application of this principle. _More_ order is experienced than
that of space and time. Certain experiences “belong together” in a
concrete manner, whereas others do not, or, at least, belong together
less intimately. And again this “belonging together” may be itself
experienced. The very moment I form this sentence of my manuscript,
a shrill, disagreeable voice begins to sing in the neighbor’s house.
The context of my sentence is something which, though extended in
time, is experienced as belonging together, whereas those sharp
notes are experienced as something foreign, as not belonging to that
context, though they are experienced _at the same time_. In this case
our principle applied formulates the proposition that _to a context,
experienced as “one thing” belonging together, there corresponds a
dynamical unit or whole in the underlying physiological processes_. In
this respect again, the order of experience is a true representation
of a corresponding functional order in the processes upon which it
depends. Perhaps it is this last application of the principle which has
the greatest importance for gestalt psychology. In this form, at least,
it becomes a radical physiological hypothesis about common sensory
experiences as well as about the more subtle and complicated processes
involved in observable behavior.
In the last case I have taken an example from outside the realm of
_objective_ experience in the more restricted meaning of this term. A
sentence I may be forming is not a part of objective experience, as
is the chair before me. Still my statement about that experience is
no less simple and obvious than were earlier ones about spatial or
temporal orders. I admit that this is not always so, however. I do
not wish to recommend the observation of _subjective_ experience for
our purposes without some limitation; as yet it looks as though only
very simple statements about one’s own _subjective_ experience can be
made with a high degree of certainty. Moreover, we have enough cases
of objective experience to provide an adequate basis for developing a
working hypothesis.
In the last paragraphs my own direct experience was taken as a basis
for constructing a general working hypothesis about the relation
between experience and unobservable constituents of behavior. The
only way in which I can bring my observations in that field of direct
experience before the scientific public is through spoken or written
language, which, as I understand it, refers to my direct experience.
However, in terms of the position already taken in this chapter, all
language as a series of physiological events is the peripheral outcome
of other purely physiological processes in my organism, particularly of
those upon which my direct experience depends. My general hypothesis
states that the concrete order of actual experience is a true
representation of the dynamic order of corresponding physiological
processes. Therefore, _if, to me, my language is an adequate “symbol”
for my own direct experience, it is an objective symbol for those
physiological processes at the same time_. We may also conclude, then,
that it does not matter very much whether I take it as a symbol for one
or the other, since in the respects in which they correspond, there is
no difference between them.
If I now go back to behavioristic psychology I have to deal with
language as one form of behavior observed in my subjects. There is no
reason now why I should not take language as a symbol, since only the
most superficial view would treat it merely as a phonetical event. The
behaviorist himself, when listening to a scientific argument, will
find himself reacting not to the phonetical properties of speech, but
to its symbolic meaning: he will react in the same way practically
to the nouns “experiment” and “_Versuch_,” to “animal” and “_Tier_,”
though in each case the first and second words are utterly different
phonetically. Why should we change that attitude when the other person
is called a “subject,” and talks during an experiment?
The language of my subject may be taken either as indicative of direct
experience in my subject or of those processes which underlie his
direct experience. If the subject says, “This book is bigger than
that,” his words may be interpreted as stating a definite direct
“comparison-experience” which he is having or the corresponding
dynamic relationship in his physiological process. Since, for my
hypothesis the same “order” is meant in both cases, the alternative
is not important. From the viewpoint of behavioristic psychology the
physiological interpretation is necessary, but the other meaning is not
excluded. This second interpretation is kept distinct from the causal
explanation of observable behavior, but if it is suggested by the words
of the subject, it can do no harm. The behavior of a chick can tell
me without words that he is able to react to one brightness in its
_relation_ to another.[8] If in the course of an experiment, a human
subject tells me that he sees one object as brighter than the other,
the scientific value of this sentence is exactly the same as that of
the chick’s behavior, whether I assume direct experience in the human
subject or not. Why then should I decline to accept language as one of
the nicest and most instructive forms of behavior? Of course, we can
obtain the same results in man as in the chick, without language, by
applying the somewhat clumsy, statistical methods of animal psychology,
and some behaviorists seem to prefer such a procedure. The only reason
I can see for this attitude is historical. Introspectionists have used
the “verbal report” in “introspection”; and I am ready to admit that
their method was something like a blind alley. But, as is usually
the case in such a situation, behaviorists have become negatively
conditioned to much more than the “harmful stimulus” alone, so that
they shy away not only from introspection but also from all things once
connected with it, and primarily from language, though it is in itself
utterly innocent.
BIBLIOGRAPHY
W. Köhler: _Die Methoden der psychologischen Forschung beim Affen_.
Abderhaldens Handbuch der biologischen Arbeitsmethoden VI, D. 1921.
III
_The Viewpoint of Introspection_
Round about the accredited and orderly facts of every science there
ever floats a sort of dust-cloud of exceptional observations.
W. James, _The Will to Believe_.
WILLIAM JAMES has well described how a vivid interest in the “irregular
phenomena” of the “dust-cloud” will often mark the beginning of a new
productive era in a science. At such times, what has been exceptional,
and a part of the dust-cloud, may suddenly become the very center
of scientific work. We shall see presently that introspection has
developed a procedure by which an orderly though artificial system of
psychology is _protected_ against a similar revolution, a procedure in
which the most interesting observations are continually exiled into the
dust-cloud.
In referring to introspection I am not considering any special school,
but all those psychologists who treat direct experience in the manner
now to be discussed.
For the most part introspectionists will agree with what has been
said about behaviorism in the first chapters; some, indeed, will have
recognized their own arguments in parts of my criticism. Wherein
lies the difference then between introspectionism and my position?
This difference, I promise, will become startlingly clear as soon as
we discuss the observation of direct experience. Since _objective_
experience is so much more legitimated by its use in physics and
behavioristic psychology, we shall first examine the treatment of it by
introspectionists, with the result that, for all practical or concrete
purposes, introspectionism will be found to be astonishingly similar to
behaviorism.
The very moment we begin to move naïvely in the field of direct
experience, protests will arise on all sides. If I say that I see a
book before me on my desk, I shall be criticized because nobody can
_see_ a “book.” Lifting the book between my fingers, I am inclined to
say that I feel the weight, external to my fingers and roughly in that
place where the book is also seen. This, a critic would remark, is an
excellent example of a typical statement by an untrained observer,
quite satisfactory, it is true, for the practical purposes of common
life, but altogether different from the kind of description given
by a trained psychologist. Even the character of being an “object,”
or “thing,” which I have tacitly attributed to the experiences I
have called “book” and “desk,” is improper in correct psychological
description. If observation is to give us simple and genuine data as
the elements of psychology, we must learn to make the all-important
distinction between _sensation_ and _perception_, between the bare
sensory material actually given to us and the host of other items
which since early childhood have become associated with it. You cannot
see a “book,” I am told, since this term involves some knowledge
about a class of objects to which this specimen belongs, and about
their use, etc., whereas in pure seeing such knowledge cannot enter.
As psychologists our task is to separate all these “meanings” from
the seen material _as such_, the manifold of simple sensations. It
may be a difficult task to effect this separation, and to behold the
net sensations which are the actual data; but the ability to do so
is precisely the special talent which transforms the layman into a
psychologist. It ought to be evident that originally, when we lift
a book, we do not have the experience of a weight external to our
fingers in space. At first there are mere sensations of touch, and
perhaps strain, in our fingers, whereas that “weight outside” has
been connected with them in a long history and through the influence
of other factors than those of pure sensation in our hand. A similar
consideration will show us easily that genuine sensory data cannot
present “objects.” “Objects” cannot exist for us before sensory
experience has become imbued with meaning. Who can deny that meaning
fundamentally determines almost all experience? Does it not finally
lead to a kind of illusion? The German noun “_Igel_” sounds to Germans
as though no other animal but a hedgehog could have this name.
“Eagle” however, which in English is acoustically the same as “_Igel_”
is in German, sounds to an American as though only an “_Adler_”[9]
could be called by this name. Here it is obvious that we have to
discriminate between a genuine sensory experience, which is the same
in both languages, and two different meanings connected with it in
different countries. Again, the symbol + _looks_ like the sign for
addition, especially if you see it between numbers; but it might as
well have been chosen as a symbol for division. If we feel at first a
disinclination to accept the last statement, the only reason for it
consists in just that sort of complexion which meaning, since we were
children in school, has given to the innocent symbol. As soon as we
realize how strong this influence is, we have to admit that probably
_nothing_ in the naïve experience of a normal adult can be free from
it. Even the most lively characteristic of an experience we may try to
describe, may be a property of that experience only in this acquired or
accidental way.
Meaning depends upon personal biography; it has a highly complicated
origin and represents a somewhat accidental trait of our material.
Therefore we must get rid of it and learn to approach actual sensations
in such a way that their qualities and laws may be discovered in their
pure form. This procedure is called “introspection.”
Probably all psychologists who are teaching at the present time have
learned this lesson thoroughly, though many perhaps implicitly and not
as a well-formulated program. Its basis seems justified; its content
seems sound scientifically. Therefore, if introspection is possible and
can lead us to actual sense data, we should be inclined to follow in
its footsteps. Direct experience as such, it is true, will not be worth
so very much under these circumstances. Of all _objective_ experience,
at most only the part cleansed and selected by introspection will
become the subject-matter of our study.
However, we should not form a definite judgment before knowing more
about the criteria for the selection of some experiences as the genuine
ones and the exclusion of those other parts of experience which are
assigned to the influence of meaning. Let us consider a few more
examples, then, which exhibit other properties than those to be found
in the cases already mentioned.
Do you see that man on the street approaching us at an even gait? Now
he is ten yards away, and presently five. What do you say about his
size at the two distances? That five yards away it was practically the
same as ten yards off? But that is impossible! A simple consideration
of geometrical optics will tell you that during the man’s approach
his height must have doubled and his breadth too, so that his total
size must have become four times the amount it was at ten yards. We
transfer the observation from the street into the laboratory. Here I
have two cardboard rectangles. The sides of the first have a length of
two and three inches, those of the second are six and nine. If now I
hold the first before you at a distance of one yard and the second at
three yards, you will admit that, optically, they must have the same
size approximately, since their linear dimensions vary exactly as do
their distances. But you tell me that the further rectangle appears
much larger to you than the nearer one? There you have it! Yours cannot
be a statement about actual sensory experience. A slight change of
conditions will make you realize that my opinion is right. Look through
the opening of this screen, which will show you the two rectangles
on a homogeneous background, but no more of their surroundings. Do
they look different now? Still a little. Then we shall be even more
cautious. I darken the room; then I turn the light on, allowing you to
make your observation through the screen, only for a small fraction
of a second, so that practically no eye-movements can occur during
that time. Now both rectangles have the same size, do they not? After
some practice you will be able, perhaps, to see them as equally
large even without the help of my screen and the short exposure, as
soon as you look upon them in the proper way--which is the method of
introspection for obtaining pure sensory experience. If you do not
accept my demonstration, you might even go so far as to believe that
the after-image of an object changes its size according to the distance
from which you “project” it upon a screen, because in untrained
observation it certainly seems to change considerably, according to
the distance of your fixation point. But that is absurd. How could it
really change, if the area of the retinal after-effect remain constant?
My second example may be regarded as a natural consequence of the
first. When dining with your friends, what forms do the plates have in
your experience, those to the left, to the right, and opposite you?
They are circular, you say, as is your own plate? Impossible again!
They must be elliptical, as you will admit after you have once thought
about their projection upon your retina; some of them must even be
very flat ellipses, and even your own plate will become an ellipse as
soon as your eye does not look down upon it vertically. It will be
worth while to apply to this case a procedure similar to the one we
used in the first example. On this plane which, as you see, is oblique
to the direction of your eye I fix a circle; on this other plane, at
right angles to the direction of your eye, I fix an ellipse, whose
shape is chosen in such a manner that from its place it is projected
upon your retina in the same form as the circle is from its oblique
plane. Though you describe the first as a circle and the second as
an ellipse, they are two almost identical ellipses in actual sensory
experience. Take this screen with two holes in it, which lets you see
both forms, but excludes, however, the data by which the angles of the
planes became known to you at first. Both forms look alike now? All
right. After a short while you will be able to see correctly without
the screen, providing you adopt the attitude of introspection. Once
more, the corresponding observation on an after-image of a given form,
“projected” upon planes of different oblique positions with respect to
your line of vision, will elucidate the case completely. Its form seems
to vary according to the positions of the plane, though the retinal
after-effect has the same form throughout. This being absurd, you
see the importance of observing genuine sensory experience by way of
introspection.
A similar paradox has been much discussed since Helmholtz wrote his
_Physiologische Optik_. Here on this desk near the window, and parallel
to it, a vertical screen is fastened. On the window side of the screen
I place a black paper on the desk, and in the shadow of the screen,
a white paper. These papers are selected in a special manner: the
dark one exposed to the direct illumination from the window reflects
the same amount of light as the white paper reflects under its low
intensity of illumination. You call the first one “black” and the
second “white”? Then again you have experiences which are not to be
trusted, because under these circumstances the images projected upon
your retina are both of them equally intense. So the sensations,
i.e., brightnesses, of both must also be the same. Moreover this can
be demonstrated. Take this piece of gray cardboard and look through
the two holes in it so that one of them is filled by a section of the
“black,” and the other by a section of the “white” paper. Now, as the
surroundings of the papers are excluded from vision, and thus you have
pure sensations before you, their approximate equality is obvious.
Again, with a little practice--most painters observe objects this
way--you will be able, without the cardboard, to destroy the illusion
of their difference--just by looking at the papers in the attitude of
introspection.
All these observations are of an illusory sort, then, the purification
of which exhibits the true sensory facts. In this and other respects
they are comparable to a great many well-known “optical illusions,”
the diagrams of which fill many pages in our textbooks. Take the
famous Müller-Lyer figures! When you observe them often and try to
pick the main line out of its surroundings, you will find the illusion
become unstable, and finally it may even disappear. It was not, then,
a true sensory fact. The same thing may be shown in another manner.
Accurately draw the two figures, one above the other. If now you fix
your attention carefully upon the two left ends of the two main lines,
you will find the one vertically above the other. Repeating the same
observation with the right ends, you find the same result. It follows
that both lines have the same length.[10] Similarly you will find
all the other illusions variable, according to the attitude of the
observer, and avoidable by practice. Therefore they cannot be classed
as genuine sensory experience.
Recently, stroboscopic movement has been investigated by some German
and American psychologists. Under appropriate conditions successive
stimulation by two lights at two points not too distant from each other
produces an experience of movement from the first to the second. But
if in making your observation in the attitude of introspection, you
find nothing but a “gray flash,” must we not be very careful, indeed,
before accepting the report of “movement”? Did not the observers of
Benussi[11] report similar experiences when their skin was touched
successively at two different points? But in their description the
movement did not usually occur upon the skin; it formed a curve through
empty space touching the skin only at those points. Evidently, such an
experience cannot belong to the world of touch alone; it is no genuine
sensory fact, then.
If all these are examples of illusions which, for psychological
purposes, have to be corrected by introspection, something must
transform pure sensory experience so that for untrained people it
assumes an illusory character. This is the case indeed, and in most of
our examples we can discover that “something” easily enough.
The man approaching us on the street does not grow larger, as for
simple optical reasons he should. The circle observed on the oblique
plane does not become an ellipse; it seems to remain a circle after
its retinal image has become a very flat ellipse. The white object
in the shadow remains white, the black paper remains black in full
light, though the first may reflect much less light than the second.
These three examples have one common property: the physical object _as
such_ remaining constant, while actual stimulation varies according
to more or less accidental conditions (of distance, of position,
of illumination), our experience agrees with the constancy of the
physical object much better than with the varying stimulation produced
by it. From this fact we draw the simple conclusion that all such
apparent constancy in experience is the product of _learning_. Since
our early childhood we have observed that a distant object is much
bigger, “really,” when we come near to it. In the same way we have
learnt that an object situated obliquely in relation to our glance
has quite another “real” form when we look at it vertically. Again in
the same manner we have observed that an object which we see under
abnormal conditions of illumination will show quite another “real”
brightness or darkness as soon as conditions become normal. As these
observations have been repeated millions of times, our knowledge of
the “real” size, the “real” form, the “real” brightness, has become so
intimately connected with what we see, that with time we seem to _see_
those constant “real” properties instead of the ever changing ones
corresponding to varying stimulation. In short, those facts, though
strange at first, are explained by _meaning_, and then completely fit
into our scientific system.
We have already seen that probably no experience escapes from the
influence of meaning. Hence we do not here introduce a new hypothesis.
If untrained people seem to “see” what from this viewpoint is a product
only of past learning, that again is an illusion, to be found generally
wherever we discover the indirect origin of the experience. Remember
the symbol +, which _looks_ like the sign for addition. Furthermore,
meaning as an after-effect of past observations must be _reproduced_
and reproduction presupposes reproducing factors in the actual
situation. Obviously, then, if we do not see the distances, the oblique
positions and the degrees of illumination, no reproduction of the
“normal” size, form and brightness will be possible. It is precisely
this which happens when the given objects are observed through holes
in a screen: since, under these conditions, the surroundings of
the objects and, with them, their _actual_ distance, position and
illumination are excluded from vision, corresponding reproductions
are made impossible. The objects must now be experienced exactly as,
in pure sensation, they have been the whole time. The same proof is
given by introspection. In our terms this procedure means an isolation
of local size, form and brightness from their context in an actual
situation. Thus it separates them from the factors determining the
reproduction of “meanings,” and under these conditions of isolation the
effect is that of pure sensation determined by the actual stimulation.
If in the case of size and form, after-images show a surprising
variation with the relative distance and position of the background,
this also is a direct consequence of our explanation. After-images are
localized upon the background. If the distance and the position of this
background operate as factors of reproduction, as in the previous case,
the same after-image will take on a great many apparently different
sizes and forms, relative to the changes of the background in distance
and position.