Sunteți pe pagina 1din 23

History of Psychology

2010, Vol. 13, No. 1, 224

2010 American Psychological Association


1093-4510/10/$12.00 DOI: 10.1037/a0017442

THE PSYCHOLOGY OF THINKING BEFORE THE


COGNITIVE REVOLUTION:
Otto Selz on Problems, Schemas, and Creativity
Michel ter Hark
University of Groningen
Otto Selz has been hailed as one of the most important precursors of the cognitive
revolution, yet surprisingly few studies of his work exist. He is often mentioned in
the context of the Wurzburg School of the psychology of thinking and sometimes
in the context of Gestalt psychology. In this paper, it is argued that Selzs emphasis
on the role of problems and schemas in the direction of thought processes and
creativity sets him apart from the program of the Wurzburg School. On the other
hand, by developing a theory of thinking that is exclusively at the intentional level,
Selz also differs from psychologists that take physics as a model for psychology,
such as the Gestalt psychology of Wolfgang Kohler. Special emphasis is given in
this paper to Selzs use of the concept of problem or task and the concept of the
schema. It is further argued that the concept of the schema is the result of Selzs
adaptation of the theory of relations as developed by the philosopher Meinong. The
paper begins with a sketch of Selzs life that ended so tragically.
Keywords: Selz, problem solving, schema, creative thinking

Otto Selz is not a household name in the historiography of 20th century


cognitive psychology. Apart from Robert Woodworth (1938), most of the standard histories and experimental psychology texts of the time that comprehensively
covered all other major strands of cognitive psychology do not mention him
(Boring, 1950; Brett, 1953). Woodworth (1938) referred to Selz seven times in his
chapter on thinking. A few years before the cognitive revolution took place,
George Humphreys (1951) Thinking appeared, in which for the first time a full
chapter was devoted to Selzs achievements. By the 1960s, when Allen Newell,
J. C. Shaw, and Herbert Simon (1958) had pointed out the similarities between
their Logic theorist and the work of Selz that had come to them via Adriaan de
Groot (1946), the first English translation of one of Selzs articles appeared
(Mandler & Mandler, 1964). The tributes paid to Selz by eminent scientists like
Newell et al. (1958), Jean Piaget (1947), and the philosopher who put problem
solving on the agenda of philosophy, Karl Popper (1928/2006), would still earn
him no place in the history of 20th century psychology. He was not mentioned in
the list of great psychologists (Watson, 1963). Commemorating Selzs birthday,
the Dutch psychologists Nico Frijda and De Groot (1981), both of whom had
known Selz during his exile in Amsterdam during the Second World War, edited
a book which includes, apart from an essay by Simon, translations of chapters of
his major work, as well as a translation of Selzs synopsis of his massive writings
Michel ter Hark, Faculty of Philosophy, University of Groningen, The Netherlands.
Correspondence concerning this article should be addressed to Michel ter Hark, Faculty of
Philosophy, University of Groningen. Oude Boteringestraat 52, 9712GL, Groningen, The Netherlands. E-mail: michel.ter.hark@rug.nl

OTTO SELZ ON PROBLEMS, SCHEMAS, AND CREATIVITY

(Selz, 1924).1 With some of his work having become accessible in the English
language one would have expected more attention to Selz, but again, if he is
mentioned at all in the history of early German psychology, it is only briefly and
subservient to other goals, such as the program of the Wurzburg School or Gestalt
psychology (Ash, 1998; Kusch, 1999).2
This article, then, tracks Selzs innovations in the field of cognitive psychology from the earliest stirrings in Wurzburg. Following a brief description of Selzs
life, the article focuses on the concept of the schema and the role that the concept
played in Selzs theory of reproductive and creative thinking. Full attention is
given to Selzs lifelong philosophical commitments, such as his embrace of the
theory of relations and complexes of Alexius Meinong and Carl Stumpf, as well
as his critical elaboration of the work of the psychologists Karl Buhler and George
Elias Muller. In closing, I discuss Selzs tense relation with Gestalt psychology.
Life and Work

Selz (18811943) lived much of his life in seclusion, cherishing the tranquility he needed to develop his epistemological, psychological, and pedagogical
ideas. Only a passport photograph has remained of him. In his scientific work,
Selz was increasingly marginalized owing to his unremitting criticism of colleagues and also to his formidably complex style of writing. Selz came into
conflict with proponents of the sciences of mind and culture, the Geisteswissenschaften, who blamed him for endorsing a mechanist view of man. Seeking to
reconstruct psychological phenomena on the basis of their elements, Gestalt
psychologists considered him an atomist. Closely allied to the Wurzburg School
of the psychology of thinking (Denkpsychologie), he did not shrink from launching frontal attacks on the ideas of some of its members. Aside from one pupil,
Julius Bahle, who closely collaborated with Selz, and the Dutch scholars De Groot
and F. W. Prins (1951), who applied his ideas in respectively psychology and
pedagogy, he never founded a school, and after 1933 his name disappeared almost
completely from the German psychological literature.
Selz was born on 14 February 1881, in Munchen, as the son of a well-to-do
partner in the banking house Frankel and Selz. His father was one of the many
children of a country rabbi and had married the daughter of a rich vinegar
manufacturer. Selzs mother was descended from a family of Spanish Jews with
a long tradition of refined culture: Well-read and a proficient pianist she cared
little for cooking or household matters (Seebohm, 1981, p. 1). The only other
child was a daughter four years younger than Otto.
At the Royal Ludwig Gymnasium in Munich, Selz was a highly gifted student
whose academic achievements were so brilliant that he was excused from the oral
part of his final examinations in 1899. The following comment by his examiners
on his way of writing would have great predictive value: Among his examination
papers the German essay in its effort at completeness allows subordinate matters
1

More recently a selection of his writings appeared including not only his work on problem
solving but also his completely neglected essays on the phenomenology of perception (Metraux &
Herrmann, 1991).
2
In more recent publications (Michel ter Hark, 1993, 2004a, 2004b, 2007), I presented a
historical reconstruction of (young) Poppers roots in Selzs psychology of thinking.

TER

HARK

to achieve prominence at the expense of emphasis on the main theme, but it is


fluently written (Seebohm, 1981, p. 2). As Seebohm comments, this tendency to
get down to the final detail, with the risk that the reader in the end loses the thread
of the argument, was to remain a characteristic of all Selzs scientific writings.
Pressed by his father into a legal career, Selz studied law, but in spite of being
admitted to the bar in 1908 he felt no vocation for an occupation as lawyer and
asked for his name to be deleted from the list again. All his life, he was to bemoan
the loss of five years. He had simply done what his father demanded. It is
hypothesized by Seebohm that his highly ritualized contacts with his father lay at
the basis of his lifelong difficulty in engaging in personal relationships. He never
married. In controversies with his colleagues about scientific matters, such as with
Narziss Ach and Muller, and even when plagiarism was at stake, such as with
Kurt Koffka, Selz refrained from personal attacks and instead defended his work
with vigor. In a letter to Bahle, Selz mentions that Buhler once said to him:
Where you have hacked about, grass will never grow again. He adds, I have
always hit out only in factual matters and fought shy of personalities (Seebohm,
1970, Appendix, p. 24).
After finishing his studies of law, Selz went his own way. In the meantime,
he also studied philosophy with Theodor Lipps in Munchen and with Stumpf in
Berlin. In 1909, he took his PhD in philosophy at the University of Munchen. In
his dissertation The psychological theory of thinking and the transcendence
problem (Selz 1910a), Selz was concerned with the question, much debated in
the 17th and 18th centuries, whether there is an objective world outside our
consciousness and how to know this world. He rejected the empiricist theory of
immanence and instead argued that the hypothesis of an objective world
independent of consciousness most satisfied our cognitive needs. Selz did not
pursue this philosophical theme after 1910, yet his critical reading of John Locke,
George Berkeley, and David Hume undeniably supplied his later arguments
against the association psychology of Muller with logical rigor.
After earning his PhD, Selz moved to Bonn to do experimental investigations
in the laboratory of Oswald Kulpe. Both Kulpe and Buhler were among his
subjects, and he probably attended some of their seminars. These investigations
resulted in his first major work, The Laws of Ordered Thinking. An Experimental
ber die Gesetze des geordneten Denkverlaufs. Eine experimenInvestigation (U
telle Untersuchung; Selz, 1913).3 Taking his cue from the theory of imageless
thought, developed in the Wurzburg School, Selz, according to Kulpe, made a
significant step forward in the psychology of thinking (Seebohm, 1970, p. 15). In
fact, the drift away from the program of the Wurzburg School was more radical
than Kulpe would acknowledge. Already perceptible in 1910, the incipient rift
between Selz and the Wurzburg School became more obvious in the wake of a
devastating review of Ach (Selz, 1910b).4
3

The first two chapters served as Selzs Habiliation at the Faculty of Philosophy in Bonn in
1912. The third chapter on the important notion of the total task (Gesamtaufgabe) was drawn from
a paper presented in 1912 at the 5th Congress for Experimental Psychology in Berlin at which
Koffka was also present. See further below on the relation Selz-Koffka.
4
See below footnote 6.

OTTO SELZ ON PROBLEMS, SCHEMAS, AND CREATIVITY

In the First World War, Selz served in the army at the Western-front, and after
having been wounded, he was decorated with the Iron Cross in 1917. After the
war, he returned to Bonn for several years as Privatdozent. He lectured on the
psychology of knowledge, the history of philosophy, and the theory of the genesis
of cognitive functions. The publication of his second major work in the psychology of thinking, On the Psychology of Productive Thinking and Errors (Zur
Psychologie des produktiven Denkens und des Irrtums; Selz, 1922) was postponed
owing to the First World War. Once it was published after the War, Selzs
intellectual prestige was incontestably on the increase, and in 1923 he was called
for the chair of Philosophy, Psychology and Pedagogy at the Handelshochschule
in Mannheim. The practical orientation of the Handelshochschule more or less
forced Selz to work also on the practical applications of his psychology. In spite
of the abstract nature of his psychology, he succeeded in pointing out its relevance
for pedagogy, in particular its usefulness in fostering intellectual achievements, a
project undertaken with some of his pupils (Selz, 1935).
Meanwhile, the Psychological Institute of Mannheim headed by Selz, who
became Rector of the Handelshochschule in 1929 1930, flourished and after receiving the Ius Promovendi, the first dissertations on the psychology of thinking and
pedagogy began to appear, among them Bahles cognitive psychological investigations of musical composition (Bahle,1930). All this came to a sobering halt after
January 30, 1933, when Hitler was appointed Chancellor of Germany. The Ministry
of Culture and Education issued an edict in the spring of 1933 demanding that Otto
Selz resign his work, the official reason being the maintenance of security and order
in Baden. Selz was no longer allowed to continue neither his teaching activities nor
his research at the Institute. On October 25, the Handelshochschule was finally closed,
and the Institute became incorporated within the University of Heidelberg. Since
Heidelberg had made no arrangements for Selz, the official reading asserted, Selz was
unceremoniously stripped of his career and livelihood. The truth of the matter was, of
course, Selzs being a non-Aryan.
Most expelled psychologists left Germany and migrated to the United States. Not
so Selz. He led a withdrawn life in Mannheim, where the opportunities to do
experimental work being gravely diminished, he threw himself into purely theoretical
work on the construction of the phenomenal world. After the Reichskristallnacht, he
was caught and deported to the concentration camp of Dachau from which he was
released again in December 1938. In May 1939 he finally emigrated to the Netherlands, first to Bilthoven and then to Amsterdam, where he lived in a small apartment
in the Cliostraat. The one desirable outcome of this shameful episode was that Selz
came into contact with the Dutch pedagogues Philip Kohnstamm and De Groot. Selz
taught at the Amsterdam Teachers Seminar (Nutsseminarium) on psychology and
pedagogy and participated in scientific discussions at the Faculty of Psychology,
hugely enriching the field of psychology. After the German invasion in May 1940,
Selz corresponded with Koffka, who had emigrated to America, but despite Koffkas
efforts nothing came out of it.5 He declined the offer of his Dutch friends to find a
hiding place for him, replying that the Iron Cross he had won during the First World
War would surely protect him. He would not be spared the horrors of the Holocaust,
5

This is explained in Beckmann (2001, pp. 1315).

TER

HARK

though. In July 1943, he was caught again by the Nazis and sent to the deportation
camp of Westerbork. A postcard stating that he wanted to give courses in Westerbork was the last sign of his life. On August 24, he was put on train Nr. DA 703
to Auschwitz. He either died in transit from suffocation or exhaustion he was
suffering from heart problems or was murdered by being sent into the gas
chambers.
The Wurzburg School

The work of the Wurzburg School has been so thoroughly discussed in review
and criticism that I shall content myself with a bare enumeration of its principal
phases, zooming in only on the work of Buhler.6 The school was founded by the
philosopher Kulpe. Kulpe was born in Candau, Courland, one of the Baltic
Provinces, and now part of Lettland. After completing the Gymnasium, he studied
history in Leipzig. There he came into contact with Wilhelm Wundt who diverted
him to philosophy. He became Wundts assistant, and, in 1894, was called to
Wurzburg as ordinarius for both philosophy and aesthetics in succession to
Hans Volkelt, founding there, by 1896, a psychological laboratory which became,
next to Leipzig, the outstanding institute of Germany. Kulpe worked there for 15
years and for that reason his pupils and followers are referred to as the Wurzburger School of the psychology of thinking. In 1909, he moved to Bonn and
next to Munich, where he died in 1915.
The psychology of thinking had its beginning with the publication of papers
by August Mayer and Johannes Orth (1901) on the qualitative nature of association, followed by Karl Marbes (1901) experimental study on judgment. These
researches of Kulpes graduates were given direction and point by his determination to show that it is impossible to analyze thought into sensory elements,
thereby discrediting Ernst Mach, who refused to recognize any other form of
existence than sensations and to contribute to the experimental analysis of thought
as a category sui generis, thereby departing from Wundt, who believed thought
was not amenable to scientific treatment. As Kulpe summarized the first achievements of the School, these were largely negative: the traditional contents of
consciousness, sensation, feeling, and imagesthe very substance of Wundtian
psychologyproved inadequate to account for the intellectual processes of
thoughtful association and judgment. In one of Marbes experiments, a subject
had to lift two weights and to judge which is the heavier. The subject reported that
different sensations and images accompanied the task, but that they were not the
elements underpinning the judgment. The judgment simply came, and usually
right, but with nothing in the subjects consciousness to indicate why they were
judgments. Marbe coined the term Bewusstseinslage (state of awareness) for this
phenomenon and used it as a mere negative description for states of consciousness
that were certainly present but seemed to defy closer analysis.
Achs (1905) careful experimentation resulted in the more precise description
of the important influence of the task on thinking and the determining tendencies which issued from it, as had already been foreshadowed by Kulpes
6
See for instance, Humphrey (1951), Odgen (1951), Titchener (1909), and more recently
Kusch (1999) and Janke & Schneider (1999).

OTTO SELZ ON PROBLEMS, SCHEMAS, AND CREATIVITY

writings. He was the first to state clearly that in the experimental afterperiod, the
period of introspection, observers often reported that a complex conscious content
was simultaneously present as knowledge. This imageless presentation of a total
knowledge-content is termed by Ach, Bewusstheit, which was translated by
Edward Bradford Titchener (1909) as awareness.7
The notion of imageless thought was further analyzed by Buhler. Buhler
joined the Wurzburger School around 1906 and remained there until 1909, the
year he followed Kulpe to Bonn. In Wurzburg, he wrote his most important
contribution to the psychology of thinking, his Habilitation, Facts and Problems
of a Psychology of Thought Processes (Tatsachen und Probleme zur einer
Psychologie der Denkvorgange; Buhler, 1907, 1908). Buhlers subjects had to
answer questions for which there could be no cut-and-dried answer, such as, Was
the Pythagorean theorem known to the Middle Ages? The subjects then gave a
retrospective account of the processes intervening between the presentation of the
task and their answer. Buhler concluded from these protocols that there were basic
unanalyzable units in the thinking process, which should simply be called
thoughts (Gedanken). Thoughts, he argued, are independent of any sensory
factors. In particular, images are too fragmentary and sporadic to be essential to
thought. Put more strongly, the occurrence of images is neither necessary nor
sufficient for the occurrence of thoughts. Buhler next distinguished between
different types of thoughts. It is here that the influence of the philosophers
Edmund Husserl and Meinong became apparent. The influence of Meinong is
evident in Buhlers discussion of thoughts in the sense of pure intentions
(Wissen um etwas). Subjects did not report any mental images at all, yet their
consciousness was not devoid of mental occurrences. They often reported a
certain direction upon the objects of thinking or a kind of order. Buhler analyzed
such thoughts in terms of determinations of logical space (logische Platzbestimmtheiten). In such cases, Buhler argued, the object of thinking is determined not
through its intrinsic properties, but indirectly through its relations with other
objects that are part of the same logical space (Buhler, 1907, p. 358). In a footnote,
Buhler referred to Meinongs discussion of indirect determinations. Selz would
pick up this idea and use it as a starting-point for his theory of successive stages
in problem solving.
The Wurzburg psychologists paid only marginal attention to the active
processes of thinking and seemed to be harking back to the static classificatory
schemes of the early associationists. Ach dubbed the term determining
tendencies for the directive influence of the task on the outcome of thinking.
This determining tendency was distinct from, and opposable to, associative
tendencies. According to Ach, determining tendencies explained the ordered
and purposeful character of thought processes; they ruled out irrelevancies and
7
In later work (Ach, 1910), Ach was concerned with the experimental investigation of
voluntary acts themselves. The debate with Selz to which I referred earlier concerned the role
voluntary acts could have in counteracting failed attempts or simply mistakes. According to Ach,
strong voluntary acts were the only means to counteract failures. To this, Selz objected that failures
could be avoided simply by focusing on the structure of the task (i.e., awareness of the total task).
There was no need to overexert oneself as Ach maintained. In his final response to Selz, Ach (1911)
honored some of his objections but stuck to his earlier view. But it led him to reconsider carefully
the pitfalls surrounding the use of introspection.

TER

HARK

prevented chance stimuli from distracting the course of thought processes.


They accomplished this by favoring those reproductive tendencies that were in
line with the aim of the subject. For instance, the presentation to a subject of
the numbers 6 and 2 may yield a reaction of 8, 4, or 12, depending on whether
the task prescribed was adding, subtracting, or multiplying. The aim of the
subject determined the reaction rather than the stimuli or their associative
tendencies.
Selzs Theory of Relations and Complexes

To a certain extent, Selz continued the experimental tradition initiated by


Kulpe and the members of his school. Like his predecessors, Selz experimented with task and stimulus words. Yet, already at this stage of his work,
he made significant alterations and improvements in the experimental design
of the study of thought processes. One of the drawbacks of the experimental
design of his predecessors was that the presentation of the task before the
display of the stimulus word led to a constant focus on the task and to
insufficient attention to the stimulus word. In this way, the exploration and
construction of the task by the subject could not be studied in detail. To
prevent one-sided focus upon the task and attendant stereotyped solutions,
Selz offered task and stimulus word at the same time. In this way, Selz could
vary task and stimulus word independently of each other and, in particular,
change the task at each new trial. Subjects were now forced to construct the
task at each new trial. In a single trial, the course of thinking could be studied
in its entirety. To study thought processes by making use of retrospection, the
nature of the task needed to be relatively simple. On the other hand, it had to
be complex enough in order to trigger genuine thinking, rather than simply
reproduction. Selz solved this dilemma by selecting task and stimulus word so
that his subjects really had to overcome a difficulty. One type of task was to
look for a whole of which the stimulus word was a part, or, to look for a part
belonging to the stimulus word. Another type was more conceptual. Thus,
subjects were asked to look for the superordinate, the subordinate, or the
coordinate of a certain stimulus word.
Selz introduced not only semantic relations (Sachbeziehungen), like part
of . . . , subordinate to . . . , in the investigation of thought processes, but
he also provided a general, phenomenological theory of relations supplemented with a theory as to how relations are extracted from the data, that is,
a theory of abstraction. Armed with this theory, Selz defined directed thinking
as the capacity to extract relational structures (Beziehungsganze, Sachverhaltnisse) from concrete experience and to store this information into memory so
that it could be retrieved for later purposes. Although Selzs theory of
relational structures contained new elements, it is also evident that he heavily
relied on philosophical sources, in particular Meinongs Hume-Studies (HumeStudien; Meinong 1877, 1882), his later theory of objects (Gegenstandstheorie; Meinong, 1904), Stumpfs Appearances and psychic functions (Erscheinungen und psychische Funktionen; Stumpf, 1907), and Husserls Logical
Investigations (1901). Because the influence of the former on Selz seems to

OTTO SELZ ON PROBLEMS, SCHEMAS, AND CREATIVITY

have been greater, it is instructive to have a brief look at Meinongs view of


abstraction and relations.8
Meinong

Meinong first of all argued against Humes nominalism, how little mere
association without abstraction was able to achieve. To give one of his examples,
suppose we have a round piece of paper, or a millstone before us, and we call this
shape. Because ex hypothesi we cannot think of a circle in abstracto; it follows
that when we happen to see, on a walk through the country, a square cornfield, it
would never occur to us to call to mind that shape and to give the name shape
likewise to the cornfield. On the view that we are capable of thinking in abstracto,
however, all would be plain sailing. Meinong further pointed out that not association, but the self-conscious activity of attention is the main originator of mental
life.
Meinong published his theory of relations in a separate volume of his
Hume-Studies, yet the connection with the theory of abstraction was evident.
Relations were important for abstract thinking, but the mainstream of British
empiricists had failed to account for them properly. Following Locke and Hermann Lotze, Meinong argued that relations involve an activity of the subject, that
is, comparing. The mental operation of comparing involves the shifting of
attention from one of the terms of the relation to another. Comparison implied
abstraction, according to Meinong, because to compare objects means to compare
them in certain respects and, hence, necessarily to leave out other aspects from
consideration.
By arguing that relations were brought into being by acts of comparison,
Meinongs stance was wholly psychological. His mature view of what relations
really are, however, was partly determined by Christian von Ehrenfelss (1890)
ber Gestalt qualitaten). Von Ehrenfels
famous article On Gestalt qualities (U
introduced the term Gestalt quality to denote the perception of form or melody
that, though based on sensory elements, can in no sense be reduced to such
elements. Meinong agreed with von Ehrenfelss negative argument but objected
to the term Gestalt property, and instead proposed the term complexion, or
what would probably be better to call complex. Taking von Ehrenfelss notion
of a foundation upon which Gestalt properties were based as his cue, Meinong
added objects that were founded on them. A founding object, he called an inferius,
as it is of lower order with respect to its superior object. Such a superius is an
object of higher order. Objects of higher order have an intrinsic lack of
independence. For instance, diversity can only be thought of in relation to
differing terms. Diversity and all objects of higher order are based on others as
indispensable presuppositions. The term complex was used by Meinong to refer
to the whole of inferius and superius. Thus, a melody is a complex, which
contains both the founding objects (the various notes sung or played) and the
founded objects, the new factor emerging when the notes are taken together.
8

The two most elaborated accounts of Selzs theory, Humphrey (1951) and De Groot (1965)
leave out the Meinongian background. The same applies to shorter treatments of Selz, for instance
Kusch (1999).

10

TER

HARK

Meinongs theory of complexes implied an intimate connection with the


theory of relations. Indeed, if one takes the meaning of the term relation
sufficiently broadly, every relational fact is also a complex fact; relations cannot
exist where there are only atomic facts, hence no relation without complexion.
The Role of Complexes in Association

Selz first of all elaborated his ideas about complexes in the context of rather
elementary forms of reproduction, such as the recognition of a melody, pattern
recognition, and reproduction of words. Like Meinong, he maintained that such
perceptual and quasi-perceptual complexes are not aggregates in which the
specific (temporal or spatial) order of the elements abcd is indifferent, but
objects-taken-in-relation (Selz, 1913, p. 97). Knowing or recognizing these temporal and spatial relations is also at the basis of the reproduction of such
complexes. Thus, the recognition of the national anthem depends on the specific
temporal order of the individual tones rather than on the tones themselves. Played
in a different key, it will have no disturbing effect upon recognition. The specific
temporal arrangement of the sounds (or the spatial arrangement of visual elements) is a property of the complex (Komplexbestimmtheit), not of its elements.
Associations between perceptual complexes, therefore, are reciprocal associations
between complexes.
It is important to note here that Selz duly acknowledged the contribution made by
other researchers, notably Muller (1913), Muller and Schumann (1894), and Kulpe
(1893). His main point in his argument with Muller, however, was that his findings
and arguments were not decisive to prevent supporters of a constellation theory from
reducing properties of complexes to associational interactions of their elements, as
was evident in the writings of William James, Ebbinghaus, and Theodor Ziehen (Selz,
1920, p. 216).9
More complicated cases of the reproduction of complexes are discussed
below under the head of the notion of complex completion (Komplexerganzung). Here, I will discuss only Selzs complex-completion on the basis of a
schema. Selz opened his discussion of schemata by suggesting that in many cases
of memory retrieval, subjects already know that the information at hand is a piece
of a larger complex. Indeed, they even often know what kind of complex to which
the piece belongs. Giving the example of a candidate in an oral examination trying
to remember the Melanchton, and who is assisted by the examiners giving the
first three letters Mel, Selz explains, The awareness of the word sought is
changed from the awareness of a not yet further specified concrete word to the
awareness of a word beginning with Mel . . . We have to conceive of the genesis
of this awareness like the way in which the empty scheme of a concrete word is
partly filled out by the insertion at its beginning of the sounds spoken in
anticipation . . . (Selz, 1913, pp. 113114). A schema is formed, which organizes
itself around a few outstanding elements. The schema is incomplete in the sense
that it contains a gap (Leerstelle), but it is this schematic whole that finally
produces the outcome.
9

By not mentioning this background Kuschs (1999, p. 66) portrayal of Selz is slightly
misleading in suggesting that the theory of complex association is one of the major features of his
program.

OTTO SELZ ON PROBLEMS, SCHEMAS, AND CREATIVITY

11

Relational Facts

Selzs next important step was to apply the laws of complex-completion to


reproductive memory processes and thought processes. Meinongs common treatment of complexes and relations surely paved the way for Selzs generalization of
the laws of complex-completion, yet the further elaboration and, in particular, the
adduced empirical evidence for the great explanatory power of this move were
unique. Selz defined relational facts as the fact that certain objects stand in a
certain relation (Selz, 1913, p. 142).10 A relational fact can neither be reduced to
the objects that stand in the relation nor to the relation itself, because it includes
both.
Put otherwise, they display the same sort of lack of independence as
Meinongian complexes. For instance, the relation of similarity that exists between
two objects can only be thought of in relation to objects that are similar.11
Relational facts, then, are objects of higher order.
The notion of relational fact was meant to be encompassing and to include
mathematical, physical, and psychological facts, yet Selz was mostly concerned
with relational facts as psychological objects. Relations between psychological
objects, for example, experiences of color, Selz argued, are themselves items in
the psychological realm. Qua relational fact, such psychological objects, are not
independent items like an individual mental image, but dependent on other
psychological objects.12 By virtue of this intrinsic lack of independence, relational
facts could be present within a subjects phenomenal field without being noticed
(bemerken). Selz here relied on A. Grunbaums (1908) influential experiments on
abstraction. These experiments showed that subjects perceived simultaneously
presented figures without becoming aware of the fact that they were similar.
Positive abstraction, however, was also demonstrated. In such cases, Selz
argued, the subject noticed the fact that certain objects stand in a certain relation.
This capacity, he maintained, is crucial for the discovery of new information.
What the subject discovered in such cases, Selz believed, is something general, a
relational structure in itself, which subsequently might be transposed to all sorts
of concrete problem situations. Selz, therefore, made a distinction between, on the
one hand Erkenntnis or Einsicht and, on the other hand, knowledge of relational
facts already at ones disposal, hence dispositional knowledge.13 Selzs protocols made clear that whether knowledge of relational facts was merely actualized
10
Selz took the term Sachverhalt from Stumpf (1907) and changed it into Sachverhaltnis.
Stumpf took the term from J. Bergmann in his work General Logic. Bergmann was a pupil of Lotze
the lectures of which were also attended by Stumpf. On the history of the term Sachverhalt, see
Barry Smith, 1989.
11
Selz did not use the term objects of higher order and instead called relational facts
mitgegeben. He distinguished between various classes of relational facts. One class of relational
facts concerns relations between objects, e.g. similarity, difference, and size. Another class concerns
relations of order, such as temporal co-existence, and spatial contiguity. Yet other classes concern
symmetrical relations and asymmetrical relations.
12
Selz speaks here of dependent moments within conscious experience (unselbstandige
Momente; Selz, 1913, p. 147).
13
In 1913, Selz speaks only of knowledge, but in 1922, he argues that what he calls knowledge
is identical with what the members of the Brentano school as well as Wolfgang Kohler call insight.
See further below.

12

TER

HARK

or newly abstracted from experience, in both cases processes of complexcompletion were involved. For instance, a subjects coming to know that death
is coordinate with sleep is not to be explained as an aggregate of separate
reproductive tendencies tied by association, but as the awareness of a relational
fact, that is, the fact that death can be conceived as coordinate with sleep.14
Selzs Theory of Schematic Anticipations

The philosophers Meinong, Stumpf, and Husserl, as well as the psychologists


of the Wurzburg School, contributed to the study of thought processes, yet no
general theory of thinking was forthcoming. Admittedly, this seemed not to have
been their ambition. For instance, Henry J. Watt and Ach were focused more on
describing certain characteristic patterns of thinking, and especially in showing
the role played by respectively the task and determining tendencies, than offering
an encompassing theory. One reason for this lack of ambition may have been the
dominance of the association theory of mind. Indeed, as Selz himself was the first
to point out, Achs view of the role of determining tendencies in the process of
thinking was still committed to the principle of association. By assuming that a
directive and selective factor, that is, a determining tendency, gives to reasoning
and thinking their most fundamental character, Ach dethroned blind association
but still subscribed to its basic tenets. In particular, determining tendencies acted
like separate factors.
Selzs goal, by contrast, was to develop a general theory of thinking that not
only explained the steps that lead from constructing a problem to achieving a
solution but also replaced the dominant associationist explanation of thinking by
means of substitution. Frequency and recency, similarity and contiguity, or any
other laws of association are inadequate to explain the particular kinds of linkage
and direction found in problem solving. Even the most sophisticated theory of
association, Mullers constellation theory, put forward as a reply to Achs
theory of determining tendencies, was inadequate to this task.
The Constellation Theory: Muller and Selz

To appreciate Selzs critique of Muller, it is necessary to explain the latters view


in more detail. Mullers constellation theory intended to serve as a correction of
associationism, according to which the mind is an unstable arena of sensations and
mental images perpetually jostling for monopoly of attention. None of the elements,
however, ever achieves this monopoly finally, so that fleeting experiences remain
subliminally present. Accordingly, a written or heard word, for example, farmer,
prompts all the mental images that were familiar to the subject from earlier experience. Assuredly, association psychology acknowledged that by no means all representations associated with farmer returned upon its recurrence. To explain this
incontrovertible fact, the theory claimed that only the association most strongly linked
to the initial mental image held permanent sway. This lawful repression of weaker
associations by the strongest associations is the only directive factor in thinking.
Muller, however, made a significant contribution to this classical picture. When
confronted with a task and stimulus word, the subjects mental state did not consist
14

The example is from Selzs protocols. (Selz, 1913, p. 29).

OTTO SELZ ON PROBLEMS, SCHEMAS, AND CREATIVITY

13

simply of associative links between the separate elements of his answer, but in
addition all the elements will be drawn toward the task. By thus functioning as a
directive representation, the task intensifies associative bonds of individual elements that belong to the content of the task, while simultaneously inhibiting those
associations that divert from it, the result being that all active representations correspond to it. In other words, the task brings about a constellation, which favors
associative bonds of elements appropriate to the task.
Selzs objection proved devastating, however. He argued that even when the
constellation is fairly strong pointless errors are not excluded by the theory. For
instance, the principles of the constellation theory would not exclude the pointless
answer line to the question What is the whole with respect to page?15 For it is
very likely that among the concepts triggered by whole would also be line, which
is the whole with respect to individual words and syllables. Moreover, it is very likely
that page triggers line as well. In his 1924 synopsis, Selz formulated his critique
even more forcefully: Suppose, for instance, that the task calls, not for the generic
concept, but for a coordinate, parallel notion. Since for the majority of concepts a
parallel notion can be found, it would follow that the majority of notions would be put
in heightened readiness. Among these there would always be notions that are indeed
associated with the stimulus word - without for that reason being related to it in the
manner prescribed by the task (Selz, 1924, p. 29). A constellation theory at best
limits the range of possibilities, but it still allows many senseless errors to occur. The
fact that such errors do not occur, neither in daily life nor in the lab, shows the
fundamental weakness of the constellation theory.
The Total Task

Selz contrasted his own complex-theory of thinking to the constellation


theory.16 The distinctive feature of the complex-theory was that the whole course
of problem solving was explained without invoking factors that lie outside the
psychological (phenomenal or intentional) realm. In addition, Selz eschewed the
principle of association, without at the same time having recourse to determining
tendencies or other mysterious effects emanating from the problem as Ach had
done. The important notion of the total task, already presented by Selz at the 5th
Congress for Experimental Psychology in 1912, served to explain problem solving as a process of tentative and phased transformations that were intrinsic to the
phenomenal material of the problem itself. In Selz, problem solving, therefore,
was essentially a dynamic process in which the construction of the total task
(Gesamtaufgabe) leads to a schematic anticipation of the goal to be achieved
(Figure 1). From the start, the solution of the problem is represented by the subject
as a more or less indeterminate part within the relational structure presented by the
15
See Selz, 1913, p. 281. Different versions of this critique can be found in Selz, 1924, and
Selz, 1927.
16
The term complex-theory was coined by Selz in his discussion with Muller. Muller,
however, did not mention this and called his own theory a complex-theory, thereby ignoring the
differences with Selz. Koffka (1925b), in his turn, first called Mullers theory a complex theory
(p. 543) and then Selzs (p. 569), without indicating the differences. This terminological confusion
prompted Selz to conclude: My complex theory is only opposed to a constellation theory, but is
itself a Gestalt theory (Selz, 1926, p. 168).

14

TER

HARK

problem situation

goal situation

Figure 1. Diagram of problem solving after Selz, 1924. The figure on the left hand
represents a problem situation. The symbol stands for a relation, for example,
part of. The symbol A stands for an item of knowledge; the black square
represents the unknown item of knowledge. The figure on the right hand represents
the solution of the problem situation. It relates to the figure on the left hand as the
completion of a schematic anticipation of a relational whole.

total task initially only in terms of its general features but made concrete by
further transformations. The dynamic interplay between superordinate and subordinate processes triggers as a whole the final solution. The solution is not related
to the problem as one heterogenous element to another, mediated by a third
equally heterogenous element, for example, Mullers goal representation. Rather,
they relate as the schema or blanket of a relational fact, with one unknown
element or relation, to the completed relational fact (Selz, 1922, p. 370).
The theory of schematic anticipations, however, was far more complex and
dynamic than this diagram suggests. For what the diagram does not represent, but
what is essential to its conception, is the dynamic role of the total task and the
attendant notion of transformation.
By his specific experimental design discussed above, Selz discovered that task
and stimulus word do not become effective separately, but that, even before the
actual process of finding a solution for the problem has begun, they are integrated
into a unitary whole, that is, the total task. The total task has the structure of a
question, or problem, so that Selzs theory of thinking has to be conceived as a
theory of problem solving. This is also evidenced by his innovative use of the term
method of solution (Losungsmethode) for explaining the subsequent course of
thinking.17 The process of problem solving, Selzs protocols demonstrated abundantly, did not begin before the subjects comprehension of the problem situation
as a whole. For instance, one observer, confronted with the task Cancer (Krebs):
Cause or effect, reported:
As soon as Id read the word Cancer, I conceived it in a zoological sense, and
had a weak mental image. I then went on and now read the task properly for the
first time. Now the thing struck me as comical, because I was still thinking of the
animal and also had this mental image. I told myself there must be a solution all
the same. Cancer must have another meaning. Then I became conscious that in fact
there is another meaning and eventually the meaning cancer as a disease came
clearly to mind . . . (Selz, 1913, p. 222)

The stimulus word Krebs is initially understood in a zoological sense, as crab,


but then it is seen in the light of the task. Once this integration of task and stimulus
17
He used this term for the first time in 1912 (Selz, 1912) and complained that it was simply
taken over by Kohler in his study of apes without crediting him (Selz, 1926, p. 194).

OTTO SELZ ON PROBLEMS, SCHEMAS, AND CREATIVITY

15

word was achieved, the subject noticed a conflict. The subject understood that
there was something wrong within the situation as given. In this particular case he
experienced this gap as comical, but in other cases it may be more troublesome. What the total task appears to call forthe cause of Krebs in a zoological
sense does not make sense. This predicament induces all sorts of control
processes, which, in their turn, lead the subject to explore the possibility of a
proper solution. The outcome is that the stimulus word has been misconceived.
Not until a meaningful total task had been constructed, that is, one in which
the meaning of the stimulus word Krebs is adapted to the meaning of the task and
taken in the sense of a disease, the process of solving the problem sets in.18 As
Selzs careful analysis of the protocols bears out, subsequent phases in the course
of problem solving were subordinated to the formation of the total task and, most
importantly, may change in character in the light of the subjects knowledge of the
whole. This was especially evident with cases of problem solving Selz called
successive actualizations of knowledge (Selz, 1913, p. 45, p. 62). These
obtained when the subject was unable to circumvent the problem immediately.
Selz observed that with more complicated problems, subjects typically first
searched for more general knowledge about the problem situation, because it was
more familiar than specific knowledge.
What he discovered at the same time was even more important. Finding a
general property of a solution to be attained, consisted in transforming (Transformation) the original problem. Consider this example. A number of subjects
confronted with the task to find a coordinate concept of railway platform
proceeded by first searching for another part of the concrete spatial whole which
included railway platform (Selz, 1922, p. 142). A railway platform is part of a
number of physical constructions, such as railroads, a train station, and so forth,
which together make up rail transportation. By first considering railway platform under the same superordinate concept to which the other parts of rail
transportation belong, subjects could solve the task of finding a coordinate
concept by searching for one of these parts. This detour via a superordinate concept
was a transformation of the original problem, which was defined more generally by
Selz: . . . the substitution for the task of another task, through whose solution the
original problem is also to be solved (Selz, 1922, p. 41). The substituted task was
a subproblem and, hence, subordinate to the original problem.19 If subjects did not
succeed in the preceding transformation of the original task, they may take
recourse to other kinds of transformation. For instance, they may seek to transform the original problem into the subordinate problem of searching for an object
that fulfills the same purpose as the objects referred to by the stimulus word. This
general description of a property of what is sought for may in its turn be replaced
by a more specific description, such as to look for an object that among other
forms of traffic fulfills the same purpose (Selz, 1922, p. 143). Via a series of such
18
The evidence for the formation of a total task came from direct reports of the subjects as well
as from more indirect sources. Thus, before the task was understood, Selz observes, subjects often
took no more than cursory notice of the stimulus-word; once the task being understood, the subjects
awareness of the meaning of the stimulus word took a more concrete shape; reports of mental images
occurred only after the task was understood.
19
Selz speaks of a subproblem (Unteraufgabe; Selz, 1913, p. 87).

16

TER

HARK

transformations, or reformulations of the original problem, general knowledge


becomes successively more and more specific.
Typically, the genesis of a problem solving process displays several transitions deviating from the mainand straightline to achieving the goal. At these
junctures, subjects learn from mistaken transformations and may revert to the
stage of one of the preceding reformulations of the problem. For instance, subjects
may revert to the stimulus word and pay closer attention to it, for example,
pronouncing it several times. Or they may set an earlier subproblem anew.
According to Selz, this retrogression (Rucklaufigkeit) of problem solving is
characteristic of goal-directed processes (Selz, 1913, p. 197; Selz, 1922, p. 122).
To be sure, subjects do not return to the earlier phase as if nothing has happened,
rather they learn from mistakes and seek to proceed in a different way from the
starting position.
The discovery of the total task implied a crucial difference with the associationist paradigm. Within Selzs conception, the relation between the different
stages of transformation is determined by the meaning or content of the problem
and not by an intermediary associative tie, the description of which does not
depend on meaning at all. In modern parlance, Selzs interpretation of the
protocols is at the purely psychological or intentional level. For instance, if the
task is to indicate the cause of frost, subjects may insert frost qua effect in
the abstract schema of a causal relation. From then on, however, if the original
problem is to maintain its meaning, the attempt to further concretize it is no longer
arbitrary. On the contrary, the sought-for concept, that is, the gap, has to conform
itself to the partly concrete and specified structure. The process of problem
solving is intentionally driven.
Creative Thinking

When a person is confronted with a new problem, Selz argued, actualizing


dispositional knowledge is no longer an option. Instead, there is a need to
actualize familiar solving-methods (Mittelaktualisierung). Applying familiar solving-methods to new problems, changing material, or in different situations, he
averred, is the route to understanding creative thinking.20 Selzs thinking here is
too complex to deal with in this essay, so I will confine myself to only one of his
heuristics, the abstraction of means. Selz described this heuristic: . . . the result
of abstraction is schematically anticipated as an, as yet unknown, solution-method
which will effect a known aim, namely to bring about a determinate partial result
(Selz, 1924, p. 55). An interesting subclass of this method is the method that sets
in when no knowledge of applied methods or other memory structures are
available. In such a case, chance may help. The solution here is not achieved
top-down, that is, by means of the schematic anticipation of what is sought for, but
bottom-up: one stumbles on the solution and then the schematic anticipation is
triggered and the solution recognized as a solution. This, in essence, was Selzs
explanation of a number of scientific discoveries. As Selz explained, Benjamin
20
He distinguished two main cases of heuristics: methods used for the finding of solving
methods (Mittelfindung) and methods for applying solving methods (Mittelanwendung). The first
group consist of three chief cases: the methods of routine actualization of solving methods,
abstraction of solving methods, and productive use of previously established abstractions.

OTTO SELZ ON PROBLEMS, SCHEMAS, AND CREATIVITY

17

Franklin had conceived the plan to draw off the electric charge of a thundercloud
by using the principle of arc discharge. His goal was to bring down the lightning
from the clouds. He knew that in order to realize this goal he had to make a
connection between the cloud and the earth. His problem was to find a means that
would establish the required connection. From the sight of kites being flown, Selz
explained, Franklin abstracted the fact that a kite may form a connection between
the earth and clouds, and this abstraction of solving methods may have led him to
actually sending up a kite on a wire.
Another example discussed by Selz is Michael Faradays discovery of electromagnetic induction (Selz, 1924). Faraday became interested in experiments of
rotating a copper disk underneath a suspended compass-needle. When this disk
was rotated rapidly, the needle was deflected in a manner quite inexplicable.
Faraday at once conceived the idea that the cause of this rotation was due to
electricity. For several years, however, he was unable to demonstrate the truth of
this assumption. But, in 1831, he began a series of experiments that established
forever the fact of electro-magnetic induction. His aim was to produce electricity
from a magnet. The partial result was to cause a deflection of a galvanometer
needle. Happening one day to be moving a magnetic core in and out of a coil
connected with the galvanometer, Faraday observed a deflection in the attached
galvanometer. This trifling occurrence, Selz argued, sufficed to make him see that
in the closed but uncharged circuit a current must have been generated by the
movement of the magnet. Selz argued that a specific event, when occurring by
accident, may prompt the scientist to abstract the appropriate means-end relation.
Accordingly, the fact that insights appear suddenly does not mean that they are
uncaused at an explanatory level. Indeed, the undeniably passive character of
sudden insights is best explained by assuming their being caused by the stubborn
persistence of engrossing problems (Selz, 1924, p. 65).
Selz and His Contemporaries: Gestalt Psychology

Selz did never partake in a research school as many of his more famous
colleagues, such as Buhler, Koffka and Kohler did, yet in a way his attitude
toward psychology was strikingly modern. Consider this letter to Bahle, November 23, 1935, when he had been dismissed from his position in Mannheim and had
to carry on his research all alone at home:
To be a theorist means first and foremost not to engage, like the old-style
humanists, in building ever new theoretical edifices but, like the great physical
scientists, to build on to the achievements of others, that is to say, to improve and
complete them but not to insist on changing everything at any price. The latter
together accomplish the building of a science, the former write one book each (like
Spranger and Charlotte B.). Our colleagues are more and more losing sight of this
goal and where Duncker is concerned I must at least give him credit for his
concern with the continuity of science (Seebohm, 1981, p. 7).

As we have seen, Selz continued the work of others (e.g., Meinong, Buhler,
Muller) and improved upon the theory of others. For instance, he could explain
many of Mullers findings in terms of his complex-theory. The same applies to a
topic I have not entered upon, Selzs explanation of Kohlers findings in animal

18

TER

HARK

psychology in terms of his theory of creative thinking.21 Selzs relation with


Gestalt psychology, however, is complicated owing to the debate over priority
with Koffka and to a lesser extent Kohler. On the one hand, there was a serious
overlap between Selzs theory of complexes and problem solving and Gestalt
psychology. Selz acknowledged the overlap and wanted the other party to do the
same. Instead, Koffka sought to set Selz apart by calling his theory mechanist and
behaviorist. On the other hand, there were deep differences of which Selz only
lately became aware. In this final section, I will first deal with the priority issue
and then move on to discuss the intellectual differences between Selz and Gestalt
psychology.
In the first half of the 1920s, Selz found himself involved in a controversy
over priority of ideas. Koffka, a leading contributor to Gestaltpsychology, wrote
a chapter on current psychology in Max Dessoirs Lehrbuch der Philosophie
(1925), which, although incontestably drawing on Selzs theories, never mentioned him at crucial places. Koffkas article provoked the outrage of Buhler who
accused him of having taken his Gestalttheory from Selz (Buhler, 1926). At the
same time, Selz responded by showing that Koffka had borrowed key concepts
from Selzs work and that his criticism of association psychology copied Selzs
own earlier rebuttal (Selz, 1926).
Koffkas article included a section on the refutation of the constellation theory
and a section on The Gestalt theory of thinking. Selz complained that he was not
mentioned, or only in passing, in the first two sections whereas it could easily be
demonstrated that Koffkas refutation exploited Selzian sources. Selz correctly
pointed out that Koffkas and Kohlers refutation was simply a terminological
revision of Selzian concepts, for example, Selzs concept of an aggregate of
elements was replaced by (the much better sounding) and-connections of
Kohler. At the same time, Selzs complaints seem a little paranoid, because at
other places both Koffka and Kohler did credit him for having refuted the
constellation theory.22
Things were different as regards the constructive part of Koffkas essay, the
Gestalt theory of thinking. Gestalt psychology replaced the law of reproductive
association by the law of Gestalt completion (Gestalterganzung; Koffka, 1925b,
p. 576), according to which every part of a Gestalt has the tendency to complete
the whole Gestalt. In the first edition of an earlier work, however, Koffka had
credited Selz for having formulated a theory that was similar to the one he
21

Selz, 1922, pp. 610.


As an anonymous referee pointed out, Koffka was supervised by Stumpf who in his work on
the perception of space and the perception of tone was clearly antiassociationist and emphasized the
role played by relations. In his article on Gestalt experiences and movement experiences, Koffka
(1915) criticized Meinong and the theory of his pupil Vittorio Benussi. In it, he gave the first clear
formulation of the Gestalt theorist point of view. Selzs critique, however, did not so much concern
the question as to the priority of having discovered the pivotal role played by relations. As pointed
out above, he relied on Meinong as well as on Stumpf. He even referred positively to an early
publication by Koffka (1912) and pointed out that his own conception of association by contiguity
in terms of complex-associations bore a resemblance with the view developed by Koffka (Selz,
1913, p. 129, Footnote). Selzs main point of critique was that he was not credited for his specific
refutation of the constellation theory, which was much more sophisticated than the theory of
association Stumpf was concerned with.
22

OTTO SELZ ON PROBLEMS, SCHEMAS, AND CREATIVITY

19

(Koffka) had outlined (Koffka, 1921, p. 176). But in the second edition, the
similarity was reduced to a mere formal similarity (Koffka, 1925a, p. 125). In
his reply, Selz painstakingly made clear that the formal similarity was, so to
speak, a mere rhetorical device for appropriating another ones ideas (Selz, 1926).
For instance, in outlining the law of Gestalt completion, Koffka made use of an
example that was virtually the same as the one Selz had used in illustrating his law
of completion on the basis of a schema. Thus, where Selz used the example of the
first three letters of Melanchton (see above), Koffka used the following example:
If one asks for the name of an animal that begins with Qual, then the answer
Qualle is extremely easy. Under different conditions, however, one will have
difficulties in proceeding from the word Qual to the word Qualle. In the first
case, the syllable Qual has precisely the character of the first syllable, but in the
second it is one syllable, a complete word (Koffka, 1925a, pp. 184 185). Despite
this similarity in content, Selz was not mentioned.
An important source of this controversy was a negative and biased review of
Selzs second volume on the laws of ordered thinking (Selz, 1922) by Koffkas
colleague, W. Benary (1923). Benary called Selzs theory a machine-theory in
which no place was accorded to the important phenomenon of insight that played such
a prominent role in Gestalt psychology.23 As Selz rightfully retorted, the similarity
between Gestalt psychology and his own complex-theory was much closer than the
use of ideological adjectives like mechanistic suggested, because Kohler himself
relied on Selzs theory of relational facts. Indeed, before his first great work in Gestalt
psychology appeared (Kohler, 1919), Kohler (1917, 1918) made use of Selzs insight
into the nature of relational facts, but rephrased it in slightly different terms. In his
study on chimps and chicken (Kohler, 1918), Kohler argued that the perception of
Gestalten, for example, two shades of color and of relations between colors has much
in common. In both cases, the individual colors attain an inner union, which is the
opposite of an association by contiguity, and which depends on togetherness or
structure-function (Strukturfunktion; Kohler, 1918, p. 16). He cited specific pages of
Selz (1913) in support of his theory. A few paragraphs further, he discussed the
application of structure-functions to the study of memory and again referred to Selzs
critique of the constellation theory and his treatment of spatial-temporal structures in
23

Selz complained that the points of contact between his theory and gestalt psychology were
obliterated by ideological phrases. For instance, both Benary and Koffka (1923) tried to distance
themselves from Selz by calling his theory mechanistic, even behavioristic. A striking example of
this ideological stance was Benarys rejection of Selzs detailed explanation of Kohlers findings
with chimps in terms of schematic anticipations (Selz, 1922). On the other hand, Selz observed that
to explain the findings of a theory by means of another theory was the ideal of verification (Selz,
1926, p. 168). Moreover, Koffkas own description of animal psychology relied on Selzian sources.
For instance, in discussing how an animal recognizes a branch of a tree as a stick to obtain a piece
of fruit which is beyond reach, Koffka argued An open Gestalt, the way to the fruit, gets its
completion from another completed Gestalt, and that happens because a part of this Gestalt
undergoes a change (Koffka, 1925b, p. 578). A branch, part of a fixed Gestalt, that is, a tree, is
abstracted from this Gestalt and jumps into (hineinspringen) another Gestalt (bridge to the fruit).
As Selz pointed out, this idea was already put forward by him in 1912, where he spoke of the
completion of a schematic anticipation. In his diagrams of problem solving, the open Gestalten, or
gaps were even pictured black (Selz, 1926, p. 169). The metaphors used by Koffka as well as
Wertheimer (i.e., umwandeln, hineinspringen) were merely paraphrases of what Selz called the
method of means abstraction.

20

TER

HARK

his chapter on complexes (Kohler, 1918, pp. 38 39). In his much more famous book
on the mentality of apes (Kohler, 1917), he again argued that what associationist
theories of intelligent animal behavior cannot explain, is the grasp of a material,
inner relation of two things to each other (sachlichen, inneren Bezugs; Kohler, 1925,
p. 219). By relation he meant the interconnexion based on the properties of these
things themselves, not a frequent following each other or occurring together
(Kohler, 1925, p. 219).24 As Selz complained, Kohlers inner relations are nothing
other than his own relational facts (Selz, 1926, p. 193194).25 The notion of insight
was also the same as the concept of knowledge used in Selz (1913), namely, the
newly acquired consciousness of a relational fact (Selz, 1926, p. 193).
What became increasingly clear to Selz, however, was the difference between his
outlook on psychology and the outlook of Gestalt psychology. What he in particular
opposed was the radical top-down approach of Gestalt psychology. In rejecting the
bottom-up approach of association psychology and opting for the opposite view of the
primacy of the whole, Selz now argued (Selz, 1927), Gestalt psychology overlooked
the implications of a theory of schematic anticipations for a bottom-up approach. By
means of phased development, as exemplified by the notion of transformation, wholes
are constructed out of elements. However, the elements are not meaningless units
subject (only) to principles of spatial and temporal contiguity; rather, the elements are
themselves from the start determined by their place, that is, their system of relations,
in larger wholes. Selz increasingly came to see the implications of his constructive
bottom-up approach to problem solving for other fields of psychology, such as the
development of motor responses or skills. In his synopsis (Selz, 1924), he explicitly
applied his earlier theory to skills. Skilled behavior, he explained, displays a structure
in which the type of response given depends on the specific nature of the conditions
of elicitation. In contrast to a purely associationist system, with skilled behavior the
character of sensorimotor reactions is relevant to the entire structure. Selz even went
so far as to say that, . . . Perhaps our era is witnessing the beginnings of a biology
of the inner. Psychology thus enters the ranks of the biological sciences (Selz, 1924,
p. 73).
Biology of the inner was not meant as a reductive proposal, ultimately
leading to the disappearance of psychology as an autonomous science, but rather
a reminder of the functional, organic organization of adapted mental activities.
The biological approach to problem solving would be taken up by both Piaget
(1947) and Popper (1972), but it was missed by Gestalt psychology with its
emphasis on the relation between physics and psychology.
In Selzs completely ignored phenomenological writings on the construction of
our perception of the outside world, the difference between Kohler and Selz came to
the fore. It is interesting to note that Selz pointed out that the associationist explanation
of the perception of the phenomenal world had something in common with the Gestalt
approach. Both schools acknowledged only a physiological-genetic approach. To be
24
Selzs use of the concept of relational fact was already known by Gestaltpsychologists, in
particular Koffka, who was present at the same conference 1912 where Selz for the first time
presented his ideas on the total task and responded to him (Selz, 1912).
25
In a letter to Bahle dated January 14, 1937, Selz reported that Stumpf had told him personally
that the theory of relational facts outlined in Selz (1913) had been of formative influence on Kohlers
work on physical Gestalten. (Seebohm, 1970, Appendix, p. 32).

OTTO SELZ ON PROBLEMS, SCHEMAS, AND CREATIVITY

21

sure, association psychology argued that the main features of perceptual wholes, such
as unity, connectedness, or size, could be explained genetically in terms of the
association between psychic elements, whereas Gestalt psychology argued that it was
impossible to come to a synthesis on the basis of elements. Gestalt psychology
proceeded from the assumption of the primacy of the whole. Nevertheless, Selz
argued, Gestalt psychology explained the construction of perceptual wholes on the
model of the construction of physical wholes and objects (Selz, 1941). The laws of the
dynamics of Gestalten postulated by, among others, Kohler, therefore, are to be seen
as laws for the causal and genetic explanation of perceptual wholes. On the other
hand, Selzs theory of the construction of perceptual wholes is based on phenomenological laws. These laws concern pure phenomenological principles of perception
and as such must not be confused with physiological-genetic laws.
Despite his outdated and complicated style of writing, Selz is a strikingly modern
author. One of his most relevant contributions is to have shown, both in his work on
problem solving and on perception, the importance of investigating mental life at a
purely psychological level. As his painstaking study of protocols demonstrated,
human thinking can be described and explained as a series of processes directed by the
subjects zooming in on the structure of the task. From this intentional and phenomenal perspective, Selz could bring into focus a number of hitherto ignored conscious
processes such as the completion of schematic anticipations. To be sure, Selz did not
deny that there is a relation between the intentional and the physiological level.
Rather, his worry was that the dominant theories of his time were the result of
premature conceptual reduction of psychology to physiology, as was the case with
association psychology, or to physical models, as in Gestalt psychology. It seems to
be the case that explanations in science in terms of more basic (i.e., underlying)
processes have occurred after the more complex observations and theories were well
established. According to Selz, psychology is no exception to this rule.
References
ber die Willenstatigkeit und das Denken [On the voluntary act and
Ach, N. (1905). U
thinking]. Gottingen, Germany: Vandenhoeck & Ruprecht.
ber den Willensakt und das Temperament [On the voluntary act and
Ach, N. (1910). U
temperament]. Leipzig, Germany: Quelle und Meyer.
ber den Willensakt [On the voluntary act]. Leipzig, Germany: Quelle
Ach, N. (1911). U
und Meyer.
Ash, M. (1998). Gestalt psychology in German culture, 18901967: Holism and the quest
for objectivity. Cambridge, England: Cambridge University Press.
Bahle, J. (1930). Zur Psychologie des musikalischen Gestaltens, eine Untersuchung uber
das Komponieren auf experimenteller und historischer Grundlage [On the psychology of musical creation, an experimental and historical investigation of composing].
Leipzig, Germany: Akademische Verlagsgesellschaft.
Beckmann, H. (2001). Selz in Amsterdam. Der Denkpsychologe Otto Selz (18811943)
im niederlandischen Exil. Psychologie und Geschichte, 9, 327.
ber die Gesetze des geordneten Denkverlaufs. II. Zur
Benary, W. (1923). Selz, Otto: U
Psychologie des Denkens und des Irrtums [Selz, Otto: on the laws of ordered thinking. II.
On the psychology of thinking and error]. Psychologische Forschung, 3, 417 426.
Boring, E. G. (1950). A History of Experimental Psychology. New York, NY: AppletonCentury Crofts.

22

TER

HARK

Brett, G. S. (1953). History of psychology (R. S. Peters, Ed.). London, England: George
Allen and Unwin.
Buhler, K. (1907). Tatsachen und Probleme zu einer Psychologie der Denkvorgange. I.
ber Gedanken [Facts and problems of a psychology of thought processes I. On
U
thoughts 1908 idem, added On connections between thoughts]. Archiv fur die gesamte
Psychologie, 9, 297365.
Buhler, K. (1908). Tatsachen und Probleme zu einer Psychologie der Denkvorgange. II.
ber Gedankenzusammenhange. Archiv fur die gesamte Psychologie, 12, 123.
U
Buhler, K. (1926). Die neue Psychologie Koffkas [The new psychology of Koffka].
Zeitschrift fur Psychologie, 99, 145159.
Duncker, K. (1935). Zur Psychologie des produktiven Denkens. Berlin, Germany: Springer.
ber Gestaltqualitaten. Vierteljahrschrift fur WissenEhrenfells von, C. F. (1890). U
schaftliche Philosophie, 14, 249 292.
Frijda, N., & De Groot, A. (1981). Otto Selz: His contribution to psychology. The Hague,
The Netherlands: Mouton.
Groot de, A. (1946). Het Denken van den Schaker [The thought processes of the chess player].
Amsterdam, The Netherlands: N. V. Noord-Hollandse Uitgevers Maatschappij.
Groot de, A. (1965). Thought and Choice in Chess. The Hague: Mouton.
Groot de, A. D. (1981). Thought and choice in chess. In N. H. Frijda & A. De Groot
(Eds.), Otto Selz: His contribution to psychology (pp. 192256). The Hague, The
Netherlands: Mouton.
ber die Abstraktion der Gleichheit. Ein Beitrag zur PsycholoGrunbaum, A. A. (1908). U
gie der Relation [On the abstraction of similarity. A contribution to the psychology of
relation]. Archiv fur die gesamte Psychologie, 12, 340 470.
Hark ter, M. R. M. (1993). Problems and psychologism: Popper as the heir to Otto Selz.
Studies in History and Philosophy of Science, 24, 585 609.
Hark ter, M. R. M. (2004a). Popper, Otto Selz and the Rise of Evolutionary Epistemology.
Cambridge, England: Cambridge University Press.
Hark ter M. R. M. (2004b). The psychology of thinking, animal psychology and the young
Karl Popper. Journal of the History of the Behavioral Sciences, 40, 375393.
Hark ter, M. R. M. (2007). Popper, Otto Selz and Meinongs Gegenstandtheorie. Archiv
fur die Geschichte der Philosophie, 89, 60 78.
Humphrey, G. (1951). Thinking. London, England: Methuen.
Husserl, E. (1901). Logische Untersuchungen [Logical investigations]. Halle, Germany:
Niemeyer.
Janke, W. S., & Schneider, W. (Eds.). (1999). Hundert Jahre Institut fur Psychologie und
Wurzburger Schule der Denkpsychologie [A hundred years the institute for psychology and the Wurzburg school of the psychology of thinking]. Gottingen, Germany:
Hogrefe Verlag fur Psychologie.
Koffka, K. (1912). Zur Analyse der Vorstellungen und ihrer Gesetze: Eine experimentelle
Untersuchung [On the analysis of images and their laws: An experimental investigation]. Leipzig, Germany: Quelle und Meyer.
Koffka, K. (1915). Beitrage zur Psychologie der Gestalt-und Bewegungserlebnisse. III.
Zur Grundlegung der Wahrnehmungspsychologie. Eine Auseinandersetzung mit V.
Benussi [Contributions to the psychology of Gestalt experiences and movement
experiences III. On the foundation of a psychology of perception]. Zeitschrift fur
Psychologie und Physiologie der Sinnerorgane, 73, 1190.
Koffka, K. (1925a). Die Grundlagen der psychische Entwicklung [The foundations of
psychological development]. Osterwieck am Harz: Verlag von A. W. Zickfeldt.
Koffka, K. (1925b). Psychologie [Psychology]. In M. Dessoir (Ed.), Lehrbuch der
Philosophie. Bd. II. Die Philosophie in ihren Einzelgebieten (pp. 495 603). Berlin,
Germany: Ullstein.

OTTO SELZ ON PROBLEMS, SCHEMAS, AND CREATIVITY

23

Koffka, K. (1927). Bemerkungen zur Denkpsychologie [Remarks on the Psychology of


Thinking]. Psychologische Forschung, 9, 163183.
Kohler, W. (1917). Intelligenzprufungen an Anthropoiden. Abhandlung der Preusische
Akademie der Wiss. Physik.-Math. Klasse No. 1. Berlin, Germany.
Kohler, W. (1918). Nachweis einfacher Strukturfunktionen beim Schimpansen und
Haushuhn [Proof of Simple Structure-Functions in Chimps and the Chicken]. Berlin,
Germany: Verlag der Akademie der Wissenschaften.
Kohler, W. (1919). Die physische Gestalten in Ruhe und im stationaren Zustand [The
Physical Gestalten in rest and in Stationary State]. Erlangen, Germany: Verlag der
Philosophischen Akademie Erlangen.
Kohler, W. (1925). The Mentality of Apes. New York: Humanities Press.
ber die moderne Psychologie des Denkens. Internationale
Kulpe, O. (1912, June). U
Monatschrift fur Wissenschaft, Kunst und Technik, 1069 1110.
Kusch, M. (1999). Psychological knowledge. A social history and philosophy. London,
England: Routledge.
Mandler, J. M., & Mandler, G. (Ed.). (1964). Thinking: From association to Gestalt. New
York, NY: Wiley and Sons.
Marbe, K. (1901). Experimentell-psychologische Untersuchungen uber das Urteil [Experimental Psychological Investigations into Judgement]. Leipzig, Germany: Engelmann.
Mayer, A., & Orth, J. (1901). Zur qualitativen Untersuchung der Association [On the
Qualitative Investigation of Association]. Zeitschrift fur Psychologie, 26, 113.
Meinong, A. (1877). Hume-Studien I: Zur Geschichte und Kritik des modernen Nominalismus [Hume-Studies I: On the History and Critque of Modern Nominalism]. In
Sitzungsberichte der Kaiserlichen Akademie der Wissenschaften. Philosophischhistorische Klasse (Vol. 78, pp. 185260).
Meinong, A. (1882). Hume-Studien II: Zur Relationstheorie [Hume-Studies II : On the
Theory of Relations]. In Sitzungsberichte der Kaiserlichen Akademie der Wissenschaften. Philosophisch-historische Klasse (Vol. 101, pp. 573752).
Meinong, A. (Ed.). (1904). Untersuchungen zur Gegenstandstheorie und Psychologie [Investigation Into the Theory of Objects and Psychology]. Leipzig, Germany: J. A. Barth.
Metraux, A., & Herrmann, T. (1991). Otto Selz. Wahrnehmungsaufbau und Denkprozess.
Ausgewahlte Schriften [Otto Selz. The Formation of Perception and Thought Processes]. Bern, Switzerland: Verlag Hans Huber.
Muller, G. (1913). Zur Analyse der Gedachtnistatigkeit und des Vorstellungsverlaufes, III
[On the Analysis of Memory Activity and the Course of Mental Images]. Teil.
Zeitschrift fur Psychologie, Erganzungsband, 8, 567.
Muller, G., & Schumann, F. (1894). Experimentelle Beitrage zur Untersuchung des
Gedachtnisses [Experimental Contributions to an Investigationb of Memory].
Zeitschrift fur Psychologie, 6, 81190.
Newell, A., Shaw, J. C., & Simon, H. A. (1958). Elements of a theory of human problem
solving. Psychological Review, 65, 151166.
Odgen, R. M. (1951). Oswald Kulpe and the Wurzburger School. American Journal of
Psychology, 61, 4 19.
Piaget, J. (1947). La psychologie de lintelligence. Paris, France: Librairie Armand Colin.
Popper, K. R. (1972). Objective knowledge. An evolutionary approach. Oxford, England:
The Clarendon Press.
Popper, K. R. (2006). Karl Popper. Fruhe Schriften [Karl Popper. Early Writings]. (Troels
Eggers Hansen, Ed.), Tubingen, Germany: Mohr Siebeck.
Prins, F. W. (1951). Een Experimenteel-didactische Bijdrage tot de Vorming van Leerprestaties volgens Denkpsychologische Methode [An Experimental-Didactic Contribution to the Formation of Learning Achievements According to the Method of the
Psychology of Thinking]. Groningen, Djakarta: J. B. Wolters.

24

TER

HARK

Seebohm, H. (1970). Otto Selz. Ein Beitrag zur Geschichte der Psychologie (Doctoral
dissertation). [Otto Selz. A Contribution to the History of Psychology]. Universitat
Heidelberg, Heidelberg, Germany.
Seebohm, H. (1981). Otto Selz. In N. H. Frijada, & A. de Groot (Ed.), Otto Selz: His
Contribution to Psychology (pp. 115). The Hague, The Netherlands: Mouton.
Selz, O. (1910a). Die psychologische Erkenntnistheorie und das Transzendenzproblem
[The Psychological Theory of Knowledge and the Transcendence Problem]. Archiv
fur die gesamte Psychologie, 16, 1110.
Selz, O. (1910b). Die experimentelle Untersuchung des Willensaktes [An Experimental
Investigations of the Voluntary Act]. Zeitschrift fur Psychologie, 57, 241270.
Selz, O. (1912). Experimentelle Untersuchungen uber den Verlauf determinierter intellektueller Prozesse [Experimental Investigations into the Course of Determined Cognitive Processes]. In Bericht uber den V. Kongress fur experimentelle Psychologie
(pp. 229 234). Leipzig, Germany: Barth.
ber die Gesetze des geordneten Denkverlaufs [On the Laws of Ordered
Selz, O. (1913). U
Thinking]. Stuttgart, Germany: Verlag von W. Spemann.
Selz, O. (1920). Komplextheorie und Konstellationstheorie [Complextheory and Constellationtheory]. Zeitschrift fur Psychologie, 83, 211234.
Selz, O. (1922). Zur Psychologie des produktiven Denkens und des Irrtums [On the
Psychology of Productive and Reproductive Thinking and Error]. Bonn, Germany:
Verlag von Friedrich Cohen.
Selz, O. (1924). Die Gesetze der produktiven und reproduktiven Geistestatigkeit. Kurzgefasste Darstellung [The laws of cognitive activity, productive and reproductive: A
condensed version]. Bonn, Germany: Bouvier Verlag. Original published in 1981
Selz, O. (1926). Zur Psychologie der Gegenwart. Zeitschrift fur Psychologie, 99, 160 196.
Selz, O. (1927). Die Umgestaltung der Grundanschauungen vom intellektuellen Geschehen [The revision of the fundamental conceptions of intellectual processes]. Kantstudien, 32(1927), 273280.
Selz, O. (1935). Versuche zur Hebung des Intelligenzniveaus. Ein Beitrag zur Theorie der
Intelligenz und ihrer Erzieherischen Beeinflussung [Attempts at Increasing the Level
of Intelligence. A Contribution to the Theory of Intelligence and its Pedagogical
Consequences]. Zeitschrift fur Psychologie, 134, 236 301.
Selz, O. (1941). Die Aufbauprinzipien der phanomenalen Welt [The Fomation Principles
of the Phenomenal World]. Acta Psychologica, 5, 735.
Smith, B. (1989). Logic and the Sachverhalt. The Monist, 72, 52 69.
Stumpf, C. (1907). Erscheinungen und psychische Funktionen. In Abhandlungen der
Koniglich Preussischen Akademie der Wissenschaften Philosophisch-historische Abhandlungen, XIX, Band, II [Appearances and Psychic Functions]. Abteiling 1891 (pp.
465516), Munchen, Germany: Verlag der Akademie.
Titchener, E. B. (1909). Lectures on the experimental psychology of the thoughtprocesses. New York, NY: The MacMillan Company.
Watson, R. I. (1963). The great psychologists: From Aristotle to Freud. Philadelphia, PA:
J. B. Lippincott.
Woodworth, R. S. (1938). Experimental psychology. London, England: Methuen and Co.

Received February 2, 2009


Revision received July 23, 2009
Accepted July 25, 2009 y

S-ar putea să vă placă și