato's word "ought" says, the
natural order is not obligatory in a rigid way; in some measure, though not
totally, it can be circumvented.
Despite what is natural and ideal, unobservables can be labeled with
words or symbols that are made specifically for them, not exhibiting in
their own first signification a connection with observation. Still, the
natural order cannot be totally circumvented, for words or symbols with
no observable first imposition can be attached to unobservables only by
a procedure that indirectly makes use of observation. The attachments can
be made in several ways: ( 1) by a definition which itself employs words
or symbols with meanings originating in observation; (2) by a context
fully enough elaborated to make the significatum of the name stand out
through familiar and properly established references, a procedure that
requires at some point words which have first impositions taken from
observation; (3) by equating the new name with an established synonym.
Although such indirect attachments are more obscure than those that
result from naturally extending the name of an observable, they
nonetheless allow the attached word to perform its function. Now it might
seem that in this last paragraph we have been speaking primarily of
mathematical symbols, and though they are instances of what we are
describing, they are not the only ones. A certain class of words-a class
that originates in another language-is less easily recognized than are
symbols as being obscure, so they make good candidates for becoming
pseudo-concepts. But let us tum again to Camap for an illustration.
After having quoted certain passages from a "metaphysician," 12 Camap
singles out the word "principle" as having no meaning in the context
which he cites; and in our judgment he is correct. The author who is
quoted does not define the term "principle," nor does the context allow a
meaning to be assigned by comparing the reality signified (none can be
identified) to something observable; nor does the context provide an
equivalent synonym. If the author has anything at all in mind, he certainly
has given insufficient evidence of it. Still, the word "principle" is trouble-
Signs, Symbols and Words 71
I. Etymologies
are accurate on their own plane, they are not equivalent to nor replace-
ments for defmitions that are systematic or scientific. Still, such first
defmitions need not be "corrected."
governance (Lord), his providence (Providence), etc. 18 All the names are
imposed upon one and the same entity, though they are taken from
different effects or activities, and they are synonyms only by reason of
that which they name and not by reason of the sources of the names.
K. Inflected languages
mistake.
Another remark seems to be in order. Given the order among the
meanings of words and the character of language as an instrument for
signifying the conceptions of the mind, it is evident that the failure to
discipline the young in their formative years to regard language with care
can be and has been mentally crippling. The first analytic training the
mind receives is in grammatical analysis, and that activity is certainly not
"purely verbal," if anything at all meets such a description. Furthermore,
the value of inflected languages in the formation of young minds (Latin
and Greek in particular) can hardly be exaggerated. A careful
understanding of the conventional regularities of language and their
careful application to a text, subject the mind to the basic disciplines
necessary for its own well being. Grammar involves the mind in the use
of principles, however imperfect they may be as principles because of
their imperfect regularity. However, before we can analyze arguments and
extract them from contexts, we must first be able to understand the
relation of the parts of the sentences in which the arguments are
expressed; and to understand the things signified, we must first
understand the linguistic order; failing the latter, the mind is left in
confusion.
Notes
Chapter 6 dwelt mainly on signs, and so now leaving them behind, let
us return to metaphors, to likenesses, this time for the sake of looking in
a general way at myth and its role in forming the mind. As soon as we see
that knowing unobservables requires the mind to employ metaphors as a
first step ultimately leading to describing or defming unobservables, then
we understand in a general way the cause of the mind's need for literary
images, figures, symbols, etc., all of which we shall include under the
general title of metaphor. We see that we are obliged to treat realities
beyond observation first through the vehicle of the imagination, whose
raw materials are observables, and whose function is not only to retain but
primarily to create. Since unobservables differ from the observables
through which they are known, the creative function of the imagination
is especially necessary for an introduction to what we conceive about
unobservable realities. This is especially true when no systematic
procedure is yet available, or when such an investigation is not
appropriate. Every society-no matter how scientific or advanced its
systematic enterprises-must base its first understanding of the
unobservables important for human life on the constructions of the
imagination. In short, epic poems, dramas, fables, folk stories, and the like
are indispensable, especially, as we have said, for the first formation of
the mind. 1 But our aim here is to consider in a special way the use of myth
as a necessary prelude to systematic investigation; myth and science ought
not to be thought of as opposed but as complementary.
according to which names are attached and the need to know a word's
first, sensory imposition. But we must also reflect to some extent on the
use of words in a metaphorical sense, a sense that depends on a relation
of similarity. Realities beyond observation we must first treat-on the way
to knowing what they are-through the vehicle of the imagination, whose
raw materials are observables, and whose function is not only to retain but
primarily to create. Since unobservables differ from the observables
through which they are known, the creative function of the imagination
is especially necessary for an introduction to what we conceive about
unobservable realities. This is especially true when no systematic
procedure is yet available, or when such an investigation is not
appropriate. Every society-no matter how scientific or advanced its
systematic enterprises-must base its first understanding of the
unobservables important for human life on the constructions of the
imagination. In short, epic poems, dramas, fables, folk stories, and the like
are indispensable, especially, as we have said, for the first formation of
the mind.
One of the more important uses of metaphor occurs in the construction
of myths, which ordinarily provide images of what are seen as causes of
natural events and things. We wish, therefore, to talk about the role of
myth in human learning. More particularly, we wish to consider the use
of myth as a necessary prelude to systematic investigation; myth and
science ought not to be thought of as opposed but as complementary.
The Ugaritic myths explain nature so as to satisfy man's craving for the
answers to the universe, and to guarantee the regularity of the processes that
84 From Observables to Unobservables
... the excesses of the nature myth school have meant, if anything, that it
has been too little considered in any serious way. 7
Whether one takes the lines "The myth can be seen as a symbolic
representation of the interplay between rain and soil that makes plants
come to life and grow," to mean that the mythical gods represent the
hidden causal efficacies of rain and earth themselves, or whether the
mythical gods are taken as causes extrinsic to nature-in either case the
myths represent unobservable causes and are symbolic representations of
them.
Kirk calls the second theory of myths that he describes "aetiological,"
so naming it because
... it implies that all myths offer a cause or explanation of something in the
real world. 9 The first proponent of this theory, Andrew Lang, is described
as holding that. .. many myths are clearly not about nature; he [Lang] was
arguing that even those that are, are more than just pretty allegorical
conceits that are explanatory in some way. 10
This second theory would seem to differ from the first by its explicit
endorsement of myths as representations of realities exterior both to nature
and to human society, realities which are gods that cause entities in nature
and have effects on human society.
The third theory Kirk describes is one he calls a "charter theory," first
proposed by B. Malinowski, who said that myths
... the definition of myth that seems least inadequate because most
embracing is this: Myth narrates a sacred history; it relates an event that
took place in primordial Time, the fabled time ofthe "beginnings." In other
words, myth tells how, through the deeds of Supernatural Beings, a reality
came into existence, be it the whole of reality, the Cosmos, or only a
fragment of reality c an island, a species of plant, a particular kind of
human behavior, an institution. Myth, then, is always an account of a
"creation"; it relates how something was produced, began to be. 13
Although Kirk presents this as a distinct theory, there is little doubt that
it is causal; and the words of Eliade quoted above need no comment from
us to make the point. Moreover, any additional ceremonial or ritualistic
role the myth might play in "reenactments" of "primordial events" or the
"creative era" does not destroy the myth's character as a metaphorical
representation of unobservable causes. One may well expect ceremonies
and rituals to depend upon something causal as that which is being "cele-
brated," as some say.
The Necessity of Myth 87
The fifth "monolithic theory" Kirk describes maintains that "All myths
are closely associated with rituals. "14 This statement by itself does not say
much; for if we reflect we see that it does not propose an opposed theory,
a theory incompatible with the others. Myths could be aetiological, natural,
etc., and still be associated (as Kirk says) with rituals. But in a more
specific form the "ritual theory" maintains that myths "are actually derived
from rituals." 15 A proponent, however, puts the issue more strongly:
D. A recapitulation
understands that artefacts are produced by men, without the child's relating
the functions of artefacts to structure and materials, so adults, early in the
generation of a culture, argue by analogy to gods as the causes of things,
without ftrst seeking to explain properties and behavior through the causes
within the things themselves.
It is interesting that the number of deities myths postulate does not seem
to be altogether a haphazard affair. Allowing for corruptions introduced
into the original stories by the entertainers, the evidence suggests that at
least in some mythologies a separate deity appears for every kind of natural
thing or phenomenon that has no observable cause. There is a god to
account for the sea, one to account for the rain, one for thunder, one for
volcanoes, and still others for plants, animals, and the remaining categories
of natural entities. Such a correlation is of course what we should
anticipate.
Aristotle is reported in his later years to have "returned to the myths" in
I\ his reflections on the ftrst cause, and we perhaps ought not to be surprised
because the indetermination of the metaphors that constitute myth allow
the mind to search for likenesses, for similarities that can enlighten the
I I
'~
',
mind on the most remote and inaccessible of all causes, namely God.
Although Aristotle argued very well to the existence of God and
understood his nature as well as one might from the evidence of nature,
most of his philosophical successors did not climb to the same heights.
When one looks at Cicero's De Natura Deorum he sees that those who talk
about God or gods describe him or them through physical attributes. This
work of Cicero's 18 presents the theological views of the three schools of
I r philosophy that were au courant in his day, and they were the Epicureans,
the Stoics, and the Academics. The following quotation illustrates what we
have just said:
For Jet us hear Plato, that divine philosopher, for ... He holds that motion
is of two sorts, one spontaneous, the other derived from without; and that
which moves of itself spontaneously is more divine than that which has
motion imparted to it by some force not its own. The former kind of motion
he deems to reside only in the soul, which he considers to be the only
source and origin of motion. Hence, since aU motion springs from the
world-heat, and since that heat moves spontaneously and not by any
impulse from something else, it fo11ows that heat is soul; which proves that
the world is an animate being.
. . . the world possesses inte11igence ... and if this be so, it fo11ows that the
world must be endowed with wisdom, for, if it were not, man, although a
90 From Observables to Unobservables
. . . But the fourth and highest grade is that of beings born by nature good
and wise, and endowed from the outset with the innate attributes of right
reason and consistency; this must be held to be above the level of man: it
is the attribute of god, that is, of the world, which must needs possess that
perfect and absolute reason of which I spoke. 19
... If again it [the being which embraces all things] be capable of reason
yet has not been wise from the beginning, the world must be in a worse
condition than mankind; for a man can become wise, but if in all the
eternity of past time the world has been foolish, obviously it will never
attain wisdom; and so it will be inferior to man. Which is absurd. Therefore
the world must be deemed to have been wise from the beginning, and
divine. 20
The point of these quotations is simply this: the human mind finds it
extraordinarily difficult to understand the nature of any substantial entity
that is beyond the physical world. The arguments for God's existence
present a maximum difficulty, as was said by Thomas Aquinas, and the
debates of which the passages are representative provide evidence of that
difficulty. Our times possess the revelations of Scripture in which God is
described as existing outside the world and his fundamental attributes are
made known to us. When, however, one has to establish from natural
evidence that God exists and what some of his attributes are, the mind
rarely gets beyond the physical. Perhaps that is why in our own day some
of the environmentalists consider nature herself to be god. Now back to
myths.
... that we ought not to intrude into the mysteries of government which the
Prince intends to keep secret: The tempests and confusions rais'd by the
loosing the winds, represent the mischiefs and disorders that arise from
such a vain curiosity in the subject .... But whatever judgment is pass'd
upon this explication, it is certainly an instance of the ill consequences of
avarice, and an unseasonable curiosity. 24
Pope's last sentence shows the confused (in the sense of indeterminate)
character of metaphors. Because Bossu starts from the bag of winds and
moves to a specific kind of evil in political societies, his explanation is
more determinate than his principle allows. On that account Pope feels
obliged to point out that if we do not admit Bossu's explanation by reason
of its not issuing determinately from the metaphor, we must nonetheless
concede that the image does draw our attention to the evil consequences of
avarice and undue curiosity, which is a much more general interpretation.
Myths, then, if they are well constructed, ought to be regarded that way:
the more indeterminate the interpretation of the images, the more certain
the myth is; the more determinate the interpretation, the more open it is to
debate. Furthermore we often see that several determinate interpretations
are compatible with the same images, a state of affairs that does not involve
any difficulty.
92 From Observables to Unobservables
... since the Egyptians had as much common sense as we have ourselves,
we may conclude with certainty that no one, except perhaps a very
unsophisticated mind, took the composite picture ofthe heavenly cow at its
face valueP
To say the Egyptian and other early peoples were less intelligent than we
is very unreasonable. To say that they viewed their mythical symbols
literally is like saying Columbus sailed west to prove that the earth is not
flat. Before his day the earth's spherical shape had been known for
centuries.
In the usual meaning of the term, "abstract" (an intellectual cut) refers
to a state of affairs one fmds in systematic considerations. The latter are
concerned with general propositions that can be cast in universal form even
when they are not true always but only most of the time. We also abstract
or make a cut when we classify things and properties in categories, for the
mind must slough off differences to obtain broader classes. Abstraction
also occurs when we leave aside irrelevant traits and events in favor of
The Necessity of Myth 93
those that are relevant to the issue being considered. In short, the ordinary
senses of the term refer to systematic considerations and do not describe
what goes on in imaginative works. We may rightly say, then, that abstract
thinking and metaphor are mutually incompatible. Of course even in
literature the mind cannot avoid using general or class terms that leave
aside individual traits; so everything is abstract in that minimal sense. Still,
there is a way in which literary works ought not to be abstract; and that is
what we would like to illustrate.
When we look at the lines below from Homer's Iliad-by way of
Alexander Pope-we see described the internal state of Agamemnon the
night after Achilles refused to return to the army, and the quoted lines
illustrate the point we wish to make about abstraction:
It must not on this account be taken as proved that Dr. Proudie was a man
of great mental powers, or even of much capacity for business, for such
94 From Observables to Unobservables
qualities had not been required in him. In the arrangement of those church
reforms with which he was connected, the ideas and original conception of
the work to be done were generally furnished by the liberal statesmen of the
day, and the labour of the details was borne by officials of lower rank. It
was, however, thought expedient that the name of some clergyman should
appear in such matters, and as Dr. Proudie had become known as a
tolerating divine, great use of this sort was made of his name. If he did not
do much active good, he never did any harm; he was amenable to those
who were really in authority and, at the sittings of the various boards to
which he belonged, maintained a kind of dignity which had its value.
Let us say at once that we are not about to enter into a discussion of the
relative merits of epic poems and novels, or any other sort of literature.
Quite apart from our not being qualified for such a task, our aim is much
more modest: we merely wish to make a few comments about the abstract
and the concrete as they reflect the mind's greater or lesser distance from
its origins in observation.
Despite the basic, abstract character of intellectual processes that we
have mentioned, the Iliad though intellectual is a concrete work. That
becomes especially plain when the selection we quoted from the Iliad is
compared to the selection from Trollop. We saw in the passage above that
Homer aims to describe Agamemnon's affective state, and he does so
through the kind of observables from which such internal states are
ordinarily known. For instance, we all experience "rowlings (rollings) in
the breast," and we also experience their emotional causes. Similarly,
"sighs following sighs ...bursting from Atrides' breast" are observables
which convey a clear message to everyone. In short, Homer successfully
focuses the listener's attention upon sensible data which give rise to
inferences the listener or reader can hardly avoid. The poet makes his point
by presenting external and internal observables that produce the conclusion
he seeks, thereby causing his own conceptions to become those of the
listener or reader. Trollop, on the other hand, does not do the same thing;
The Necessity of Myth 95
Notes
I. All that we have done in our considerations on observables implies that the use
of the imagination solely for the purpose of stimulating emotions without regard
to their subordination to and control by the intelligence is a kind of disorder. An
emotional state by itself can hardly be the primary aim of education or
investigation. The current tendency in the arts to make the stimulation of emotion
the artistic goal is clearly a kind of anti-intellectualism not in the best interests of
human life and culture.
2. Mythology, a Mentor Book (New York: The New American Library, 1942),
p. 19.
3. Mircea Eliade, Myth and Reality, Harper Torchbook (New York: Harper &
Row, Publishers, 1963), p. 5.
4. "Canaanite Mythology," in Mythologies ofthe Ancient World, ed. Samuel Noah
Kramer, Anchor Books (Garden City, New York: Doubleday & Company, Inc.,
1963), p. 183.
5. See G.S. Kirk, The Nature ofGreek Myths (New York: Penguin Books, 1974),
c. 3.
6. Loc. cit.
7. Loc. cit.
8. Loc. cit., p. 46.
9. p. 5.
10. Ibid.
I I. Ibid. p. 59.
12. Ibid., p. 63
13. Myth and Reality, HarperTorchbook (New York: Harper & Row, Publishers),
1963, p. 5.
14. Op. cit., p. 66.
15. Ibid.
16. Robert Graves, The Greek Myths (Baltimore: Penguin Books, 1955), Vol. I,
p. 10.
17. We ought to note that Fustel de Coulange in The Ancient City maintains that
mythical figures first began as household gods. Family ancestors first came to be
honored and then were gradually transformed into family gods, who ultimately
came to be honored beyond the family of their origin. And though the interest of
Fustel de Coulange is in such gods, he does grant that gods of nature were also a
part of the culture. In a way, family gods are even more closely linked to
observation than gods of nature, for they originate from the conviction that
The Necessity of Myth 97
the intelligence, and the observations themselves are (at least in some
measure) a priori. Now if the intelligence is responsible even in part for
the content of observations, then it has to be independent of observables
for the origin of its concepts. Now in what follows we shall take as
representative of the view that observation is theory-laden a work of
Norwood Russell Hanson, 1 whose approach is more particular and more
empirical than most, and whose claims provide an opportunity to consider
a number of important points and difficulties. Also we shall briefly
consider some views of others.
B. Hanson's position
... saying that Kepler and Tycho see the same thing at dawn because their
eyes are similarly affected is an elementary mistake. There is a difference
between a physical state and a visual experience. Suppose, however, that
it is argued as above-that they see the same thing because they have the
same sense-datum experience ... .If this is argued, further difficulties
obtrude. 2
Fig. 1
Do we all see the same thing? Some will see a perspex cube viewed from
below. Others will see it from above. Still others see it as a kind of
polygonally-cut gem. Some see only criss-crossed lines in a plane. It may
be seen as a block of ice, and aquarium, a wire frame for a kite_ or any
of a number of other things. 3
Theory-Laden Observation 101
Though Hanson says we do not all see the same thing, he will not allow
us to explain our differences by saying that though the seeing is the same
the interpretations are different because
This sounds as ifi do two things, not one, when I see boxes and bicycles
[which means that] the concept of seeing which is natural in this
connection does not designate two diaphanous components, one optical
and the other interpretative. Fig. I is simply seen now as a box from
below, now as a cube from above; one does not first soak up an optical
pattern and then clamp an interpretation on it. 4
About another line drawing (not reproduced here) Hanson says that
although he sees a bear climbing a tree, someone else will not. He adds
that paintings are not only colors, plots are not merely details in stories,
tunes are not just notes because organization gives the elements of each
of these a shape or pattern, without which what we see would be
unintelligible:
Organization is not itself seen as are the lines and colors of a drawing. It
is not itself a line, shape, or a color. It is not an element in the visual field,
but rather the way in which elements are appreciated. Again, the plot is
not another detail in the story. Nor is the tune just one more note. Yet
without plots and tunes details and notes would not hang together.
Similarly the organization of Fig. III (not shown) is nothing that registers
on the retina along with other details. Yet it gives the lines and shapes a
pattern. Were this lacking we would be left with nothing but an
unintelligible configuration of lines. 5
. [which] perhaps constitutes their seeing the same thing. 6 both Tycho
and Kepler have a common visual experience of some sort ..
When he says that seeing is at least a visual copying and then goes on
to add that it is more than that, he tells us he is equivocating. "Seeing" has
an ordinary meaning according to which it has to do with "visual copying"
and refers solely to the sensation of sight; but although Hanson qualifies
his view by saying that "There is a sense . . . in which seeing is
theory-laden," he would make "seeing" signify something beyond what
it actually is and what the ordinary man understands by it. Given the
equivocation on "seeing" to mean at one time only the external sensory
act ofseeing and at another to mean the operations ofimagination and/or
memory as well, one might readily admit on the basis of such an
ambiguity that observation is "theory-laden." Moreover, if observation is
taken to include, in addition to sensation, the intellectual act of inferring
that y follows from x then for even stronger reasons observation is indeed
"theory-laden."
Other remarks he makes leave little room for doubting the
equivocation:
Without doubt Hanson thinks the intelligence puts something into the
content of sensation itself, and the reason appears to be his
misunderstanding of what is required for multiple operations. He seems
to think that if the operations were several, then they must be successive.
Simultaneously occurring, ordered operations do not appear to him to be
what takes place. Moreover, that he is inconsistent can hardly be denied
in the light of his saying that seeing "embraces the concept of visual
sensation and of knowledge," while maintaining at the same time that
there are not "two diaphanous components ... there are not two opera-
tions." But let us consider further what he has said.
First we should note that Hanson has not distinguished the several
cognitive capacities the human species has. That they are several would
hardly seem open to doubt, and that they operate together with a kind of
unity also seems not open to doubt. In fact, the close coordination of
activities is the reason why they are run together and why the distinction
of capacities is often overlooked. On the other hand, Hanson does
recognize the multiplicity, even though he wants to introduce the effect
of coordination into the act of seeing. Or more properly, he wants to
introduce the knowledge of the pattern, the order-which is something
known by the intelligence with the aid of other cognitive capacities-into
the operation of the senses.
But the failure to distinguish cognitive capacities is not peculiar to
104 From Observables to Unobservables
Hanson; for the British Empiricists as a group do the same thing. John
Locke, David Hume, and George Berkeley, for instance, identify the mind
with imagination, and usually they are not taken to task for their mistake.
We are hardly surprised, then, to see an error such as Hanson's passed
over. In the next chapter we shall, in association with a discussion of
innate ideas and the a priori, discuss how one goes about determining the
distinction and number of operational capacities that a species possesses.
Now, however, we shall limit our remarks to some particular points
Hanson has tried to make.
When Hanson says it is a mistake to claim that Brahe and Kepler see
the same thing on the grounds that their eyes are similarly affected, he
does not give a reason why. He says only that there is a difference
between them, and apparently that is supposed to establish his claim. Now
let us concede at once that there is a difference between physical and
psychic activity, but that concession does not entail our having to grant
that people do not see the same thing when they are affected by the same
object. Such an argument amounts to maintaining that because two things
differ in species they do not belong to the same genus or category, and the
fallacy is patent. Sensation is passive in a very determinate way, as we
were at pains in an earlier chapter to show; and insofar as it is passive in
the face of a stimulus, it has something in common with purely physical
activities, such as being heated, undergoing magnetization, etc. So just as
we may infer that several bodies proximate to a Bunsen burner will all be
heated by it and that the heating is the same in kind (though not all bodies
will absorb heat at the same rate), we may also infer that two people
exposed to the same visual stimulus will undergo the same kind of effect.
In short, the primary effect of a visual stimulus, for instance, is the same
in each person, though secondary effects such as emotional responses may
well be different. Thus when the agents and passive recipients are the
same, the effects brought about are the same in kind, though they are not
numerically -individually-the same. But now to another matter.
On considering Hanson's claim that in the line drawing introduced
earlier one might "see" a transparent cube either from above or from
below, we now understand that "seeing a cube" involves more than the
sense of sight. In the drawing, sight attains a two dimensional continuum;
that is what we actually see on the page. But a cube is a three dimensional
continuum; and to "see" a cube in the two dimensional drawing one must
simultaneously employ his imagination, which is where such "seeing"
actually takes place. Furthermore, since the imaginations of different
Theory-Laden Observation 105
people can do different things with the same initial observables, we ought
not to be astonished when they do not "see" the same thing.
Hanson also said that in the line drawing one man may see an ice cube,
another an aquarium, another a frame for a kite, etc. Ice we take to be a
substance, while an aquarium and a kite are artefactual things made of
glass and other materials that also are substances. The latter, however, are
not (in the strict sense) observable; only their properties are. We do not
see, feel, etc., water, glass, metals, and other substances; rather, we sense
their qualities of being wet, cool, mobile, hard, dry, etc., no one of which
is a stuff or substance. So because "ice cube" and "aquarium" and all other
such things include a substance or substances in their conceptions, we
must say that in the proper sense of the term-the one at issue here-they
are not observable; they do not themselves affect the senses; only their
properties do. As we said earlier, substances are understood by the
intelligence as a consequence of data provided by the senses, but the latter
are not themselves stimulated by the substance. If they were, we would
not err, for example, by confusing real gold with fool's gold. In short, we
do not see ice, an aquarium, etc., except insofar as "observable" or
"visible" is taken in the secondary sense of the term which includes in its
signification the intelligible stuff or substance in which the properties are
found. 11 If, then, we take "observable" in its secondary sense that includes
what only the intelligence attains-substance, for example-then we can
understand why Hanson would think that the intelligence does contribute
to observation. But as we have said, here more than sensation is involved.
Still another point of importance, one that we treated briefly earlier,
arises in connection with words that signify observables in relation to
unobservables, a point that we treated briefly in chapter 3 and which we
shall repeat here. It has to do with words that signify an observable along
with a relation of the observable to an unobservable. For instance, the
word "wrinkled" said of the brow is neutral in the sense that it signifies
only an observable configuration of the forehead; it has no unobservable
element. On the other hand, "smirk" is not neutral, for in addition to a
configuration of the lips, it signifies an unobservable internal state.
According to Webster's (2nd ed.), "to smirk" is "to smile in a conceited or
affected manner," and conceit is clearly an unobservable state or
disposition in the one in whom the smile occurs.
Many words are like that: "leer," "scowl," "chortle," "sarcasm,"
"groan," "scoff," and "scream" are only a few. "Groan," for instance, is
defined by Webster's as "a low moaning sound ... " and if the definition
106 From Observables to Unobservables
substantive thing or stuff; and once again different people might "see" or
"observe" different things. But now back to Hanson.
Some of the examples that Hanson adduced ran properties together
with the realities of which they are properties. Music, for example,
although it is not a substance, is more than just sound. Music is ordered
sound, and order is an unobservable relation (as Hanson admits) perceived
by the intelligence; we do not hear the relations. More generally, any
organization of colors, lines, etc., is a set of relations that is unobservable,
even though the relations may be founded on observables. On that account
we can say that a visual, auditory, or other observation involves an
"interpretation" whenever the observation is accompanied by a
simultaneous recognition of an unobservable element by the intelligence.
And though the recognition is not an interpretation in the sense of an
explanation, it does require the activity of the mind as a cognitive capacity
distinct from the senses. So when Hanson says that (interpretative)
"seeing" is not an interpretation, only the meaning of "interpretation" as
explanation would make that statement true.
That the mind is involved in many judgments about unobservable
entities which have observable properties is perhaps the most commonly
occurring reason for saying that observation is theory-laden. When we
drive our car down the street just after filling the gasoline tank and fmd
that it comes to a stop, we quickly infer that something is wrong. We
observe the cessation of sound and motion, and then without hesitation
infer the conclusion. The inference is trouble-free and instantaneous, and
so we are inclined not to notice it; nonetheless it is an intellectual and not
a sensory act. Similarly, the neophyte prospector who thinks he holds gold
in his hand has inferred the substance from the properties he detects, and
his mistake is accounted for by his "interpretation," that is, by the
inference his intelligence has made from inadequate observations. The
same can be said of such claims as "I see a cat on a roof," or "I see an
aquarium," etc. In each of these cases the mind concludes to a substantive
entity from observable properties; and the entity is not observable in the
primary sense of the term but only in its extended, secondary sense,
according to which "observation" is indeed "theory-laden." Nonetheless
in those cases the "theory" does not contribute to the content of the
sensation itself, rather, the "theory" consists in an inference from the
observable.
Before we continue, let us note that obviously observable and
observation-term are not the same. It is one thing to say that an
108 From Observables to Unobservables
C. Other voices
Thomas Kuhn, for instance, says that Aristotle and Galileo in looking at
a pendulum do not see the same thing 13 because Aristotle sees
"constrained motion," while Galileo sees "free fall." The reader will
recognize, however, that this is only another instance of an observable
named in relation to an explanatory cause; and so he has introduced an
element that is not part of the observation itself.
P. K. Feyerabend takes another position. 14 He quotes Galileo when the
latter deals with an argument saying that if the earth moved, then a stone
dropped from a tower would not appear to fall on a perpendicular line.
Galileo grants the appearances, namely, that the stone does move without
any observable deviation from the perpendicular; but he asks that we use
the power of reason to confirm or disconfirm the appearances. The role
of reason in Galileo's mind is supplied by the following illustration:
One may learn how easily anyone may be deceived by simple appearance,
or let us say by the impressions of one's senses. This event is the
appearance to those who travel along a street by night of being followed
along the eaves of the roofs. There it looks to them just as would a cat
really running along the tiles and putting them behind it; and appearance
which, if reason did not intervene, would only too obviously deceive the
senses. 15
and also be corrected by the mind, then the sensory capacities are the
locus of two different passivities, one regarding the stimulus, the other
regarding the mind. But such passivities would be of such radically
different orientations that we could hardly consider them to belong to the
same operational power. In other words, we would have two kinds of
"senses," one passive with respect to the activity of the observable
stimulus, the other passive with respect to the activity of the intelligence.
And just as the ability to be affected by gravity is different from the
ability to be affected by a magnetic field and the ability to be molded, etc.,
so the two abilities of the sense would also be distinct. But that the senses
should have two such passivities is not a tenable view; so now let us turn
to Feyerabend's own remarks.
About the illustration we quoted above he says:
To start with, we must become clear about the nature of the total
phenomenon: appearance plus statement. There are not two acts-one,
noticing a phenomenon; the other, expressing it with the help of the
appropriate statement-but only one [emphasis his], viz. saying in a certain
observational situation, "the moon is following me," or, "the stone is
falling straight down." We may, of course, abstractly subdivide this
process into parts, and we may also try to create a situation where
statement and phenomenon seem to be psychologically apart and waiting
to be related (this is rather difficult to achieve and is perhaps entirely
impossible). But under normal circumstances such a division does not
occur; describing a familiar situation is, for the speaker, an event in which
statement and phenomenon are firmly glued together. 17
There are not, Feyerabend maintains, two acts but only one when one says
in a certain observational situation that "the moon is following me," etc.
But how his claim is anything other than gratuitous is hard to see,
especially when a few paragraphs later he says (after granting what he
thinks is a doubtful assumption to the effect that sensations which enter
the body of science are independent of their linguistic expression):
their views in our context here. For instance, one might see W.V.O.
Quine, Theories and Things (Cambridge, Massachusetts: The Bellknap
Press of Harvard University, 1981), Chapter 2, entitled, "Empirical
Content." The reader would see that his discussion is in terms of
sentences, and that would entail much analysis before it could be made to
fit our essay. To analyze a position on observables, on properties of
realities, that would start from sentences would be lengthy, to say the
least.
One could also see W.H. Newton-Smith, The Rationality of Science
(Boston Massachusetts: Routledge & Kegan Paul), 1981. For similar
reasons this author too must be left aside, and so with this our
consideration of theory-laden observation is closed, and we shall now tum
to other matters. 19
Notes
16. We shall consider innate ideas and the a priori in a later chapter. Here we
content ourselves with noting that regularly men think themselves to know, to
some extent, things "out there" as they are "out there."
17. Op. cit., p. 72.
18. Ibid., p. 73.
Chapter 9
The A Priori and Innate Ideas
A. Introduction
to bend and retain the shape after the action of an external force; viscosity
is a property of stuffs that flow, one part moving in relation to the another,
the movement resulting from gravity or some other agency exerting an
influence on it. And so we could continue. The point, however, is evident
enough: these physical properties, as well as others we did not mention,
imply an ability to undergo the action of some agent cause.
But certain spontaneous movements of natural entities brought about
by fundamental natural forces are determinative of things in a way others
are not. For instance, gravity, whatever its nature, is an active source of
motion in everything that has mass. It also is the cause that determines the
formation, shape, movements, etc. of galaxies and their components, as
well as the orbits and movements of planets. Speaking generally, gravity
may be said to determine the relative position of one body in relation to
another within the universe. And we see that the earth's constituting
materials exhibit an order that resulted from gravity at a time when the
earth was in a molten stage. Our point in all this is simply that gravity is
the major determinant of the character of the physical universe insofar as
the latter is constituted by an arrangement or position of bodies or stuffs.
Ultimately the earth's order is an arrangement of elements and
compounds, which are substances that are complete in species. 2
Another spontaneous natural movement is the one that is brought
about by the electrostatic or electromagnetic force that binds charged
particles; and because all elements are composed of such particles, their
electrostatic forces can be thought of as the "glue" which holds things
together. Nuclear forces, of course, play an analogous role. But our point
is a general one: to each of these active causes there corresponds an
appropriate passive property. Gravity has no effect on something that has
no mass; electrostatic forces do not affect particles or bodies that lack a
charge; a magnetic force cannot affect something that has no ability to be
magnetized. In short, natural activities of the sorts we have just mentioned
require their own passive principles in the bodies they act upon, and
conversely, the passive principles require a corresponding active source
in order to be affected. Gravity cannot affect a charged body by reason of
its charge, and a charged body cannot affect some other body by reason
of the other's mass alone.
If we now consider living things, we find that they possess active
abilities or capacities, and these capacities have operations which are
directed to definite kinds of objects in a clear and fixed way. Now
operational powers or potentialities or capacities are active; they are
The A Priori and Innate Ideas 117
things by one mind and other things by a second mind; and again we
would be two persons, not just one. But our last remark requires
elaboration.
Some years ago the Russians were featured in certain popular
magazines for having grafted the head of one dog onto the body of
another, the grafted head functioning in some measure and deriving its
sustenance from the body of its host. Something similar occurs, of course,
when a calf is born with two heads. Furthermore, if in regard to the latter
we ask whether there are two calves or one (assuming both heads are
functioning), we have to reply that despite the obvious deficiencies, there
are two calves. The first criterion of individuality is the numerical
distinction of principal operations; so if there are two visual activities,
there are two organs, and if there are two acts ofhearing, again there are
two organs. More directly, if there are two distinct sets of principal
cognitive activities, then there are two dogs or calves. Similarly, if we had
two similar but numerically distinct intellectual actions we would be two
persons." To repeat, wherever the principal actions are numerically
multiplied, so are the entities, and were we in possession of two radically
distinct kinds of intelligence, we would be two radically distinct kinds of
person. It would seem, then, that the human mind is empirical and
unendowed with either a priori concepts or innate ideas, for none of us
has two or more intelligences or is more than one person.
Notes
I. We treat this issue at greater length in Substance and Modern Science, cc.
11-12.
2. Just as an arm is not complete in species because it is only a part of a
substance, it would seem that elementary particles, too, in a somewhat different
way but for even stronger reason, are not complete in species. That is why we
speak here primarily about the effect of gravity on elements and compounds.
Our remarks do not deny the reality of pulsars, black holes, and other esoteric
entities astrophysicists discuss, entities which are not composed of matter as
we know it here on earth; nor do our remarks deny the effect of gravity on
particles existing separately.
3. These notions, too, are discussed in S&MS, c. 23.
4. Descartes is not talking about a supernatural infusion of knowledge which
presupposes a natural mode of knowing as already established; that is a
different matter and not at all the issue here.
Chapter 10
Regularities, Experience and Experiment
A. Introduction: a note on systematic considerations.
That is the problem: how do we establish the truth of the premisses, which
themselves are not deduced from other propositions. That is the principal
issue before us in Part II, and we shall begin our discussion with a
discussion of regularities, which play a foundational role in every
systematic consideration that pretends to explain the world of nature.
Moreover, the regularities play different roles in science and philosophy,
and the differences are very important when it comes to the question of
premisses or starting points. And so to them we now tum.
122 Observables and Unobservables
B. Concepts
common sense laws; 3 and although the statement tells about a change that
occurs in the human (and other) species, it is not "behavior" in the usual
sense of the word, despite its obvious reference to a change that happens
to men.
The illustrations above all indicate the use of "law" to signify an
empirically obtained regularity that is obscure and requires an argument
to show why the law is so; in other words, a law of this sort becomes a
conclusion of an explanatory argument. But there is another use of "law"
according to which it signifies a regularity (usually behavioral) that is a
first principle and on that account an explanatory premiss, examples of
which are the law of universal gravitation and the laws of
thermodynamics. To be sure, most of the "law-premisses" or
"law-principles" are hypotheses, but that does not affect our point; for
empirical regularities that are not assumptions and serve as argumentative
starting points are also laws in this sense of the term.
In the ordinary speech from which its scientific use is derived "law"
signifies a legal directive or proscription that regulates human behavior
and is a directive issued by civil authority. Still, a civil law does have
something of the character of a proposition insofar as it communicates
what the citizens ought to do, or insofar as it represents the de facto state
of affairs in some society. Formally, however, a law is a directive that
obliges citizens to initiate or inhibit certain kinds of behavior under
specific circumstances. Hence the ordinary sense of the term must be
distinguished from the scientific, according to which "law" signifies a
regularity which we understand but which is not a directive we issue.
There are, however some caveats.
Although laws are expressions of regularities that can carry universal
quantifiers, we must be careful not to say that every proposition with a
universal quantifier is a law; for appearances can be deceiving. As will be
plain to many readers, if we were to say "All of the men in this room have
pneumonia," or "All of John's brothers are physicians," or again, "All of
the people in test group A showed symptoms of the disease," not one of
the statements would be genuinely universal, despite the occurrence of the
word "all" in each ofthem. On the contrary, every one of the illustrations
is a singular proposition, for in each of them the individuals about which
something is said are capable in principle ofbeing designated. And as the
reader will know, when a proposition talks about individuals that can be
designated, it is singular. The indication of more than one individual does
not alone warrant saying the proposition is general; it must also say
126 Observables and Unobservables
something about a class; that it, the proposition must signify primarily that
something belongs to the instances to which it can be applied, either as
pertaining to their defmition, or as following from or explained by the
definition. When we say "Elephants are able to remember," the
proposition is general and capable of universal quantification because
"able to remember" is understood to accompany the nature of elephant
generally, and we understand remembering to belong to the individuals
as a consequence of their having that nature. But when we say something
about a group that can be designated, we understand the attribute to
belong first and foremost to individuals. Put another way, at first we know
the attribute only as belonging to some determinate individuals, and
another step is required to go beyond them to say something about the
class commonly. Thus, test groups and the like provide singular
propositions because the propositions talk about something that can at
least in principle be designated, and by themselves they do not allow us
to move to the class generally, which means the quantifier in those cases
is not truly universal. It simply means "the whole of." Again: though such
a proposition signifies the whole of the group being described, it does not
signify or express a regularity. 4
When one recognizes that systematic considerations bear on
regularities and hence on classes, he sees at once that the object of a
theoretical scientific investigation does not to seek to understand the
singular instance in its singularity. The biologist who dissects a shark does
so with a view to discovering the common anatomical features of the
species and perhaps of vertebrates more commonly, but he is not
interested in the peculiarities that belong solely to the individual specimen
on his lab table. Similarly, a chemist is not interested in the individual
samples as such which he puts in a test tube in order to observe a reaction.
On the contrary, he is concerned with the samples only in the measure that
they illustrate the character of the reacting substances generally.
Applied or practical disciplines are a different matter, however, since
their concern is with doing or making something, and doing or making
always bears on the singular. Consequently, we ought to keep in mind that
physics and chemistry are not the same as mechanical and chemical
engineering. Engineers and other practical scientists may begin with
general, theoretical considerations, but they design things that will be
produced as individual entities. Perhaps the "most singular" of the
mechanical disciplines is civil engineering, for the civil engineer has to
design a bridge for this location on this river, etc. His individual
Regularities, Experience, and Experiment 127
construction has to take into account all the individual conditions of the
location in which the construction will occur. In sum, there is a "polar"
difference between those disciplines whose concerns are solely investi-
gative or theoretical and those whose concerns are fully practical.
E. Experience: what it is
those that he has. In the past some whole societies have forgone any large
measure of technological development solely because they preferred
non-practical cultural pursuits as their social ends. Thus in man "problem
solving ability" refers not exclusively but primarily to theoretical
activities that are legitimately goals in themselves; in that way man is
most fundamentally distinguished from other species. But one final and
important note: the obscurity of natural entities is due principally to the
obscurity of natural causes 9 together with the unobservability of natural
substances. Were everything in nature observable, there would be no
obscurities, no problems, and no arguments. No need would exist to
illumine one proposition through others that implied it.
But experience does more than supply us with problems; it also shows
us regularities which, when expressed in propositions, function as
premisses in systematic accounts. It seems fair to say that stated in one
way the second law of thermodynamics is a generalization from
observation: "Heat passes from a body at higher temperature to a body at
a lower temperature." Whatever heat is ultimately conceived to be,
whether the fluid, "caloric," whether random molecular motion, etc., we
certainly must admit that when a cold body comes in contact with a hot
one, the cold body becomes warmer and the warm one gets cooler. That
is the first correlation we discover in relation to "heat transfer." Moreover,
if one asks how we know that the more sophisticated formulation of the
second law given by Clausius-it is impossible for any self-acting
machine to convey heat continuously from one body to another at a higher
temperature-is true, 10 he would have to start with the experience
described just above. Other illustrations of empirically obtained premisses
could be given, and later we shall point some out when we discuss another
topic. But for the moment, we shall make do with those we have
presented.
In summary, it appears that experience supplies some principles for
systematic procedures (we do not claim that all principles are so
obtained); so it would seem unnecessary to labor the point further.
Moreover, the mind's dependence on such regularities is plainly a
consequence of its dependence on observation, and that is central to our
considerations.
130 Observables and Unobservables
H. Experiment
It need not be said that a great deal has been written about the
characteristics of experiment. The control an experimenter exercises over
his observations, his use of instruments in making them, the active
relationship he has to nature when he varies objective conditions for the
sake of determining whether corresponding variations occur in the
property being investigated-all of these characteristics and others as well
have been described and discussed in detail by a number of authors.
Needless to say, a thorough knowledge of the character of
experimentation requires one to understand them all; but a remark by
Claude Bernard allows us to put aside most of them because it draws
attention to the feature of experiment that is most important here. We
refer to his statement that experiment teaches. 11
Because an experiment teaches, because it enlightens the mind in
regard to something initially obscure, it differs from that which is only
observation. To be sure, experiments are made for the sake of
observations, but they do more than allow us to observe. As we said
earlier, observations acquire a scientific or systematic utility when
memory collects them into what Aristotle called experiences; and that is
the key: experiment provides the mind with deliberately acquired
experience, thereby making man more fully the master of his own
learning. And because .experiment brings about deliberately acquired
experience, we have no trouble seeing why scientists are not satisfied until
an experiment or set of experiments has been successfully conducted by
several laboratories. Experiment, then, has the same teaching function as
experience in general, and it can lead to regularities that (1) give rise to
problems, or (2) that provide generalizations which can function as
premisses, or (3) that confirm or disconfirm hypotheses.
Some people are inclined to assign other roles to experiment, roles that
depend on the physical activities that are a part of nature's being pushed
to show herself in the laboratory. Much has been written about the
modification of experimental objects that occurs in certain experiments.
Sir Arthur Eddington suggested that some of the entities we "discover"
may well be the results of the experimental activities and not at all things
of nature. Certainly laboratory activities can modify realities so
extensively that the realities are outside their normal state. The last
elements of the periodic table are cases in point; for they are spoken of as
"artificial," and we do not encounter them outside the laboratory. They
Regularities, Experience, and Experiment, 131
illustrate our point even though they are not "by products" of experiments
intended for other purposes.
To the extent that instruments may modify the objects which they help
us to know, to that extent they impede nature's teaching function with
regard to us. But no matter the role of instruments, no matter the
modifications brought about by experimental activities, no matter the
unnatural state of some of the end results, the essential nature of an
experiment is to instruct the intelligence; that is why we perform them. So
to repeat our main point: the goal of experiment is the acquisition of
experience, which means that experiment may appropriately be defmed
as deliberately acquired experience; that is what it is and why it teaches.
Notes
1. Hans Reichenbach, The Philosophy ofSpace and Time tr. Maria Reichenbach
and John Freund (New York: Dover Publications, Inc., 1957).
2. Even when there is no behavior in the proper sense of the word, we sometimes
speak as if there were. For example, mathematicians occasionally talk about the
behavior of sets or functions when in point of fact they are speaking of relations.
In that case the word "behavior" is predicated as a metaphor. Also, in the course
of defining what color is, one could very well ask himself what a color "does."
It reflects light, of course.
3.. The Aim and Strncture of Physical Theory, trans. Philip P. Wiener (New
York: Atheneum, 1962, p. 144.
4. It need not be said that the so-called "accidental universal" is not properly a
universal at all, as its name indicates. It amounts to a "counting" of the
individuals.
5. Metaphysics, Bk. I (alpha major), c. 1.
6. The medievals had much to say about experience in this sense, and those
interested ought to see a fine article by James S. Stromberg, "An Essay on
Experimentum," in the Canadian journal, Laval Theologique et Philosophique,
Vol. XXIII, 1967, No. 1; and Vol. XXIV, 1968, No.I. Needless to say,
Stromberg treats the issue at length and in detail.
7. It is perhaps apparent that a man can have a memory full of observations but
lack experience in the sense we are discussing because he may fail to see the
correlation, the regularity they illustrate. Twins could in principle have had
precisely the same sensory observations yet only one of them have had
experience in the sense at issue here.
8. The word "chance" has a number of meanings, and we hope no reader will take
to mean a statistical probability. We discuss this topic in our Nature's Causes,
where we point out that the relevant sense in natural events is that some
unforeseen effect comes about from an incidental union of two agents whose
132 Observables and Unobservables
We are now in a position to discuss the flrst of our two main issues,
which is the relation of hypotheses to observables; and in a particular
way, the issue bears on the relation of theoretical entities to observables,
because theoretical entities seem to be constructs of the intelligence that
in the minds of some have little or no relation to observables. That is to
say, they seem to originate from the intelligence itself, and that will be the
main focus of our discussion here.
Philosophers have written much about the reality of theoretical
entities, about particles and forces, and their main positions go under the
titles of realism and antirealism. Many aspects of the issue have been
considered, and no one could consider them all even in a reasonably long
book. Here, however, we wish to talk about hypotheses, and though in
some ways they are well understood, it nevertheless happens that some
philosophers regard starting points of "metaphysical" philosophical
theories to be hypotheses when they are not. In his day too Newton had
much to say about what appeared to be opinions on hypotheses that would
not stand the light of day. 1 So it seems desirable to make very plain
exactly what we think hypotheses are, and later in another chapter to show
why this is so. Our task, then, will be ( 1) to show the function of
hypotheses as explanatory premisses; (2) to show in what measure
postulated realities, whether forces or elementary particles, are real; (3)
to show how they are related to observation and observables.
Let us begin our discussion with a commonplace illustration. Suppose
that Socrates returns home one day to fmd that his lawn mower is missing
from his garage. Because he has an inquiring mind, he wishes to
understand why it is missing; but we must quickly add that any attempt
he makes to account for its absence will require that he assume some
134 Observables and Unobservables
the cause; for particles of some sort there are. It is, as we said, a question
of the nature of the particles. What, then, are the defming attributes of
those particles? That would seem to be the issue.
Dalton assumed that among other things the atoms or particles were
solid and indivisible, from which it followed that bombardment should
not show elementary substances to be porous, nor should it be possible to
break something off from their atoms. Experiment, however, proved
otherwise, and so the particles that entered into chemical reactions were
soon known not to be elementary; that is, they were known to be
composed of other particles. What that meant, of course, was that the
definition of the chemical particles had to be modified to show their
composition. Let us repeat: the experimental evidence did not show there
were no such particles; it showed only that their character had been
wrongly conceived, requiring that their descriptions or defmitions
undergo modification, a process that still continues, leading as it has to the
realm of charmed quarks.
To consider, now, hypotheses from the point of view of modus tollens,
we know that they can be disconfumed with certainty because such an
argumentative process warrants a necessary inference. One may argue that
if( chemical) atoms are indivisible (uncomposed), then no particles can be
broken off from them; but experiment shows that particles are broken off;
hence (chemical) atoms are not indivisible (uncomposed). The destruction
of the consequent does not destroy the entire conception of chemical
particle but only a part of the concept, which plainly accounts in large
measure for the steady improvement of the atomic theory over the years.
With each new development only a part of the earlier definition or
description is modified; a substantial remainder is left. The very general
notion that natural substances are particulate is left undisturbed, and the
steady development of the theory shows in more and more detail how that
general statement is true. In short, the theory approaches, as a kind of
limit, a more and more perfect understanding of the first, most general
proposition bearing on the particulate constitution of natural substances.
It approaches a precise understanding of what the elementary particles
are; yet we shall never be certain that we have attained the limit.
And so it would seem that neither a straight yes nor a straight no can
be given to the question, "Are elementary particles real?" Each time a new
one is discovered, it is discovered because some observable effect
(directly or indirectly, with or without intermediary instruments) has been
isolated; and no effect is capable of bringing itself into existence.
138 Observables and Unobservables
Necessarily, then, something real and particulate is the cause, but exactly
what the nature of the new particle-cause is remains uncertain. At this
point we are constrained by the limits of our defective process, and we
cannot assert that we know the particles to be constituted exactly as we
conceived them. Thus in a limited way we know theoretical particles to
be real, and in another way we are uncertain about them.
The character of hypotheses as we have described them seems to be
well expressed by a physicist who discusses the earth's composition,
having to start from considerations of its diameter, the precession of its
axis of rotation, its moment of inertia, its volume, mass, and average
density:
As we had guessed, the moment of inertia of the earth is less than it would
be if its density were completely uniform. This indicates that the mass is
concentrated toward the center, but does not allow us yet to find the
detailed distribution; it is useful in eliminating many distributions that
might be imagined without giving a single unambiguous answer for the
actual distribution. Nevertheless, on the basis of the results already
established, we can begin the construction of provisional models for the
density distribution inside the earth; models which will almost certainly
have to be revised as evidence accumulates. 3
This illustration seems to make our point, although we admit that it does
not bear on theoretical entities. Nevertheless, the difference between
understanding the composition of the earth in a general way in contrast to
understanding it in detail illustrates the point of interest. The physicist, on
the basis of what he has established very generally about angular rotation,
knows that the mass of the earth is unequally distributed; he has no doubts
about that. The precise character of that distribution is, however, obscure.
If we now look back at our illustration involving Socrates and his lawn
mower, we see an important difference between the assumptions which
account for chemical reactions as well as other such phenomena and the
assumptions which accounted for the absent lawn mower. Atoms and their
constituents enter into propositions which are in principle incapable of
being verified inductively, that is, by observation, because the effects to
be explained are not sufficiently intelligible for us to see in them the
precise nature of their cause. On the other hand, although the assumptions
Socrates made to account for his absent lawn mower were not verifiable
in that instance, nevertheless considered generally they are empirically
verifiable and are therefore independent of the argument in which they
Hypotheses and Reality 139
occur. Furthermore, the causes they contain are known to be real and their
character understood, whereas the causes in current scientific hypotheses
are inherently obscure. To go beyond our example: if one calls the
assumptions of the social sciences "hypotheses," he is talking about a
different kind of animal; such hypotheses are not of the same nature as
those of physics and chemistry.
cause? Or we might put the question another way and ask, just what does
the theory do?
In view of the large amount of literature that has appeared over the
years on the nature of scientific knowledge, it might seem presumptuous
on our part to offer an answer to the question just raised in the small
amount of space we shall utilize here. Yet we wish once again to remark
that our aim is to consider the issues only insofar as they are related to the
general thesis of this essay, namely, the way in which the intelligence
learns about unobservables through observables. On that account, then,
the present considerations will be limited; and we shall not dispute the
many positions that have been taken on the large number of topics
associated with the general issue but instead confine ourselves to a few
points that are rather directly related to the discussions that have gone on
in earlier chapters.
In line with what we said earlier, we ought to note again that the
question whether something is must not be identified or confused with the
question what something is or with the question why it is. Each of these
is distinct, and the first can be answered even when the others cannot.
Moreover, the reader will recall that the inference from an observable to
an unobservable cause is precisely of that nature: it is an inference to the
actual existence ofthe cause, not to its particular nature. On that ground,
then, I would like to argue once more that certain distinctions must be
made apropos the reality of theoretical entities; for there is a sense in
which we are certain that theoretical entities exist, that they are real, and
there is another sense in which we cannot know that they are anything
more than constructions of the intelligence.
If we consider planetary motion, for instance, we know from the start
that the motion is not responsible for itself. It does not determine itself to
be elliptical, it does not slow itself down in one part of the orbit and speed
itself up in another, etc. In short, we are certain that the motion and its
characteristics do have a real, an existing cause. Again, real effects cannot
issue from fictional or other non-existent causes, and so we do know that
there is a referent for "cause of the motion." To sum up: we are certain
that the motion of the planets, the falling of bodies, the movement of
charges, etc., all have some real cause, to which is attached the general
name "force."
The force responsible for planetary and other, similar motions, has
received the name "gravity." But once it has been named, what more can
we say about it? What additional attribute or attributes other than
Hypotheses and Reality 141
existence can we assign to it? Such was indeed the quandary Newton
found himself to be in; he could determine no defining or descriptive
attribute of the cause and so declared that he did not fashion hypotheses.
Thus "gravity" remained for him and appears to remain for us a
metaphor. We do not have an analogy, the reader will recall, until we can
extend the naming word to the unobservable cause on the basis of
something belonging to the latter that is distinct from the characteristics
of the observable effect. But when gravity and other forces are conceived
as fields, then they are conceived in the mode of substance, a point we
discuss elsewhere. 4 If, then, gravity or any other force is postulated to be
a field and thus a substantial entity, the notion of the force would seem to
lose its metaphorical character, at least in a significant measure.
At this point it seems advisable to repeat, perhaps with some emphasis,
a point we made earlier when we said that the obscurity of the effect is
precisely what demands that we introduce hypotheses in order to explain
gravitational motion. When we consider movement that is change of
position, we see that as a reality it is minimal. That is, the "ontological
content" of such motion is next to nothing. Motion, either translational or
rotational, does not introduce modifications into the physical properties
of natural substances; it does not produce a qualitative alteration of
properties, as does heat. This this inherent ontological obscurity, owing
to the impoverishment of motion as something real, is the proximate
source of the mind's inability to infer the nature of the cause from the
effect; and the deficiency is not so much one of the mind as it is of the
reality known. After all, an argument cannot yield in the conclusion more
than is contained in its principles. In sum, then, the mind's inability to
infer the character of the cause, its inability to get beyond a metaphor in
its conception of gravitational force except to say that it is a field is the
consequence of the radical unknowability of the effect upon which the
inference to the cause depends.
As evidence that what we are saying here is actually the course the
mind must follow, consider the development of the notion of force in the
mind of Kepler. According to Max Jammer, Kepler's early conception of
force made the latter to be a soul, a conception he later modified as
follows:
And so from soul to chains to ... ? Truly, we cannot escape the use of
metaphors. But if our knowledge of the nature of gravity does not pass
beyond being metaphorical, in what way can gravitational theory be
explanatory? Is it not, as is claimed, merely descriptive?
In reply, let us note that given our analysis above, we can say that the
theory is explanatory insofar as it tells us how the cause operates; that is
to say, when we know that the planets, as well as all other bodies, "attract"
each other with a "force" that is proportional to the product of their
masses and inversely proportional to the square of the distance separating
them, then we have described the mode in which the cause -whatever its
nature-operates. But of course that is not an explanation in the full sense
of the term; it is only partial and obliges us to say, therefore, that we have
a description of the behavior of the cause through its effects but that we
do not know the inherent attributes of the cause and so do not have an
explanation in the full sense of the word. Put another way, we know that
there is a referent, but we do not know its character, its nature; hence we
have an explanation in a diminished, qualified sense. So it seems to us that
the realist and the anti-realist are each partly right, which makes our
answer somewhat similar to but also somewhat different from the one we
gave in connection with particles. Our knowledge of the latter is not
altogether metaphorical, since we know that they are parts of substances;
still, we cannot say without qualification that the particles as conceived
are real, since we cannot determine that our definitions describe them as
they actually are. It would seem, then, that from the realist point of view
our knowledge of those forces which are understood in no greater
measure than gravity is inferior to our knowledge of elementary particles,
at least in some measure.
It might seem that at times theory has missed the boat on both
accounts, failing even to determine that a cause exists. An example which
might appear to make the point is the ether, which was postulated as a
medium for the propagation oflight waves. Yet we may argue that further
Hypotheses and Reality 143
Notes
sciences.
We must grant at once that philosophy is not quantitative; it does not
measure, and its arguments are not mathematical. But lest one think that
the measurements and the mathematics which are employed by physics
and chemistry are the solely the consequence of the genius and a new
"insight" of men such as Galileo, Kepler, and Newton-lest one think that
is the case I wish to argue that the use of mathematics is per se to the
science of physics and can be accounted for very simply by considering
the cause oflocal motion, change of position as the physicist says. But we
must explain ourselves.
Opinions on what is called the "Copernican revolution" have, as
everyone knows, been many, and they are summarily represented as to
type by Alexander Koyre in a passage it is useful to quote:
Some others stress the practical attitude of modem man, who runs away
from the vita contemplativa, in which the medieval and antique mind
allegedly saw the very acme of human life, to the vita activa: who
therefore is no longer able to content himself with pure speculation and
theory; and who wants a knowledge that can be put to use: a scientia
activa, operativa, as Bacon called it, or, as Descartes has said, a science
that would make man master and possessor of nature.
The new science, we are told sometimes, is the science of the craftsman
and the engineer, of the working, enterprising, and calculating tradesman,
in fact, the science of the rising bourgeois classes of modem society. 2
thereby that the contemplative function of the mind either was not
considered a real option, or that whatever there is of the
contemplative-the theoretical understanding of nature for its own sake
-in human heads is subordinated to practical benefits. To be sure, one can
hardly deny that the technology accompanying the scientific revolution
accelerated the development of the modem commercial era and gave men
an extensive dominion over nature; but the question is: did the
development of an interest in trade and manufacturing bring about the
quantification and mathematization of the study of nature? Now such
interests may well have disposed scholars of the times to look at the world
mathematically, yet we shall argue that it is not the fundamental reason
why physics became mathematical..
But Koyre offers another explanation, one that he endorses himself:
I am convinced that the rise and growth of experimental science is not the
source but, on the contrary, the result of the new theoretical, that is, the
new metaphysical approach to nature that forms the content of the
scientific revolution of the seventeenth century, a content which we have
to understand before we can attempt an explanation (whatever this may
be) of its historical occurrence.
them, just as a photon, once it exists, does not require the action of an
extrinsic agent to initiate its motion.
Turning now to the heavens, in addition to the earth, there were the
celestial bodies which were held to be radically distinct in kind from
bodies on the earth. The celestial bodies were thought to be incorruptible
and everlasting and on that account not constituted of the same sort of
material as things of the earth. Aristotle considered such a supposition to
be necessary because according to the astronomical records available to
him no one had observed either a star or planet come into existence or
pass out of existence, which made reasonable the hypotheses of a different
kind of material constitution and the consequent incorruptibility of
celestial bodies. 4
But with the advent of the new data that grounded a new astronomy,
that of Copernicus, Galileo, Kepler, and Newton, the understanding of the
cosmos underwent a very radical change. First in importance was the
implication that celestial bodies are not composed of a material radically
different from the substances found on the earth. Once the face of the
moon was seen (with the aid of the telescope) to contain many irregular-
ities, once the earth was understood not to be the center of the universe,
once the planets were held to revolve about one star among millions, then
it became more reasonable to assume that terrestrial and celestial bodies
are composed of the same kind of materials. From then on celestial bodies
could not be thought incapable ofbeing produced or destroyed by natural
agencies, however infrequently such events might happen. And if the
celestial bodies are made of the same kind of materials as terrestrial
bodies, then there is no reason to think of them as being moved by
agencies different from those that move bodies on the earth. Furthermore,
if the earth revolves about the sun, then the elements do not move toward
the center of the universe or toward its boundary by reason ofbeing heavy
or light. In short, once the dike had been pierced-to use a cliche-the
trickling water soon became a flood, sweeping before it the entire edifice
of Aristotelian science. Furthermore, space now came to be considered an
existing container, the extent of which was infinite, or at the least
unknown, so that there could not be a question oflocating moving bodies
in relation to the boundaries of space, of the universe. Such conditions
clearly required a new understanding of the cause of local motion.
The Mathematical Character ofPhysics
151
There seems little doubt that the most serious weakness in the
Aristotelian theory of local motion was the account it gave of projectiles,
which are instances of what Aristotle called constrained motion, a motion
he thought could be continued only by reason of a complicated movement
of the air through which it traveled. From early times this part of his
theory met opposition, but not until the medieval period did its chief
opponent, the impetus theory, become accepted. The important advantage
of the impetus theory was that it did not require a continuously acting
agent to sustain the constrained motion; instead it held that contact
between an agent and a moved body was necessary only at the beginning
of the motion. Thomas Aquinas stated the relevant principle as follows:
of a necessary property.)
Physicists are quick to point out that an absence of extrinsic forces can
be understood in two ways: either no forces whatsoever are acting on a
body, or the forces acting on it are balanced, which means that their
vector sum is zero. Hence the first law also says that no unbalanced force
is affecting the body, which means that the only effect for which we can
detect an agent cause-the unbalanced force-is an acceleration, which
by defmition is a change of speed or direction. Now no determinate
direction in relation to the universe or to space as a whole can be assigned
to a motion; so to repeat: according to Newton the only observable effect
for which a cause can be assigned is an acceleration. Thus unlike
Aristotle he could not attempt to describe a particular motion by referring
to the extremes of the universe.
In the absence of known limits of the universe an acceleration can be
established only by reference to a set of coordinates, and whether the
change be one of speed or one of direction, acceleration is a variation in
a motion that is a more or less. That differences in speed are differences
of more or less is obvious, for speed is an indication of how fast or slowly
a body is traveling. But in the new physics a change of direction is also to
be regarded as a difference of more or less; for when directions can no
longer be defined absolutely in relation to parts of the universe they must
be described as angles measured from coordinates (that are not established
by the boundaries of the universe), which means that direction too
becomes a difference of more or less. That is to say, directions are distin-
guished quantitatively rather than qualitatively. Even movements that are
oppositely directed and represented as positive and negative come to be
regarded as having no absolute qualitative differentiation or direction in
relation to the universe. In short, Newton's first law tells us that the only
difference in the motion of a body that can be causally explained is a
difference ofmore or less. On that account a science of moving bodies can
no longer be founded on the natural, qualitatively distinct kinds of local
motion that characterized Aristotelian physics.
And so now we see that if they are to be known precisely, differences
of degree must be measured. Consequently the new physics was obliged
to become quantitative, for according to its fundamental philosophical
position only variations in speed or direction, only differences of degree
had causes that could be detected. And once the official data of physics
became numbers standing for measurements, then all correlations of data
had to be represented by mathematical propositions; for every relation
The Mathematical Character ofPhysics 153
Notes
1. The quotation we give is from Gerald Holton and Duane Roller, Foundations
of Modern Physical Science (Reading, Massachusetts: Addison-Wesley
Publishing Company, Inc., 1958, p. 22. The quotation is, we think, representative.
2. Newtonian Studies (Cambridge, Mass., 1965), pp. 5B6.
3. Ibid.
4. Aristotle was well aware that his account of celestial movement was founded
on an hypothesis, for he says: "The mere evidence of the senses is enough to
convince us of this, at least with human certainty. For in the whole range oftime
past, so far as or inherited records reach, no change appears to have taken place
either in the whole scheme of the outermost heaven or in any of its parts." (De
Caelo et Mundo, Bk. I, chap. 3, 270b, 11 16.) Thomas Aquinas comments on this
passage as follows: "This is not necessary but probable. To the extent that
something is long-lasting, more time is required in order that this change be
detected; just as the change in a man will not be detected throughout a two or
three year span during which the change in a dog or some other animal of shorter
life will be observed. Therefore, someone can say that although he heaven is
naturally corruptible, it is so long-lasting that the total time span of which we have
memory does not suffice for detecting its change. (In I de Caelo, Iect. 7, n. 6.) The
translation is ours.
5. Questiones Disputatae de Potentia, (Rome: Marietti, 1949), q. 3, a. 11, ad 5.
6. Plainly an analogue device such as a mercury thermometer is closer to
experience in this regard than a digital counter. One might even ask if the latter
would be intelligible if it were not known to do the sort of thing that a
thermometer does.
Chapter 13
General Considerations on Philosophy
A. Introduction: the separation of philosophy from science1
premisses, we are talking about the kind of thing man is; we are talking
about a nature that is found in every individual. If it is not, then the
individual is not of the human species. Our point is that we cannot do
without these species-names, for we have no systematic understanding of
the specimens or individual members without knowing first of all the
nature that is common to them all; so we must have words that signify the
species. Furthermore, that the intelligence can understand realities
according either to their existence in the world "out there" or according
to the existence they have within the intelligence-their intentional
existence-is well understood by philosophers, if not so well understood
by many others. Now the difference in the kind of existence that can
accompany what a word signifies can bring about difficulties, which a
certain use of the word "man" illustrates. Consider the statement "Man
evolved from a lesser primate." If one now wishes to use the statement as
a premiss and say that "John is a man," and then say therefore "John
evolved from a lesser primate," he draws an utterly false conclusion, for
we may not say "John evolved from a lesser primate." On the contrary, he
was born ofhuman parents. Hence neither "person" or "human being" can
be substituted for "man," since they will render some propositions false.
For instance, no Christian may say that Christ was a "person like us in all
things but sin," because they all understand him to be a divine, not a
human person, and so not like us in that way. Nor as we just said can we
substitute "human being" for "man" because "human being" too signifies
a person. 7 In sum, man is a species and the word "person" cannot be
substituted for "man" whenever the species as such is being discussed.
We must now look at hypotheses for a second time for the sake of
seeing better how they function as premisses in explanations so that we
may contrast them to premisses that are explanatory in philosophy. Earlier
we noted that atomic theory provides hypotheses as principles of
explanations, as explanatory premisses, which means that the laws
obtained by induction from which the problems arose are now seen as
deductive conclusions of arguments. The investigators knew that in its
essentials the explanation would read: chemical reactants combine in
definite proportions (in multiple proportions, conserve their mass)
because p. In other words, they already had conclusions; what they
needed was a premiss. Dalton's assumption that chemical reactants are
General Considerations on Philosophy 163
D. Syllogistic arguments
The same procedure can be used with all the other examples; for instance
the argument on herding goes as follows:
All animals that are better protected from predators by herding do herd;
small animals are better protected from predators by herding; therefore
small animals herd.
E. Explanation in philosophy
F. A summary
Notes
a syllogistic inference.
11. Plainly our explanation is a set of bare bones. Questions having to do with
pressures on the exercise of liberty that come from emotions, etc., are left aside;
they are not necessary for the purposes we have in mind here, namely, the
illustrating of philosophical explanation, since they bear on the application ofthe
notion to human acts, not on the justification of the notion itself.
12. It goes without saying that the character of the intelligence must be known in
detail to appreciate the efficacy of the explanation fully. A mere statement ofthe
explanatory premiss cannot do that sort of job.
13. An extensive comparison of calculational, truth functional logic of modern
times to the kind of logic first worked out by Aristotle is made by Henry B.
Veatch, Two Logics, (Evanston, Illinois: Northwestern University Press, 1969).
Veatch considers our present issue, namely the unsuitability of relational or
symbolic logic for philosophy, in a thorough and competent way. We can touch
only on a central point or two, whereas he takes up all the major topics associated
with the character oflogic. We recommend the reader see his work for his more
thorough treatment.
14. E. J. Lemmon, Beginning Logic (Indianapolis: Hackett Publishing Company,
1978), p. 11.
15. As my colleague James Stromberg notes, this way of defining a valid
argument begs the question, for one cannot presuppose the truth of the conclusion
because that has to be established by a formal syllogistic pattern.
Chapter 14
Philosophical Principles: Preliminaries
A. Introduction
When we compare the adult to the infant mind, we see that the
intelligence, like the body, undergoes development over the course of
time, our descriptions of the mind employ the language of the body.
When we say, for example, that this or that student is developing well,
that he is improving his physics or mathematics, that he is strong in
chemistry, we extend the language applied to observable entities to the
unobservable intelligence. And just as the physical is subject to certain
general developmental principles, so too the intelligence, which moves us
to ask the question we now must consider, namely, what are the general
principles that determine the course of the mind's progress, its
development from a state of ignorance to a mature grasp of a systematic
discipline or disciplines? Let us note at once that the issue we shall
discuss does not belong to the "psychology of learning," as that phrase is
commonly understood. We shall have nothing to do with emotional
preconditions, "motivation," ambient circumstances, etc. Instead, our
concern is with the procedures that are innate to the intelligence and that
are not subject to variation by human manipulation. Stated another way,
our concern is with the mode of knowing that is built into the mind and
necessarily manifests itself in every functioning human being,
independently of his cultural or environmental circumstances.
There is still another way in which the mind develops by going from the
confused to the distinct, which we shall discuss by making use once again
ofPiaget's interrogations of children. In Chapter V, he describes interviews
which show that at an early age children ordinarily identify conscious
activity with activity generally:
During the first stage everything is regarded as living which has activity or
a function or a use of any sort. During the second stage, life is defined by
movement, all movement being regarded as in a certain degree
spontaneous. During the third stage, the child distinguishes spontaneous
movement from movement imposed by an outside agent and life is
identified with the former. Finally, in the fourth stage, life is restricted
180 From Observables to Unobservables
The child's failure to grasp the distinction between living and non-living is
quite different from the attribution of life to the inanimate that an adult
might make for the purpose of defending a panpsychist position, a
confusion that is an error rather than an imperfection. We all understand
imperfection to be part of the normal developing process native to the
intelligence, but the adult who erroneously assigns to an entity an attribute
which the mind has come to understand with some distinction is not in a
state of imperfection but of error, a quite different state of affairs. Error is
a confusion of quite another sort, and as we said, pansychism is an
illustration of it.
Although the data in the illustrations above are new, the principle is not,
for Aristotle illustrated the developmental character of the mind when he
pointed out that little children call all men "father" (or all women
"mother"). His remark does not mean that children do not recognize their
parents on a sensory level, for they obviously do, and at an early age.
Rather, his statement tells about the child's intellectual apprehension. A
neighbor's boy once asked me, inquiring about my wife, where my mother
was, which showed that to him "mother" was a synonym for "adult
woman." He had not yet come to understand that his mother is the
particular woman who was a cause of his existence.
There are other ways, too, in which this sort of development shows
itself. If, for instance, a hunter looks across an expanse of prairie to the foot
of a hill and sees something on the slope, he will know that it is a
four-legged animal, though he may not know whether it is an antelope, an
elk, a deer, or something else. The hunter may be certain about the animal's
generic character yet be ignorant of its specific kind. Similarly, a physicist
may see an unexpected track in his cloud chamber and not know what
caused it, except that he is certain the track was caused by something real,
and here, too, the statements he makes are limited to generic attributes. He
is unable to assign those that are specific, which shows the character of this
sort of development, namely: we can know the general with certainty
without knowing the specific, and that is to know in an indeterminate,
confused way. Let us repeat: the mind develops by going from the general,
which is indeterminate or confused, to the specific, which is more
determinate or distinct. (Unfortunately the bureaucratic mind too often
goes in the other direction.)
It is, of course, characteristic of the generic that it is "empty" relative to
Philosophical Principles: Preliminaries 181
D. A summary
from the confused to the distinct, and that kind of development occurs in
two ways: first, when we go from knowing wholes in a confused manner
to knowing their parts and then return to the wholes with a more distinct
understanding; and second, when we proceed from considering the general
to the more specific by way of distinguishing the species or subdivisions
under the genus or more universal class. We see that these modes of
development are natural to the mind, for they are ubiquitous and
unavoidable: they stem from the constitution of the intelligence. Yet we
must not forget that we are talking about systematic or scientific
knowledge.
Notes
I. Op. cit., p. 173. This time we shall use only Piaget's summary remarks and not
the interviews themselves.
2. Op. cit., p. 194.
3. This more empty character of the general is of course why Hegel had to claim
to have discovered a "new" universal, the "concrete" universal, which supposedly
contained the differences within it. Without such a claim he could not have so
audaciously pretended to deduce the species from the genus.
Chapter 15
Natural Philosophy
B. Preliminaries
physics makes of the notion and without being an expert on the subtleties
of attempting time's measurement; and of course to measure time we must
have at least a general understanding of what time is. We must add, too,
that other qualitative distinctions pertaining to common natural properties
are presupposed to the investigations of the natural sciences, distinctions
that need not be understood by such sciences according to anything more
than our common conceptions.
Perhaps it is even more obvious that considerations on what motion is
or what change in general is in no way affects mechanics or any other part
of physics, or any part of the other natural sciences, for that matter.
Furthermore, if a scientist makes a general distinction between categories
of change on the ground that some are chemical and some are physical,
then in the process of distinguishing them he will have to make use of the
notion of substance, since a chemical change involves a change of the
latter. But just what is involved when a modification affects a stuff rather
than one of its modifying properties is a matter that requires systematic
examination; and here, too, it may be said that the consideration in no way
affects the hypotheses that the chemist may have formed about the
processes by which the changes in stuff are thought to be accomplished.
In sum, the sorts of questions described above are the sort of thing natural
philosophy can and ought to do, and they separate it from the
experimental sciences themselves.
To continue the point, biology is a natural science that deals with
living entities; yet the biologist does not systematically examine, for
example, whether organisms are aggregates of particles, what an
aggregate is, what internal states as distinct from external operations are,
etc. These issues do get discussed, not by the biologist but by
philosophers.
And so having made these preliminary remarks, we may now ask, the
question, "What is natural philosophy"?
to make the primary focus of his labors differences of more or less? But
then why, if the distinction of the natural sciences from philosophy is
found principally in the use of measurement, is ethology not a part of
philosophy? An ethologist must, to be sure, make more observations than
a philosopher, but can one look at that as an essential characteristic?
We must concede at once that no strong case can be made for putting
ethology on one side of the fence rather than the other if we depend on
such grounds. If ethology starts from observations, if it argues
systematically, defines, classifies, etc., by taking into account similarities
and differences that are not reducible to measurements, then little
substantive ground can be found for distinguishing it from natural
philosophy, especially if the latter considers all "mental" activities of
organisms. Quite the contrary, on the basis of an absence of measurement
a strong argument could be made for saying that ethology is an instance
in which the names "science" and "philosophy" can be applied to the same
discipline. But such an identification is unlikely to meet with favor from
ethologists, owing to impressions they may have as to what "philosophy"
does in practice, and owing also to the insistence on keeping philosophy
and biology distinct in accordance with current academic custom; and
assuming that such a longstanding custom cannot be wholly in error, the
objection is understandable. Moreover, the distinction between them must
in some measure be accommodated according to the points that have been
already made; for the considerations of the ethologist are more particular,
while those of the philosopher are more general. That is central to the
distinction between natural science and natural philosophy.
In the area of sociology and anthropology Lewis Mumford regards
himself as a generalist, and the name suggests how he views his function.
It indicates the kind of considerations Mumford makes on the detailed
investigations of his colleagues. He compares the particulars, the details,
to discover the general regularities that might belong to many distinct
societies. Similarly, one might describe the function of the natural
philosopher in that way: he is a generalist insofar as he considers the
causes and properties common to all natural things; and the distinctions
he makes are evident on the basis of ordinary experience, however much
he may also avail himself of the data and regularities uncovered by the
natural sciences.
It bears emphasis that generalizing supposes data and information that
are sufficient to supply an adequate inductive base for the generalizations.
Thus, like the generalist in sociology, the natural philosopher needs to be
Natural Philosophy 189
reasonably acquainted with the data and the more important observational
regularities the scientist comes to know, although the natural philosopher
need not be a scientist nor have a scientist's understanding of his subject
mater. In our time, acquiring knowledge of natural regularities is for most
people a vicarious business. We live in artificial environments that
preclude our observing the behavior patterns and traits of many things in
nature, and from this point of view the American Indian was considerably
our superior. It is easily seen, then, that the natural sciences provide the
philosopher with a vicarious observational contact with reality, a not
insignificant service. But we must consider the point more determinately.
The general principle according to which one orders the parts of a
science which have the same category of subject matter is revealed in the
following quotation:
... in a certain way the sciences follow on things; for [operational] habits
are distinguished according to their objects, from which the habits derive
their species. But the natural scientist considers motion and mobile things.
The Philosopher [Aristotle] says in the Physica, Bk. II, that anything
which moves belongs to the investigation of some physical science.
Therefore it is necessary to distinguish the parts of the science of nature
according to differences in motions and mobile entities. Now the first kind
of motion is that which is local [change of position]; it is more perfect
than the other kinds and is common to all natural bodies .... Hence after
the consideration of motion and moveable entities that is common and that
was taken up in the Physica, [Aristotle] first considers bodies according
as they are moved by local motion, and he does this in the De Caelo,
which is the second part of natural science. There remains, then, the
consideration of other motions which follow, those motions that are not
common to all bodies .... Among these, production and destruction
[generatio et corruptio-what we today call chemical change] have the first
place. Alteration [what we now call "physical change"] is ordered to
production as to a goal. Growth also is consequent upon production and
destruction [chemical change], for growth does not occur without certain
productions and destructions by means of which nutritional materials are
converted into that which is nourished by them; for as the Philosopher
says in the De Anima, Bk II, just as food nourishes insofar as it is
potentially flesh, so, too, it increases magnitude as it is potentially
quantified flesh. Thus it is necessary that these motions be considered at
the same time because in a certain way they follow on production and
destruction. 5
The principle according to which the sciences are to be ordered for the
190 From Observables to Unobservables
The reader is certainly aware that some of the issues that we have
claimed to be the province of natural philosophy do get discussed under
the title of philosophy of science. When one "explicates the concepts of
the natural sciences" he can hardly avoid touching some of the topics that
belong to natural philosophy. Nonetheless, these two parts of philosophy
ought not to be regarded as equivalent because questions of scientific
methodology, which form a large part of the philosophy of science, do not
fit into the same category as substantive questions that bear directly on
natural entities and their properties. Nor do the epistemological topics,
those that bear formally on the kind of knowledge each is, belong to
natural philosophy. In short, the philosophy of science has its own
character and its own role to play, and one ought not to confuse it with
natural philosophy.
In closing, I would like to note that although the foregoing discussion
is but an outline, it does indicate what, in the minds of many, natural
philosophy ought to do. It would seem evident that a complete
understanding of nature requires, to the extent possible, a systematic
examination of the general character and attributes of natural entities
insofar as they are qualitatively distinct and insofar as their differences are
not reducible to differences of degree (a state of affairs that precludes
philosophy disappearing into science). The latter task, I have argued,
belongs to philosophy, even though one cannot claim that its contem-
porary mainstream succeeds in or even attempts to accomplish that task.
But the actual state of affairs matters little for the theoretical issue, and the
work is there to be done.
E. Epilogue
Earlier we made the point that philosophy starts from and is concerned
with regularities, as are the sciences of nature. But before we leave the
192 From Observables to Unobservables
Notes
1. The position presented here regarding the relation of natural philosophy to the
natural sciences, except for the shortcomings that are mine, is substantially that
of the late Charles DeKoninck, of Laval University in the city of Quebec.
2. An ordinary understanding derived from our common and indeterminate notion
of these properties founded on common experience suffices for such purposes.
3. What we say here about why physics measures and how natural philosophy is
to be distinguished from the natural sciences is treated at length in our essay, "The
Character ofNatural Philosophy," in The New Scholasticism, Vol. LI, 3, Summer,
1977.
4. Ethologists do measure some phenomena to find out, for example, if one mode
ofbehavior is more efficient than another, and such differences of more or less do
occasion some measurements. But these do not tell against our point.
Natural Philosophy 193
Definitions are answers to the question "what is it?" and our desire to
defme arises from an obscurity in some class of things, an obscurity that
we want to overcome. The procedures the mind employs in arriving at
defmitions results, however, from the way the mind develops, and that is
what we wish to show. But first, some preliminaries.
B. Packaged formulas
C. Nominal definitions
Without doubt when the answer to the question, "what is it?" yields the
name and only the name of the object of inquiry, we are confronted with
Definitions 197
D. Systematic definitions
The term fluid refers to a substance that does not have a fixed shape but
198 From Observables to Unobservables
that is able to flow and take the shape of the container .... 5
That is the first line of the chapter on the statics of fluids. Then later the
i
authors say something else:
From this definition we shall be able to derive all the laws that govern the
experimental behavior of ordinary liquids and gases in static equilibrium.
Plainly the first defmition the authors give does no more than tell the
meaning of the word "fluid" and therefore is a nominal defmition stated
in terms of its characteristic observable property, the flowing motion
common to liquids and gases. However, the second definition is
substantially different from the first, for it is a principle of inference
insofar as it implies the laws (the regularities) which reveal what fluids do
under conditions of static equilibrium. Again: the second defmition allows
one to infer theorems that are conclusions derived from the defmition as
an explanatory principle. It allows the physicist to see why fluids under
static conditions behave as they do, which is the ground for our calling it
a "systematic definition." By calling it that we focus attention on the role
of the defmition as an explanatory principle, and it has that role because
it tells us what a fluid is in such a way as to allow us to account for other
behavioral properties a fluid exhibits.
Elaborating a bit, we wish to add that although the definition as stated
does not declare that fluids flow, nonetheless it presupposes such a
motion. Only because a fluid flows is it unable to exert tangential or
tensile forces, etc., and so the nominal definition is presupposed to the
systematic, and it provides the proper starting point, by reason of its own
light. Hence to account for the reaction of the fluid to exterior forces
exerted on it is to account for its passive behavior on the basis of a
fundamental motion that characterizes the substance. 6 But we do not
pretend to pass a physicist's judgment on the merits of the defmition
quoted above. Our aim is only to note the existence of a kind of defmition
which tells what something is as well as possible, providing thereby a
Definitions 199
Our point has been to distinguish two main kinds of defmition on the
basis of two distinct functions they can have. One function is to attach a
name; the other is to say what something is. As we mentioned above,
there are many kinds of nominal defmition, some of which approach more
closely to being systematic than others. Often, too, they ground
inferences, although not of the explanatory kind. 9 In fact, the reader can
see that any defmition which falls short of being systematic in the sense
described above ought to be classified as a nominal definition. 10
Definitions 201
Notes
All of the phenomena that we call weather, from the most gentle of
breezes to the driving winds of a storm, or from clear skies to massive
downpours of rain or snow, are the results of the unequal distribution of
heat energy in the atmosphere.'
recall from our considerations in chapter 12 that first principles are not
necessarily hypotheses; on the contrary, they can be empirical regularities
obtained inductively from observation. But the issue is not whether there
are first principles, for everyone knows that they occur in every science;
the issue is whether they are anything more than regularities obtained
from observation.
To say that the nature of first principles which are not hypotheses is a
somewhat confused issue is not to overstate the case, and an illustration
of how propositions are thought to be divided may help to set up the
problem we wish to discuss:
We are all aware that there are some first principles which are
employed by every systematic discipline. The statement, "Anything either
exists or does not exist," or what is the same, "It is impossible for
something to be and not to be at the same time and in the same respect
(the principle of non-contradiction)," is the most obvious case. The mind
cannot take a single mental step without recognizing, at least implicitly,
that the principle is true; for anything whatsoever that we consider either
does or does not exist. Another example, one that is sometimes not
understood properly is, "A whole is greater than any of its parts." Still
another, less frequently articulated common principle, is "Anything
received in any way whatsoever is received according to the limitations
of that which receives." These statements are clearly true, and they clearly
apply to every category of reality. Wherever there are existing realities,
or wherever there are wholes and parts, or wherever something passively
receives an action by another, these principles apply.
Still another example of the same sort is "Two things equal to a third
are equal to each other," a statement upon which the whole of
mathematics depends. 3 Of course such propositions do not enter directly
into the arguments of the sciences; they do not provide explanations
proper to the special subject matters, but instead form part of the
background. 4 Nonetheless they are recognizably indispensable first
208 From Observables to Unobservables
The burden of the first part of this essay has been to show how the
intelligence starts from observables and moves from them to an
understanding of unobservables. The position defended is realist, and
consequently we argued that philosophy bears on regularities derived
from sensory data. We have maintained that the origin of our knowledge
is singular entities, since they are what actually exist and what actually
affect the sensory powers. As a consequence of the mind's dependence on
sensory data for the origin of its concepts, we do not possess concepts that
are unrelated to observable properties of singular physical entities.
Furthermore, if the first propositions that we know are derived from
experience and observation, then they are affirmative and have what in
the past has often been called "existential import"; or as the medieval
Aristotelians would say, they have "personal supposition." One might
state the matter still another way and say that such propositions have
"referents." We also maintain that negative propositions have "existential
import" insofar as they say something about the real world; for one must
keep in mind that negation presupposes affirmation. Thus, if negative
propositions are to be formed at all, they must deny in some manner or
other that which something else has affirmed and therefore is taken to
exist. That is how a negative proposition has "existential import." Using
the Aristotelian vocabulary, we may say that affirmative propositions that
are non-supposing (that do not have referents or existential import) are
false. Only a negative proposition can be true when non-supposing, that
is, when it is not applicable to some particular entity or class of entities
"out there." On that account, the square of opposition for the real sciences
is that which Aristotle presented, and it may not be denied that the
systematic disciplines (we except mathematics) do indeed deal with
propositions that have personal supposition or existential import or
referents.
But if this is the position one takes in regard to propositions in general,
it is also the one he must take in regard to first principles. That is to say,
first principles in genuine disciplines cannot be of such a nature as to deny
their relation to real, individual things; they cannot be propositions that
deal with concepts alone in separation from the things to which the
Analytic Propositions 209
Therefore we shall take up the issue and present what we think is the
position of the ancient Greeks, a position which says that although
analytic propositions are true by definition, and although subject and
predicate signify the same thing, the propositions themselves are neither
conceptual nor linguistic identities. First, however, we must undertake
some considerations on necessity in things; for if analytic propositions are
not linguistic or conceptual identities, and if somehow they depend on
realities, then in one way or another there must be necessity in contingent
entities. So to begin, let us look once more at the examples of common
principles we introduced earlier. 7
It is not hard to see why the principle of non-contradiction has to be
true, for "exists" and "does not exist" are absolutely incompatible in the
real world. The issue is not just one of words or an identity of concepts,
for non-existence is the absence "out there" of something existing. First
we note the difference between existing and not existing and then we
attach the two names. Nor is it much more difficult to see that "Every
whole is greater than any of its parts," is also necessarily true and that
"whole" and "part" are attached to extra-mental realities. And since these
statements originate from our experience of real things, they are first
obtained by an induction from observation, even though the ease with
which we see them puts them, paradoxically, outside the list of inductions
we most quickly recognize. To be sure, the inductions themselves do not
show that the propositions are necessary; yet we nonetheless recognize
that they are. Their necessity must come, therefore, from the realities that
ground the inductions, and the necessity must be accessible to the mind
through something other than the inductions themselves.
At this point we must note apropos "The whole is greater than any of
its parts," that there is a tendency for some to think that "whole" and
"part" can be used for any arbitrary association whatsoever, even an
association that is wholly mental. Grand Central Station and a patch of
blue sky do not, to take an example, constitute a whole, except insofar as
"whole" has been given an arbitrary extension that does not respect the
original attachment of the name. Convention first imposes the words
"whole" and "part" upon physical units that are integrated out of what we
recognize as subunits constituting the complete units themselves.
Sometimes the units are amounts of homogeneous stuffs that are divisible
into homogeneous subunits, while at other times they are made up of
heterogeneous parts. But in either case it is upon such realities that we
impose the names "whole" and "part." A random collection such as an old
212 From Observables to Unobservables
beer bottle, an egg shell, a tree branch, and a bird bath, even though they
are all found in the same back yard and make a "pile," they cannot be
integrated into a whole, except in the most incidental way. Thus, to extend
the names "whole" and "part" to anything whatsoever we wish to consider
as "united" whether or not they are in the world outside is to distort the
language. It is to ignore the restraints placed upon the intelligence by the
first impositions of words according to which the words are attached to
independently existing realities. Put another way, we must respect the
noetic and hence linguistic limitations imposed upon us by the realities
that are signified, and we must not think that we may without sacrifice
extend a name to whatever suits our fancy. Contrarily, if the imposition
of words were purely arbitrary, if it were completely ad libitum, then the
mind would indeed construct what it knows.
Perhaps it is unnecessary to say that our discussion of first principles
does not include practical rules that tell us how to design, make, produce,
or otherwise cause or carry out some work of human invention. As we
said when speaking of defmitions, works of human origin are excluded
from our considerations, an exclusion that applies here, too; for
disciplines that tell us how to produce or do something provide
immediate, practical principles that are unlike the kind of first principles
nature provides for investigative, purely theoretical sciences. On the other
hand, since practical sciences presuppose natural entities either as
materials or as agents, they make use of some first principles that belong
to the natural sciences and are not properly their own. But putting that
aside, let us face the basic issue directly and ask how the necessity in
things that founds analytic propositions is accessible to us? How do we
obtain necessary propositions from contingent entities? What is our
ground for denying Hume?
As the reader well knows, beginning with John Locke and David
Hume, philosophers have been inclined to think that, since natural entities
are contingent, the mind cannot derive necessary propositions from
sensory impressions. A necessary proposition about a contingent reality
is taken to be a contradiction in terms; so that would seem to be the first
point to consider: how can there be necessary propositions about
contingent realities?
To start, let us note that there is a difference between the contingent or
Analytic Propositions 213
'
'
they can come to be and pass away. And if that is what contingent entities
are, then it seems that necessary entities must neither come to be nor pass
away and so be everlasting .. Such would seem to be one meaning of
"necessary," and certainly there is a clear opposition between "contingent"
and "necessary" as they are so described. And were that the the sense of
"necessary" we must use, then clearly only the conceptual can be in any
way everlasting; nothing in nature meets the description. We must not,
however, overlook another distinction that nature imposes on us.
If one starts not with entities but with properties, then he can see
another kind of contingency. Man, for example, is able to sit or to stand,
to walk or to run, etc. because he has operational capacities that are open
to alternatives. On that account we say that sitting and standing are contin-
gent-men contingently sit and contingently stand. Now this notion of
contingency does not have to do with the coming to be or passing away
of the entity itself, but with the coming to be and passing away of a
property in it. The opposite of this sort of contingency is a determination
inherent in the constitution of a thing which does not allow such
alternatives. We are not, for instance, able to opt for flying instead of
walking. Our operational capacity is limited, determined, to the action of
walking; it is not open to flying. In short, there is a necessity in things
which is none other than a determinate orientation to one property,
operation, or characteristic rather than to more than one.
An example of still another kind of contingency is skin color in the
human species. Skin color is not determined by man's nature; for if that
were so, we would all be alike, according to one pigmentation or another.
Yet skin color is not separable in a physical way from the individual, as
are standing and sitting. The contingency, then, is on the universal level,
insofar as the constitution of man which one conceives to be distinct from
color and other properties does not imply one pigmentation to the
exclusion of others. In contrast to these contingent properties stand others
which, although they belong to contingent entities, are not variable within
the species in either of the ways described above. Consider, for example,
a photon, which, physicists say, cannot stop moving at 300,000 kilometers
per second as long as the photon exists. To arrest it is to cause it to cease
to be. Moreover, it is not the case that some photons travel at the named
speed while others travel at one that is lesser or greater. Hence, although
the photon is a contingent entity, the relation between it and its motion is
necessary. That is to say, as long as they exist photons necessarily move
at 300,000 kilometers per second. But photons are theoretical entities;
Analytic Propositions 215
Though the point has already been made, it is worth repeating that not
every natural entity has a fixed set of properties which admits of no
variation; for what we have said about certain inanimate substances is not,
as everyone knows, true of organisms. Individuals within a species
obviously can vary, a state of affairs apparently not possible with
chemical elements. 11 Horses, for instance, can vary in size, color, shape,
running abilities, hair length, etc., without thereby belonging to different
species; and what is true of them is true of other organisms as well.
Nonetheless, there are certain attributes and certain operational capacities
without which one does not have the species, which means that some
attributes are contingent and others are necessary. There is a difference,
then, between the animate and the inanimate: the former have both contin-
gent and necessary properties.
The existence of necessary relations in contingent entities should not
be denied on the ground that individual organisms can depart from the
norm. Man, we say, is an animal that is able to speak; yet some
individuals are born dumb and remain forever so. But to hold therefore
that man is not necessarily a loquacious animal is to miss the point; for the
necessary relation is between the species-kind and the property that is
consequent upon the kind. Accidents can and do happen to individuals
both before and after birth, accidents that prevent the exercise of the
normal functions; and the ability to suffer accidents is indeed still another
kind of contingency. Yet, as we said, such states of affairs do not allow
us to say that man is both a loquacious animal and a non-loquacious
animal, for we cannot hold that both muteness and the ability to speak are
states natural to the species in the sense that they follow on it. Neither are
they contingent in the manner of alternatives, or in the manner of a
contingent property. In short, we recognize dumbness as abnormal and
hence as not rooted in the human constitution but as coming instead from
the capacity the materials have to be affected by foreign influences during
the generative process.
But this amounts to still a third kind of contingency, one that stems
from the inability of the constitution of the organism to prevent exterior
agents from introducing unwanted modifications into it. Such a
contingency is unlike both the capacity for alternatives and the possibility
of the species to receive more than one kind of property within its genus,
for instance, skin color. Thus, although accidents can and do happen to
individuals, we are nonetheless entitled to say that a necessary relation
exists between, say, the rational nature which defmes man and his ability
Analytic Propositions 217
To begin, let us note that analytic propositions are twofold, those that
are common, and those that are proper. Analytic propositions are common
when they are used by or presupposed to every science; but they are
proper when they pertain to a determinate species or genus of reality and
cannot be extended beyond that. An illustration of a common principle is
"Every whole is greater than its part," while an instance of a common
principle is "Man is a reasoning animal," a proposition that will be
discussed more at length later.
If we now ask the question, how do we know that the common
principle, "Every whole is greater than any of its parts," is necessarily
true, the answer has to be: we see its necessity as soon as we know what
wholes and parts are. That is not to say we see the proposition to be
necessarily true because in conceiving the subject we conceive the
predicate, nor is it to say that as soon as we know to what realities the
names are attached we know the proposition to be an identity, a tautology.
An analytic proposition is neither a conceptual identity nor a linguistic
convention. Rather, we may rightly say that we know the proposition to
be necessarily true as soon as we know what the terms mean, provided
218 From Observables to Unobservables
that "knowing what the term means" is taken to say that in knowing the
meaning of the terms we actually get at what the realities are. And though
in the case of common principles such as those which served as
illustrations earlier, there is no difference between attaching the name and
knowing what the reality is, such a state of affairs does not always obtain,
a point we made at length in our chapter on definitions. Indeed, in the
case of proper principles outside mathematics, we do not know what
things are until after we have named them and investigated them at some
length. Yet analytic propositions are true by definition; and that, of
course, is what has occasioned both Kant's and the analysts' views. Put
another way, because our first definitions of common realities suffice to
tell us what things are, we precipitously tend to think that every necessary
proposition comes either from a noetic identity of subject and predicate
or from the arbitrary attachment of the name to the object signified. The
consideration of proper principles, however, shows how far from the truth
such a confusion is.
The classic example of an analytic proposition that is a proper and not
a common principle is the statement, "Man is a rational animal,"
"rational" having to be understood as "able to deduce conclusions from
premisses (or theorems from axioms)." 13 At first sight the example does
seem to be an identity in Kant's sense; it seems to be a statement in which
the predicate declares what we actually think in the subject and therefore
comes more from the intelligence than from induction and observable
realities. But the first impression is deceiving. Certainly the proposition
is an identity in the sense that what is signified by the subject term is
identical to what is signified by the predicate; the referents, one might say,
are the same. Yet there is an important difference, for the term "man" does
no more than name the entity; it does not signify distinctly what that sort
of animal is. In short, because the word "man" properly signifies an
unobservable substance, what man actually is cannot be known except
through the predicate, which is taken from an activity known by
experience, thereby revealing the character of the underlying substance.
Again: the predicate signifies more clearly because it comes from a reality
known by observation, both internal and external, whereas the interior of
the substance is known by inference from them. That, of course, is
precisely what Kant overlooked, and it is why he thought the predicate is
already conceived in the subject. Thus, although the subject and predicate
terms signify the same reality, the proportions of the subject and the
predicate to the human mind are considerably different, so different that
Analytic Propositions 219
the mind cannot extract the predicate from the conception of the subject.
Certainly an analytic proposition is not of the form, xy is y.
The example, "Man is a reasoning animal," does present a certain
difficulty insofar as our familiarity with ourselves tends to make us look
on the defmition as nominal; it seems to be the phrase by which the name
of the species is attached. But a truly nominal defmition of man would be
"two-legged, featherless, TV watching animal," or something similar, in
which the describing traits would not be liable to confusion and would be
attained by external observation. Furthermore, when we understand "able
to reason" in a thorough way, we see in it why man is political, why he is
artistic, as well as why he has many other properties, none of which can
be read from a merely nominal definition. In short, the defmition of the
species man does what a systematic definition is supposed to do: it shows
why certain properties follow necessarily on his constitution. 14
It would seem to be plain by now that analytic propositions reflect the
mode ofknowing peculiar to the human mind, namely, that unobservables
(substances especially) must be known through observables (properties),
and the generic must be known (systematically, distinctly) before the
more specific. On that account the subject does not already- upon being
named.;_tell us what its predicates are; the subject does not contain them
noetically, though it does ontologically. Analytic propositions that are
proper principles, therefore, assign systematic definitions to their subjects;
nominal defmitions do not suffice. A systematic defmition, if it is
taken-as it ordinarily is-to be constituted by a category or genus along
with a difference that determines the species, cannot be formulated for
categories which are themselves first and most common and have nothing
more common over them. Although they are seen to be necessarily true
as soon as we know what the terms signify, what they mean, the
statements that contain them are not systematic formulations in the
ordinary sense but instead are circumlocutions; for common principles
cannot be stated in terms of a common genus divided by a difference that
is not intrinsic to the genus. And a single term applied to more than one
of them is necessarily equivocal, although in an orderly, non-incidental
way. Yet the awkwardness of speaking about such common categories in
circumlocutions does not prevent the mind's knowing what they are in
themselves.
And so to speak generally, we may say that an analytic proposition is
one in which the predicate enters into the definition of the subject in some
way, a formulation that covers both common and proper principles. We
220 From Observables to Unobservables
must add that we cannot say the subject enters into the definition of the
predicate, though we cannot discuss the point here. Such a proposition is
not analytic but a priori synthetic, to use a term favored by Kant.
Furthermore, a systematic definition of a property demands seeing the
cause of the property in the subject; hence until such causes are known,
no analytic proposition about properties is truly possessed. 15 So to repeat:
the predicate of an analytic proposition must belong in some way to the
defmition of the subject; yet such statements are not redundancies or
tautologies or pleonasms, in which subject and predicate signify the same
thing without any noetic difference.
are described or defined in terms that are common and related to the very
constitution of the human species; that is, to that in man of which he is not
the cause. But we need the examples to illustrate the point. 16
I. Closing considerations
Notes
3. A more generalized form of the proposition, one not tailored to the category of
the measurable, is: "Two things which are the same as a third are the same as each
other."
4. We shall say more about the metaphor, "background," when we talk about
hypotheses.
6. Critique of Pure Reason, tr. Norman Kemp Smith (New York: St. Martin's
Press, 1965), p. 48.
7. Henry Veatch has written a thoroughgoing critique of Kant's doctrine, leaving
nothing undone. His exposition is superior to anything I might do here, and I
recommend that the reader see his Two Logics, (Evanston, Illinois: Northwestern
University Press, 1969), chapter III.
8. What we shall say here about analytic propositions, propositions that have been
called per se nota is not new and will be familiar to many. We must take up the
issue, however, to relate such propositions to observation.
9. The terms "contingent" and "possible" are often taken as synonyms.
10. We need not say that we are excluding flavors, colors, and other properties
that come from the introduction of foreign substances into water to make a
solution or suspension.
II. The issue of the necessary relations among properties is discussed at length
in ourS&M.
12. It would seem that isotopes are not a problem; the reader will not find the
differences they represent to cause trouble.
13. The reader ought to see John D. Beach, "The Act of Analysis," The Thomist,
Vol. 44, No. I, Jan. 1980. This is a fine article and covers aspects of analytic
propositions we cannot take up here.
14. One must not take "rational" in some vague sense equivalent to an ambiguous
"intelligent," a word too often applied to non-human animals and covering many
cognitive operations that are not intellectual in the human sense of that term.
15. Actually seeing the cause-effect relation between rational and the ability to
speak, etc., entails a detailed understanding of the nature of the intelligence, an
understanding that is not had simply on the basis of our spontaneous familiarity
with our own internal processes.
16.The position we would defend regarding the relation of a property to the
subject for purposes of definition is only in a small way similar to the Kantian
view. We would agree with Kant that the concept of the subject does not include
in itself the concept of the predicate, even though the subject must be seen to be
the cause of the property signified in the predicate.
17. Rosenberg, Op. cit., pp. 17- 18.
18.It seems to us that this example is offered as an instance of a proposition that
is per se in the fourth mode, to use the Aristotelian description of the type. That
mode is the one in which the formality under which the subject is seem is that of
cause in regard to the predicate; the cause-effect relation is evident in the example
given.
224 From Observables to Unobservables
19. For a fuller discussion of our ability to define natural species we refer the
reader to S&MS.
Chapter 18
Recapitulation
Having made our case as well as we are able, perhaps it is
advantageous to recapitulate what we have done in the chapters that have
gone before. The center of focus is of course the understanding-
particularly among those who are realists-that human intellectual
knowledge depends necessarily and in a fundamental way on the sensible,
observable traits of natural entities, not only for its understanding of the
sensible traits themselves, but also for its understanding of the
unobservable realities to which we constantly have recourse, not only in
the sciences, philosophy, and theology, but in our ordinary discourse as
well.
We began by posing the problem and then pointing to our commonly
accepted starting point, namely that we must reduce what we say to facts,
which are either empirically given or are inferred from empirical origins.
But that required that we first of all point to the properties of things that
are admittedly observable, and then show that the mind's first movement
to an unobservable depends on a causal relation between the observable
and the unobservable. Once unobservables come to be known to exist,
they can then be compared to observables on the basis of the similarities
and dissimilarities the we see in them.
Expanding on points already made, we saw the constitutional
dependence of the rational intelligence on observables reflected in our use
of analogies and metaphors, as well as signs and symbols, the principal
signs being words. Because language is our ordinary means of
communication, it must be carefully examined so that we can see the first
meanings of words, which ordinarily are those that are attached to
observable traits, either external or internal. Having looked at these issues
we then moved to broader considerations on myth and its role in the
development of human learning, after which we showed the errors of
those who claim that observations are theory-laden. Finally in Part I we
226 From Observables to Unobservables
argued that the human intelligence cannot have innate or a priori ideas or
concepts.
Beginning with Part II, we looked at regularities as the observable
foundations of systematic sciences, moving from there to the regularities
that are the principles of the natural sciences. Following that with some
preliminary reflections on philosophy, we next looked at its principles,
which are analytic propositions and defmitions. That terminated our
labors.
BIBLIOGRAPHY
Atheneum, 1962.
Eddington, Sir Arthur. New Pathways in Science. Cambridge: At the
University Press, 1947.
Eliade, Mircea. Myth and Reality. Harper Torch book. New York: Harper
& Row, Publishers, 1963.
Feyerabend, P.K. Against Method. London: NLB, 1975.
Fodor, Jerry. "Observation Reconsidered," Philosophy of Science, 51,
(1984).
Galileo Galilei. /1 Saggitore. Reprinted as "Two Kinds ofProperties," in
Philosophy of Science, ed. Arthur Danto and Sidney
Morgenbesser. Meridan Books. New York: The World
Publishing Company, 1960.
Gordon, Cyrus H. "Canaanite Mythology," in Mythologies ofthe Ancient
World. Ed. Samuel Noah Kramer. Anchor Books. Garden City,
New York: Doubleday & Company, Inc., 1963.
Graves, Robert. The Greek Myths. Baltimore: Penguin Books, 1955.
Hanson, Norwood Russell. Patterns ofDiscovery. Cambridge University
Press, 1958.
Huygens, Christian. "The Wave Theory of Light," in Classics in Modern
Science. Ed. William S. Knickerbocker. Boston: Beacon Press,
1962.
Halliday, David and Robert Resnick. Physics. New York: John Wiley &
Sons, Inc., 1962.
Hamilton, Edith. Mythology. A Mentor Book. New York: The New
American Library,
Holton, Gerald and Duane Roller. Foundations of Modern Physics.
Reading, Massachusetts: Addison-Wesley Publishing Company,
Inc., 1958.
Jacobs, Noah Jonathan. Naming Day in Eden. New York: The Macmillan
Company, 1958.
Jammer, Max. Concepts of Force. A Harper Torchbook. New York:
Harper and Brothers, 1962.
Kant, Immanuel. Critique ofPure Reason. Tr. Norman Kemp Smith.
New York: St. Martin's Press.
Prolegomena to any Future Metaphysics. Library of Liberal
Arts. New York: Bobbs Merrill Company, Inc., 1950.
Kirk, G.S. The Nature ofGreekMyths. New York: Penguin Books, 1974.
Koyre, Alexander. Newtonian Studies. Cambridge, Massachusetts:
Harvard University Press, 1965.
Kuhn, Thomas. The Structure of Scientific Revolutions. Chicago: The
Bibliography 229
E G
F I
Q definition of, 61
imitate observables, 63-64
qualitative differences, 185-186 similarity relation, 33ff
quantitative approach to science, singular,
146-147 not object of science, 126
Quine, W. V. 0., 1-2, 47 sound, see wave
speech, modified vocal sound, 63
R Spiegelberg, Herbert, 53ff
spontaneous movements, 116
realist position, presupposed, ii state and activity, 46
recognition, primary and statue, Ill
secondary, 108 Stromberg, James S., 127n6
regularities, 124-126 subjective, 55
examination of examples, 220- substance, unobservable, 27-28,
221 105
regularities and experience, 124- known through observables,
126 27-28
Reichenbach, Hans, 115nl syllogism, 164-166
relations as unobservables, 26-27 systematic considerations
relations of predicability, 171 bear on regularities, 121
relative opposition, 23 systematic definitions, 197-199
Roller, Duane, 146n3 as explanatory, 198
Rogers, Eric M., 37n6 symbol
Rorty, Richard, i, 150-161 as distinct from sign, 60
Rosenberg, Alexander, 206 as likeness, 61
critique of, 220-221 implies human making, 61
s T
with, 12
source of knowledge of, 11-13
v
Vander, Arthur J., 37-38
Veatch, Henry B., 170nl3
voice as a sign, 62
w
Wallace, William, 136n2
Wartofsky, Marx W., 4-5
wave, see light
whole and part, 217
Williams, Dudley, 197-198
words,
as pseudo-concepts, 66
cognates as obscure, 71
extensions of, 68-69
functions of, 195-196
kinds of existence signified,
162
many meanings of, 67-69
order of signification, 69-73
signify by convention, 63
signify observables in
relation
signify concepts, 62
to unobservables, 29
that from which and that
upon which, 74-75
..