Sie sind auf Seite 1von 20

The Subject Matter of Psychology

Gustav Bergmann

Philosophy of Science, Vol. 7, No. 4. (Oct., 1940), pp. 415-433.

Stable URL:
http://links.jstor.org/sici?sici=0031-8248%28194010%297%3A4%3C415%3ATSMOP%3E2.0.CO%3B2-4

Philosophy of Science is currently published by The University of Chicago Press.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained
prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in
the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/journals/ucpress.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For
more information regarding JSTOR, please contact support@jstor.org.

http://www.jstor.org
Fri May 18 08:19:09 2007
T h e Subject Matter of

Psychology

GUSTAV BERGMANN
$1. INTRODUCTION

HE coordination of both logical syntax and


methodological analysis of scientific language by
scientific empiricism (logical positivism) has
achieved a twofold result:
I . Elimination of philosophical pseudo-pro-
blems and purification of the empirical sciences
from the biases engendered in them by these problems.
2 . Creation of an instrument apt for exact formulation, loca-
tion, and analysis of the real meaning of the valid philo-
sophical issues.
While in the earlier phases of the logico-positivistic movement the
first of these two major achievements has been particularly
hailed by its students and likewise resented by its opponents,
emphasis now has shifted towards the second. Language analy-
sis shows that
2a. one class of the questions mentioned in (2) can not be
formulated except by reference to a chosen language,
that consequently the answers to be given are contin-
gent upon that choice, and in this sense are of a prag-
matic nature (formation and transformation rules,
primitive predicates, criteria of meaning); while
2b. a second class points to problems of the empirical
sciences.
4I 6 Subject Matter of Psychology
Thus scientific empiricism is in a position to redefine the meaning-
ful content of many of the old issues on the linguistic level and to
help in formulating just those problems scientists sometimes sur-
mised to be forbidden by a too narrow epistemological rigorism.
Having reached this mature state, scientific empiricism now is also
psychologically much better equipped to overcome resentment.
As far as psychology is concerned, it is here submitted that
this recon~tructivetask has already been performed implicitly by
Carnap in his paper on "Testability and Meaning".' The
following discussion should not be considered as anything else
but a tentative interpretation of his work with a special view
towards the problems of psychology. I realize, however, that
this attempt to come as near as possible to actual psychology of
today had to be paid for by the serious disadvantage of very
aphoristic and groping formulations, for whose rashness I shall
have to assume the exclusive responsibility. Starting on such a
discussion it might not be inappropriate to ask, whether method-
ology has not lately become inclined to exaggerate and over-
dramatize that resentment against scientific empiricism, to fight
valiantly against strawmen and windmills. I do not feel that
this is the case. The positivistic-pragmatic trend of more
recent scientific thought, especially in this country, has, to be
sure, succeeded in quelling many remonstrations, though there
are still violent outbreaks openly raising philosophical issues
such as Koehler's last and otherwise so impressive book.2 But
there is likely to be a good deal more of repressed and under-
standable uneasiness which needs analysis before it degenerates
into resignation.
As a matter of fact, if there has ever been any reason for that
fundamental astonishment which is said to be at the basis of all
philosophy, a bird's eye view of modern psychology can not
fail to arouse it. Psychophysics and molar behaviorism, both
philosophically sophisticated, and psychoanalysis as a thrivingly
unsophisticated successor to introspectionism, all claim to deal
somehow with the same 'subject matter'. Consequently and
Philosophy of Science, 3, (1936),418-471; 4, (1937),1-40.
Koehler, W., The Place of Values in a World of Facts (New York: Liveright Pub-
lishing Corporation, 1938).
G . Bergmann 4x7
most fortunately, actual research makes every effort for coor-
dination and translation and all kinds of isomorphisms are sought
also by those who dislike the term. But is it not utterly un-
pleasant to find oneself confined to the laboratory and to patch-
wise theoretical guesswork even in such a fundamental matter,
to be left entirely without the reassuring help of an integrating
philosophy in the very search for the common subject of all these
psychologies? Let us illustrate one aspect of that vexing situa-
tion in plain and enjoyably metaphysical language: While be-
haviorism ascetically endeavors to 'construct' the phenomena of
the 'mind' on a second level of reality over the allegedly solely
honest level of physical reality, Koehler's phenomenalism en-
joys all the forbidden pleasures of immediate mental reality and
so, apparently, does psychoanalysis. This illustration by the
central issue of philosophical realism, famous for the fallacies of
the material mode of speech, has not been selected a t random.
There seems to be no way out for theoretical psychology either,
as long as it does not free itself from the ambiguities of this mode.
I t is a case for language analysis. To anticipate the result of
such a discussion, drastically condensed and put into a nutshell:
traditional epistemology became disentangled only after this
c'
unclear mixture of logical and psychological components"
(Carnap) was separated into its parts by cautious distinction
between the material and the formal idiom. In the same way
psychological methodology should benefit from an elimination of
its pseudo-epistemological enclosures. But that, of course, can
be achieved only after the logical and epistemological criteria
for any science have been laid down, and relative to this frame
of reference. Then, the 'psychological' or, rather, methodological
problems of philosophy (logic and general methodology) and the
'philosophical' or, rather, methodological problems of psychology,
though pragmatically connected, will prove to be on different
systematic levels.
52. THE LANGUAGE O F SCIENCE
Let us begin by restating, without any detail, the basic prin-
ciples as formulated by Carnap. Syntactically, the full (indi-
vidual) sentences and the general statements (laws) of science
418 Subject Matter of Psychology
can be interpreted as constructions over a basic class of so-called
primitive predicates. In this structure the regulative principle
of empiricism determines the choice of both, the primitive predi-
cates and the syntactical rules. The primitive predicates must
be observable, 'observable' being a basic term and as such, of
course, not defined in the theory. The second and only other
descriptive term realisable necessary for the finer empiricist theory
of meaning, is not relevant for our much simpler purpose. The
syntactical rules, on the other hand, are selected in such a way
as to admit only sentences which are confimable or testable.
This however is performed by formal language analysis. Test-
ability or confirmability are purely linguistic criteria, character-
izing the relations of compound sentences to the observable
predicates.
These criteria being agreed upon, the second methodological
decision is the choice, by characterization rather than by enumera-
tion, of the observable primitive predicates. According to the
thesis of physicalism, any scientific language can be constructed
on the basis of the language used in physics, the physical language.
In selecting the primitive predicates of this language, there are
three possibilities, viz. :
a. the thing-language; example: 'this time-space point is red.'
b. the phenomenological language; examples: '(I am) now
hungry.' '(I am) now seeing a red spot.'
c. the psychological language; examples: 'Mr. A is now angry.'
'Mr. A is now seeing a red spot.'
While (c) entails certain factual inconveniences as to validity
and reliability, (b) implies moreover the clumsiness of so-called
methodological solipsism. But still the choice among those three
alternatives is by no means a problem of psychology, but a purely
methodological decision, analytically prior to psychology as well as
to any other science. Physicalism, as now formulated by Car-
nap, holds that (a) is a sufficient and suitable basis for any science:
The physical language is the universal language of science.
Its primitive predicates belong to the thing-language.

I t should be noted, that the term, physicalism, is used here in


its larger meaning. A second, more specific meaning has been
G . Bergmann 419
distinguished by Feigl.3 Furthermore, one important point must
be emphatically stressed here. The physical language is not the
language of theoretical physics. The latter is a formal system-
language. I t becomes confirmable, empirically meaningful, and
thus, in any of its everchanging forms "until further notice",
integrated into the universal language of science only after the
necessary coordinations between its terms and those of the thing-
language have been established. This distinction is essentially
the same as the more familiar one between pure (axiomatic) and
applied geometry, the latter being a part of the physical language.
That is one of the focal points upon which this discussion centers.
After all these preparatory steps have been made, and only
then, one can proceed to the problems of special methodology,
i.e., to an analysis of the 'languages' of the various disciplines.
After what has just been said, the main task of such an investiga-
tion is to clarify the position and the peculiarities of the language
in question within the universal language of science, whether and
in what sense it can be said to be a distinguishable part of it, and
so forth. This purely linguistic investigation is nothing but what,
in the material idiom, is usually called the determination of the
subject matter of a discipline. The underlying facts to be kept
in mind are:
I . In a rather vague sense it can be said that those sentences
of any single discipline which belong to the thing-language
are constructed from a certain descriptively characterized
subclass of predicates. This is what is meant by the subject
matter of a discipline. But
2. I t is not a t all necessary and not even likely that these predi-
cates are a subclass or constructed from a subclass of the
observable primitive predicates basic to the universal sci-
entific language, or that they have some formal linguistic
properties in common; while
3. The hypothetico-deductive system-languages used in the
various disciplines might very well be entirely independent
system-la nguages (theories) without any common term
or law.
a Feigl, H., "Unity of Science and Unitary Science," Journal of Unified Science, I,
(1939).
4 2o Subject Matter of Psychology
The so-called thesis of the unity of science stresses (2) without
denying (3).
A few clarifying remarks on this rather important point may be
added. Take the sentence 'Mr. A operates an elevator.' Within the
frame of linguistic reconstruction, such a sentence gets its meaning only
by means of a number of lengthy chains starting from primitive predi-
cates 'dealing with' colors, sounds, spatial structures, and so forth.
Moreover, one might rather safely challenge everybody who holds that
there is any class of primitive predicates which is not represented among
the linguistic ancestors of the material of any science. But it is like-
wise true that our sentence 'Mr. A operates an elevator' and the terms
occurring in it belong to the linguistic material of several disciplines,
as psychology, sociology, economics, and technology (applied physics).
And quite obviously the 'derivation' of the thing-language sentence is
not affected by the scientist's choice of the context in which he is going
to put it according to the 'subject matter' just investigated.
The term 'system-language' is equivalent to the term 'theory'. There
is no better way to explain its meaning than by referring to the hy-
pothetico-deductive system-languages of theoretical physics and axio-
matic geometry. These problems have been most exhaustively treated
and there is no need and no space here to go into them with more detail.
I t might be mentioned, however, that the homophony usually obtaining
between many terms of the thing-language and those of the less ad-
vanced theories has always been a very embarrassing source of confusion.
As a rule it is responsible for theoretical short-cuts, but sometimes it is
also a t the basis of methodologically unjustified resistances. Psy-
chologists will here remember the resistance against the 'anthropo-
morphic' terms of purposive behaviorism (see $3 B).

$3. THE LANGUAGES O F PSYCHOLOGY


Applying these principles to an examination of the actual con-
tent of psychology, one will have to distinguish three different
sublanguages of psychology in the sense of $2, I and accordingly
three different system languages in the sense of $2, 3. If one
prefers to put the same thing more paradoxically, three different
subject matters of psychology will emerge. But all these lan-
guages with their respective theories are, of course, comprehended
in the universal language of science. The thesis of so-called
logical behaviorism (Hempel) is nothing but the statement of this
G. Bergmann 421
thesis of scientific empiricism for psychology: Any possible psy-
chology has to be formulated according to the rules and in the
terms of the universal language of science.
A. I'he M-Language. The sentence 'Mr. said (wrote) yester-
day, "I am enjoying myself",' is a meaningful sentence con-
structed from the thing-language, and so are all sentences of the
form: 'X says ". . .".' Hence the class of all these sentences
(quotation sentences) or the subclasses of those sentences regis-
tering all the utterances of the individual 'Mr. A' or of the mem-
bers of a certain group, e.g., the French or the British, are a legiti-
mate basis of a scientific theory. In the material idiom, the
study of the linguistic utterances (of an individual) is an empirical
science. Of course there are several points of view from which
such a study may be undertaken. Some of them concentrate
upon what is contained between the double quotation marks.
Thinking for instance of the class of 'sentences' uttered by the
French or the British, one sees that the history and grammar of
the empirically given 'languages' are possible lines of such a
research. But there is another view point, more important for
our subject, from which such an investigation may start; namely
the study of the 'language' of an individual and its 'meaning'.
Here two remarks must be made:
a. The psychological terms 'sentence', 'language', and 'mean-
ing' have been put in quotation marks in contradistinc-
tion to the homophonic terms used in logical syntax and
methodology.
b. All our scientific knowledge about this 'language' and its
'meaning' is in principle arrived at in a purely empirical
(behavioristic) way and without any resort to introspection
or any kind of mentalism.
Because human life, from a very early stage to its end, is
accompanied by an everflowing stream of actual and easily
aroused verbalization, and because of the unequaled importance
of just this human activity, the intrinsic consistency of this
gigantic protocol could not fail to provoke an effort a t intrinsic
explanation. Thus it stimulated one of the earliest scientific
achievements: the concept and the theory of human mind. As a
42 2 Subject Matter of Psychology
matter of fact, the study of Mr. A's 'language' is most powerfully
aided by the invention of a theoretical system-language, the
M-language, containing terms as 'Mr. A's thoughts', 'conscious',
'sad', 'to wish', and so forth. Of course, the term 'Mr. A' in such
a theory does not designate the organism of Mr. A or any part
of it, but is a term coordinated to a construct over sentences as
'Mr. A says "I .. .".' Whether it will be a basic term or not
depends upon the syntax of the special system. At any rate, we
are dealing here with a very rough illustration. I t is clear, how-
ever, that the term 'Mr. A' designates the organism of Mr. A only
if it occurs outside of the double quotation marks. In the same
way, and in contradistinction to "I am enjoying myself" which
for itself is no sentence of the thing-language at all, but part of a
sentence of this language, the proposition 'Mr. A is enjoying him-
self' is a sentence of the theoretical M-language. Whether this
sentence will be asserted or denied, depends on our knowledge of
Mr. A's vocabulary as well as upon our theory. To be sure, it
will be often asserted without a homophonic utterance of the
subject in question.
Here an important objection will probably be raised. Mr. A's
high spirited state of mind might as well be inferred from his
humming his favorite tune, from a mirthful expression of his face
or many other entirely unverbalized behavioral observations. As
far as the tune and certain 'physiognomic' expressions are con-
cerned, this remark touches upon the question, how to define the
psychological term 'language'. That is obviously a problem of
psychology. Furthermore, it is not held here that there are no
other than linguistic cues for the construction of an empirically
meaningful M-language, or that the exclusive use of quotation
sentences as the basis of a theoretical system-language is in any
sense commendable. I t is not claimed either that the existing
M-languages (every day language and the mentalistic, introspec-
tionistic, and 'depth-psychology' theories) are logically elaborate,
theoretical structures, or that their respective coordinations
always show desirable precision and unambiguity. The facts to
be stressed are rather, that (a) the inconveniences just mentioned
are all of a merely pragmatic character, (b) M-languages as such
G. Bergmann 423
are perfectly legitimate system-languages (theoretical approaches
to psychology), and (c) these languages are patterned after and,
a t least predominantly, based upon the quotation sentences of the
universal scientific language (the phenomena of human
'language').
Let us add a few explanations and consequences of this analysis:
I. Apart from all questions of reliability, M-languages have the
same methodological position as e.g., the system-language of
theoretical physics. A translation of this result into the
material idiom, though dangerous and misleading, might be
of some elucidative value a t the present state of affairs. As
far as I can see, the translation suggesting itself most readily
would be: Mental phenomena, thoughts, wishes, percep-
tions, emotions and so forth, are as real as velocities, masses,
forces, electrons, and atoms, in physics. Still more viciously
and confusingly, if that is possible: Psychic events are as
directly observable as those of physics. There is no use in
elaborating here again how objectionable such formulations
are. Instead it seems appropriate to stress that behaviorism
as a method of psychological research must not be inter-
preted as a pseudo-philosophical denial of the underlying
methodological meaning of these formulations. Logical
behaviorism does not imply any philosophical aspersion
against any line of psychological research just because it has
to depend upon a study of the phenomena of 'language'.
Some of the most important and relevant problems and
aspects of human psychology do undoubtedly largely depend
upon this 'material'.4 Both logical behaviorism and psychol-
ogy may freely admit that anthropomorphism and even
empathy need not always be methodological sins, detours
reluctantly conceded because of our human weakness.
Sometimes they may be methodologically sound and con-
sistent preparatory steps directly towards the construction
' A similarly dangerous and misleading hyperphysicalization of psychology, the
confusion between "logical behaviorization" and universal "experimentability" has been
discussed in a previous paper, "On some methodological problems of psychology," Phi-
losophy of Science, 7 (1g40),205-219.
Subject Matter of Psychology
of a genuinely scientific and empirically meaningful M-lan-
guage. After an educative period of ascetic behaviorism,
psychology of today seems resolutely bent upon the attain-
ment of this reconstructive insight.6
Ia. I t should be clear by now that this introduction of
M-languages without the detour of either behavioristic
theories (B-languages) mainly based on overt behavior,
or physiological psychology (P-language) does in no way
revive the mentalistic prejudices of introspectionism
with its qualities and its dangerous "philosoph~cal"
implications. At a later state of this argumentation
($3 B 2 ) it will become still more obvious that M-lan-
guages are not conceived here as opposed to B-languages,
but rather singled out as a special type of B-language,
distinguished by the fact that they are predominantly
based upon and patterned after a special class of sen-
tences of the thing-language. For the elimination of
certain biases, however, which after all are but emotional
residues from the heroic period of behaviorism and logi-
cal positivism, it might not be superfluous to emphasize
the thesis of the preceding remark. Only together with
these biases will eventually disappear what is still left
of resentment against scientific empiricism.
2. T hose who appreciate synoptic generalities might gather
some intuitive confirmation for this linguistic rehabilitation
of 'mentalism' from the important r61e language plays in the
most recent scientific approach on the M-level, psycho-
analysis, an importance so overwhelming that to it is devoted
one of those rare passages in Freud's Introductory Lectures
(chapter I ) distinguished by a strong and eloquent emotional
tinge. But we shall prefer to mention a more technical
consequence of the preceding discussion for psychoanalysis:
the automatic elimination of one of its most annoying
pseudo-problems. I mean all the speculation centered

6 Allport, G. W., "The psychologist's frame of reference," Presidential address delivered


a t the meeting of the American Psychplogical Association, (1939).
G. Bergmann 425
around the terms: conscious, preconscious, unconsciou~.~I t
is not difficult to give a t least some prima facie evidence,
that these terms can be defined (or, rather, a t the present
state of affairs, characterized) formally and without any
pseudo-psychological reference. Which of these predicates
is attributed to a given 'psychic event' depends only upon
the way in w$ich the corresponding utterances have been
elicited. From this very formal standpoint, psychoana-
lytical technique itself is nothing but a way of eliciting such
'responses'. And according to these differences in the mode
of elicitation, the position of an utterance or a group of utter-
ances in the whole string of consecutive responses of the
subject, with regard to both order and 'content', is different.
Hence, as the terms, 'conscious', 'preconscious', 'uncon-
scious' are nothing but names for different types of such
structures, they are from the very beginning on a par,
and there is no methodological (philosophical) meaning in
the statement, that 'the unconscious' is merely a speculative
and doubtful extension of 'the conscious'. Of course there
is no scholastic objection against an abbreviated and well-
defined use of these nouns. Any ontological hypostasis,
however, is hopelessly metaphysical.
There are other extrapolating extensions of the psycho-
analytical M-language. Some of these have the character
of scientific hypotheses and are, therefore, methodologically
unobjectionable. At least one other, so-called bioanalysis,
an alleged application to biology, is merely speculative and
hardly a scientific approach to anything a t all. This has
been very clearly pointed out by psychoanalytical method-
ology itself.? On the other hand, from the purely formal
and intrinsic viewpoint here adopted, what is usually called

Specific problems arising from the fact that psychoanalytic theory is an M-language
must be distinguished from the general requirement of 'logical behaviorization' of psycho-
analysis. As to the latter, see the article quoted, footnote 4, 197-236.
Bernfeld, S., "Zur Revision der Bioanalyse," Imago, 23, 1937. Some psychoana-
lytical approaches to the social sciences raise similar problems. See Bergmann, G., "Zur
analytischen Theorie literarischer Wertmasstaebe," Imago, 21, 1935,498-506.
426 Subject Matter of Psychology
Freud's biological speculation, i.e., the hypothesis that
phylogenetically inherited patterns are equivalent to indi-
vidual experiences, is nothing but a set of syntactical rules
(in Carnap's terminology, P-rules). But there is, to be sure,
still a different methodological angle ($4) doing more justice
to the factual importance of this "biological" aspect of psy-
choanalytical M-language.
3. As it has been shown before ( $ 2 , 3), one of the main tasks
of a scientific, i.e., non-ontological study of the subject
matter of any science is to investigate the relations of its
system-language to those of other disciplines. In this
respect, M-languages show a peculiarity which is a t the root
of one of the most famous philosophical fallacies, the mind-
body problem.* M-languages contain only one coordinate
which allows for a formal and pragmatically successful
identification with the time axis out of the four-dimensional
spatio-temporal frame of reference used in theoretical
physics. Even apart from the snares of the material idiom,
the statement that the subject matter of psychology is non-
spatial is, therefore, somewhat incautious. M-languages do
not contain these dimensions, and the criticized formulation
is not at all so much comparable to a proposition like 'the
surface of a sphere has no third dimension', but rather of
the type 'gravitational fields have no citizenship rights'
(legal orders as system languages). Though, to avoid mis-
interpretation, it should be mentioned in this context that
the above criticism does not apply to the use of the mathe-
matical term 'space' and other geometric terms to represent
and manipulate structural properties of any system lan-
guage. That holds true also for 'dynamic' analogies.
Whether such an apparatus is applicable and whether its
use is economically justified is entirely puaestiofacti. Lewin
has always been careful to make this distinction in explain-
s See Feigl, H., "Logical analysis of the psychophysical problem," Philosophy of
Science, I, 1934, pp. 420-445; an integration and elaboration of the "contribution of the
new positivism" to this issue, anticipating most of the results of consistent language
analysis.
G. Bergmann
ing his 'mathematical method'. As far as psychoanalysis
is concerned, it is perfectly clear, that its 'mechanical'
models and occasional 'geometric' pictures are nothing but
analogies and visualizations.
Before turning to the two other psychological system languages
it shall be repeated that the preceding paragraphs do not pretend
to be more than a short and hardly adequate attempt to call
attention to one aspect which seems to suffer from neglect. That
is certainly not the case with the other two languages. Therefore,
as the purpose of this discussion does not allow one to omit them,
it will a t least limit itself to a few remarks. Many adequate
formulations have been presented recently.
B. The B-Language. Constructed upon the basis of primi-
tive predicates of the thing-language, there is a large class of
behavior-sentences, which of course also belong to the thing-
language. They speak of human and subhuman activities like
'running', 'VTE-ing', 'lever-pressing', 'eating', and 'speaking'.
They deal with what psychologists call 'behavior' and what they
try to define as "that part of the functioning of the organism
which is engaged in acting upon or having commerce with the
outside world" and regarding which "it is often desirable to deal
with an effect rather than with the movement itself, as in the
case of the production of sounds9' (Skinner). According to $2, I ,
there is no merit in efforts to circumscribe the class of these predi-
cates and relations syntactically. I t must be stressed, however,
that all these terms and sentences, so far, belong to the thing-
language and not to any system-language. I t is especially impor-
tant to keep this in mind, because in contrast to the procedure in
physics, most of the work on this 'subject matter' has been spent
in the sharpening of the thing-language. That is what is usually
called the well-developed operational side of behaviorism. In this
respect, standards of breathtaking rigorosity have been pro-
claimed only recently by workers in this field.9 Nevertheless,
theoretical systems, B-languages to fit this refined thing-language,
have been of the utmost scarcity. If some theorizing occurred, it
Skinner, B. F., "The behavior of organisms," (New York: D. Appleton-Century
Company, 1938), Chapters I, XIII.
4 28 Subject Matter of Psychology
was fortunate usually to escape discovery, because the homoph-
ony of its terms with those of the elaborate new thing-language
was mistaken for identity. I t is not so long ago that the labeling
of some structures and persistent sequences of such terms in the
experimental or observational protocols with new terms intro-
duced by this very procedure, as 'purpose', 'demand', 'goal',
'learning', and the cautious and sober statement of a minimum
of laws obtaining between these terms of the next linguistic levels,
required the intrepidity of scientific pioneers. Today Tolman,
Hull, Lewin, to mention only three representative theoretical
programs most widely discussed among psychologists, engage
openly in the construction of what I believe to be B-languages.
But it might still be worth while calling attention to the fact that
by this very procedure, the behavioral predicates from which
such constructions start become the basic terms of a theory, how-
ever tentative and rudimentary such a construction might be.
Example. Basic terms: being starved, running, entering blind alleys,
VTE-ing, eating. Introduced terms of the next levels: hunger, learning.
Basic law$ (postulated P-rules generalized from the thing-language
data): increasingly steep negative slope of the error curve with increas-
ing hunger; the configurational connection between the VTEcurve
and the error curve (Tolman's hypothesis).
VTE is an abbreviation for 'vicarious trial and error'; VTE-ing a
very characteristic behavior of maze-learning rats. The number of
VTE's displayed increases steeply and shows a maximum shortly before
the error curve (number of entries into blind alleys of the maze) shows
a marked and persistent decline.
T h e following two remarks go back to the relation between M-lan-
guages and B-languages ($3 A ~ a ) .
I . Quite obviously linguistic behavior is also behavior and thus
part of the material of behavioristic theories. Possibly no
somewhat satisfactory M-language can be constructed with-
out some unverbalized behavior-material, and certainly no
meaningfullo behavioristic approach to human behavior can
lo Of course the expression 'meaningful' must not be understood here as a methodologi-
cal term. Also in some other respects the diction has been loosened in this and later
paragraphs. Where no danger of confusion exists, the material idiom has been used for
convenience.
G. Bergmann 429
disregard the linguistic material. It is clear, moreover, and
has been pointed out (93 A b) that linguistic behavior gets
its psychological meaning, scientifically as well as geneti-
cally, only within the broader texture of human behavior of
all types. The point is that after what can be gleaned from
the present state of affairs, M- and B- languages seem to be
of very markedly different intrinsic structures. T o use a
suggestive analogy for illustration of this difference: A sen-
tence in a play may have an important place in the develop-
ment of the plot, and a t the same time serve as a cue for
some actor.
2. T o explain what is meant by the vague expression 'different
intrinsic structure', let us, for instance, consider psycho-
analysis. Assume that in an elaborate psychoanalytic
M-language, terms as 'libido', 'Id', 'cathexis', will be the
basic ones, while the postulated laws obtaining between
them and the introduced terms as, e.g., 'anxiety' 'uncon-
scious wish', 'complex', will be shaped after mechanisms like
'sublimation', 'repression', 'displacement'. In a B-lan-
guage, on the other hand, these terms and laws will probably
only be found in the way of isomorphism (94) and on a very
high level of introduced terms and derived laws. At the
beginning of the chains introducing these highly com-
pounded terms one will probably find terms corresponding
to certain very elementary behavioral units, both verbalized
and unverbalized, but probably predominantly or exclu-
sively unverbalized.11 And with the same degree of prob-
ability the laws postulated between these basic terms and
their more immediate derivatives (e.g., laws of the Condi-
tioned Reflex type) are unlikely to show any structural simi-
larity with the 'introspectionistic', M-shaped laws of psycho-
analysis.
C. The P-language. I n contradistinction to M- and B-lan-
guages, the P-language is not especially devised for psychology.
11 There is no contradiction between this expectation and the previous statement that

no meaningful behavioristic approach can disregard the linguistic material. In such a


theoretical B-structure all linguistic behavior may be located on the level of introduced
terms and derived laws.
430 Subject Matter of Psychology
Rather, the P-language is the system-language of biology (physi-
ology), replaced by and reduced to the system-language of
physics, wherever this is possible a t the present state of these
disciplines. The problem whether biology can eventually be
dispensed with by this approach, in the same way as chemistry
has been recently reduced to physics, is of no concern for the
purpose of this discussion.l2 The subject matter of this approach
is the material discovered by neuro-physiological research, so-
called psychophysiology, or any other experimental investigation
on this physicalistic level. The methodological difference be-
tween this type of experimental work and the theoretical ap-
proach of so-called Stimulus Response theory has been suffi-
ciently clarified by such psychologists as Hull, Hilgard, and
Skinner. But it may help to clarify the line of thought followed
in this paper if it is recalled that Hull's system is essentially a
B-language, though consciously and consistently built with a view
towards P-language, i.e., towards the factual isomorphism to be
established ($4). Consequently the terms used are homophonic
and the laws obtaining between them are formally isomorphic to
those discovered by the experimenters of the Pavlovian school.
There is no need to insist here on the enormous merit such an
orientation has from any theoretical standpoint. But still the
whole argument moves on the B-level.
Finally, as Hull's and Lewin's theoretical thoughts are often
mentioned together, I would like to stress here a rather important
difference. As far as I can see, Lewin's theoretical attempts are
also directed towards the construction of a B-language, but the
formal isomorphism after which this language is to be moulded
refers to the M- rather than to the P-level. "Regions" are
arranged beside each other and change their relations in combina-
torial patterns similar to those of our interacting thoughts and
adequately integrated perceptions, "barriers" grow up and vanish,
connection^'^ appear and disappear, "paths" are singled out by an
interplay of motivational "forces" and perceptually given struc-
IZThe thesis of strict physicalism is the affirmative answer to this empirically open
question. It should also be noted that the expressions 'replaced by' and 'reduced to'
as used here must be understood in the sense of factual isomorphism (8 4).
G. Bergmann 431
tures, according to changes in the M-universe of thoughts,
memories, expectations, perceptions, wishes, and fears. But here
again all the efforts towards a theoretical systematization move
unmistakably on the B-level, though this 'introspective orienta-
tion' might account for much of the emotive appeal and the
empirical fruitfulness of this approach.
$4. THE RELATIONS BETWEEN THE LANGUAGES

The distinction just made between formal and factual iso-


morphism needs explanation. Of course the term can only have
formal meaning in mathematics and logic. If a one-to-one or
multivalued correspondence between the elements and a one-to-
one correspondence between the operations (defined exclusively
and independently within each of these manifolds) can be estab-
lished in such a way that it is not disturbed by these operations,
i.e., if the results of corresponding operations applied to corre-
sponding elements are again corresponding elements, formal iso-
morphism obtains. E.g., the positive and the negative integers
(+n) and (-n) are isomorphic with respect to the operation '+'.
Isomorphic correspondence might also be established between
only part of the operations in the two fields or between parts of
the fields or between one field "P" and part of another field "W".
This latter case shall be symbolized here by: W -t P. Linguistic
investigation can not be pushed further and there is no formal
meaning in the question of whether or not any kind of identity
has been confirmed by the discovery or, rather, construction of
such an isomorphism between the terms and laws of two system-
languages. As introduced into theoretical psychology and meth-
odology by Koehler, however, the term has a different meaning.
I t implies the question whether such a tentative formal 'identifica-
tion' between two theoretical systems proves pragmatically suc-
cessful and allows for a detailed experimental substantiation in
full accordance with the standards of our empirical knowledge
and techniques. In this case one better speaks of 'factual iso-
r n ~ r p h i s m , 'or,
~ ~ rather, intertranslatability. As pointed out by
I8 This necessary distinction must not be interpreted as a criticism of Koehler's thought.
I t should rather help to appreciate this powerful, anticipatory vision as what it is, the
432 Subject Matter of Psychology
Boring,l4 the term isomorphism itself is misleading and danger-
ously conducive to anticipatory simplification.
At the present state of affairs, and in spite of the heuristic and
directive value of the model-isomorphisms proposed by or at least
implied in some recent approaches to psychology, almost every-
thing has yet to be done for the discovery of the factual isomor-
phisms which will eventually allow for a unification of the three
languages of psychology. Hull, Koehler, and Lewin, though with
different accents and in different terminologies, all agree upon
that. But, to be sure, no philosophical problem, no need for an
ontological search of a unitary subject matter will ever arise.
On the language level the isomorphism can only be a formal one,
and its empirical discovery and validation is a matter of fact
and of facts.
If one tries to locate the result of actual "isomorphic" re-
search, it seems that, e.g., so-called brain physics is more di-
rectly bent upon establishing the P -+ M isomorphism, while
reflexological research is mainly concerned with the relation
P --+ B. But one might wonder whether such a distinction has
more than an illustrative value even a t the present state of our
empirical knowledge. At any rate, it is more than likely that
with the progress of science the importance of the distinction
between B and M will fade and eventually disappear. Yerstehens-
psychologische explanations, on the other hand, and the 'inter-
pretation' of human behavior as symbolic or substitutional activi-
ties by psychoanalysis, show the direction M + B. But it has
never been maintained by any serious scientist, that a consistent
isomorphism of this type exists. Such a claim, however, would
constitute a possible meaning, absurd certainly, but still an
empirical meaning, of spiritualistic systems as, for instance, that
of Hegel. Taking into consideration what has been said before

legitimate heir of the great German tradition of natural philosophy, brought to the stand-
ards of present time empirical science. As to its linguistic and physicalistic interpretation,
see also a previous paper, "On physicalistic models of non-physical terms." Philosophy
of Science, 7 (19~01,151-158.
l4 Boring, E. G., Psychophysiological Systems and Isomorphic Relations. Psycholog-
ical Review, 43 (1~36), 565-587.
G. Bergmann 433
($3 B I, 2), there is much more hope for a factual isomorphism in
the direction B -t M. Finally, if one considers the pattern thus
anticipated,
P+B, P-M, B-tM
one sees: No arrow points towards "P", no arrow issues from
"M", and if one arranges the three levels in the order "P", "B",
"M", the arrows point all to the right. That reflects the tend-
ency of our scientific thought usually called naturalism.
The State CTniversityof Iowa.

Das könnte Ihnen auch gefallen