determine introspectionism’s criteria for “true” sensory experience,
we find the following two: _First_, local sensation depends upon local
stimulation; it does not depend upon other processes in the nervous
system, not even upon those issuing from neighboring points in the
same sense organ. There is only one physiological assumption which
would explain such an independence of local sensation, namely isolated
conduction from one point of the sense organ toward one point of that
region the activity of which is accompanied by sensory experience.
This, however, is nothing but the first half of a reflex arc, so that
in this respect introspectionism agrees completely with behaviorism. If
experience does not seem to obey this principle universally, that is
because we have a _second_ principle: At a higher level of the nervous
system at least, connections may be built during development, which did
not exist originally at birth. A corollary of this principle is: that
certain experiences will be followed and accompanied by others, or by
reproduced material corresponding to them, in the form of “meaning.”
Though this term sounds a trifle more refined than “conditioning,” in
the physiological considerations of most introspectionists it means the
automatic formation of new conducting connections, just as conditioning
does for behaviorists. Again there is no real difference between the
one viewpoint and the other.
During their lively controversy over whether introspection, or the
objective observation of behavior, is the true procedure, it does not
occur to either party that some other theme might be more important
and urgent, namely, that concerning their common convictions as to
the basic type of nervous process underlying experience or behavior.
This assumed basis appears self-evident to both of them. And since
so much is taken for granted by both, we should not be surprised to
find the same conservatism in introspectionism which is so striking in
behaviorism.
That psychology is a very young science and that therefore its future
and its progress will probably depend upon discoveries unsuspected at
the present time, seems not to be realized by most introspectionists.
In sensory experience, at least, the essential traits of all possible
and acceptable observations are fixed for them before they begin to
observe. Consequently, they show an almost negativistic attitude
whenever an observation does not agree with that preëstablished
truth; and their experimentation tends to become a rather defensive
and critical procedure. If others point to some new observation
that does not exactly conform to the fundamental concepts mentioned
above, they are eager to quell the disturbance, and to dismiss it by
appropriate introspection and discussion. I admit that criticism of
new observations is a healthy procedure in science; yet I have known
introspectionists who spend their scientific lives in a critical
defense of what they believe to be unquestionably true. In these
circumstances one cannot expect them to have a youthful interest in
free observation and naïve discovery. If the main points are already
established, and if further observation cannot give us essentially new
hints, why should we strain our eyes in the chaos of a large world
behind which, anyway, the old truth _must_ be hidden everywhere? As
William James would say, whatever happens is a “nothing else but” for
them; it can be reduced to an old story. In this attitude again there
is no real difference between behaviorism and introspectionism, though
each defends its conservatism by different methods. But there is a
real difference between preferring to be at rest upon preëstablished
ground or preferring adventure and intellectual curiosity which
take you full-sail into the open sea of experience. If you choose
the latter, perplexity will be your usual feeling, and a glimpse
of clear-shaped coasts a rare reward on some Sunday of research.
But the next generation will have more Sundays as a result of your
exploration. Certainly, neither this nor the next generation is needed
in psychology, if we stay at home and know beforehand that no other
sites can be found in the future.
I do not see, then, why introspectionism should be preferred to
behaviorism, or vice versa. They are so much alike in their
fundamental opinions and in their general attitudes, that all their
wrangling seems like a family quarrel to the onlooker. And it is among
precisely those themes, over which they do not quarrel, that we shall
find the problems of _gestalt_ psychology. For the present, however, we
shall pursue this question: Is it true that the processes underlying
experience and behavior are determined mainly by the connections of
given nerve-paths, and that some change in the conductivity of these
connections is the secret of an individual’s development?
BIBLIOGRAPHY
M. Bentley: _The Field of Psychology_. 1924.
K. Koffka: _Gestalt Psychology_. The Psychological Bulletin. 1922.
D. Katz: _Die Erscheinungsweisen der Farben_. Ergänzungsband 7 der
Zeitschr. f. Psychol. 1911.
W. Köhler: _Uber unbemerkte Empfindungen und Urteilstäuschungen_.
Zeitschr. f. Psychol., 63. 1913.
W. Köhler: _Akustische Untersuchungen_ III. Zeitschr. f. Psychol.,
72. 1915.
IV
_Dynamics as Opposed to Machine Theory_
SOMETIMES people are conservative and right at the same time. Still it
seems highly improbable that in our very young science conservative
opinions should be right, opinions which are opposed to almost all
experience, and which have scarcely been examined, because they have
been protected by the “meaning theory.”
Some of the physiological presuppositions lurking behind introspection
are hardly convincing if they are once examined critically. In one of
our examples, apparent movement, produced by touching successively two
different points on the subject’s skin, is not admitted as a “true”
sensory fact because usually it forms a curve through empty space, only
the ends of which are felt on the skin (cf. p. 79). But why should all
the experiences which depend physiologically upon the stimulation of
a sense organ be localized in the same place where that sense organ
is localized as an experience? In vision, this is not the case; forms
or colors are not seen at the place where we feel our eyes. Neither
are sounds heard generally where we localize our ears. Behind the
introspectionist’s argument there seems to be some primitive confusion
of the peripheral processes produced by stimulation with sensory
experiences depending upon them of the physiological locus of the first
with the sensory localization of the second.
I think this case is typical because it shows that the apparent
self-evidence of the argument from physiology prevents a critical
consideration of it, whereas any examination of the case would efface
the self-evidence of the assumption. In this chapter I shall try to
show that even the main assumptions of introspective psychology and
behaviorism are by no means axiomatic, though they agree with a common
prejudice which is thousands of years old.
We have seen that those assumptions are utterly dependent upon the
explanatory force of the meaning theory. Unprotected by this theory
those assumptions would be destroyed by any observations which
contradicted them. Recent experimental work on some of the experiences
discussed as examples in the last chapter bears upon the explanation
of those experiences in terms of meaning. In order to _learn_ that
sometimes white is black and vice versa, an individual needs time and
_much_ experience, largely because he has to learn it so thoroughly
that finally the products of learning will be “projected” into his
field of vision as definite nuances of brightness instead of the
“true” ones. We should expect, then, that young or primitive subjects
would not show the “constancy of brightness” to any considerable
degree. But when the experiment was made with young chicks under very
severe conditions, they were found to possess approximately as good a
“constancy of brightness” as I do.[15] Similar experiments made on the
“constancy of size” with children (from two years up), and with young
apes, also gave positive results.[16] Though it is difficult to exclude
slight influences of meaning upon the phenomena in question, it seems
altogether unlikely, even without further work, that _in principle_
these phenomena are produced by meaning. I do not deny that objective
experience is imbued with acquired meaning in many respects. But, as
has been said before, where this influence is not demonstrated in the
history of individuals, or somehow demonstrated independently, no
indirect argument is strong enough to be accepted in lieu of such a
demonstration.
Since in these instances the meaning theory begins to relinquish the
field, a radical change in fundamental principles becomes unavoidable.
The phenomena we have here discussed, such as constancy of size, form,
localization, speed and brightness, the stroboscopic movement, the
well-known illusions and so forth, become just as decisive in our
understanding of genuine sensory experience as the so-called “normal”
cases of introspection, i.e., as the observations that at a given
distance and on a homogeneous background seen size depends mainly upon
retinal size; that (apart from contrast and other exceptions) in a
given illumination seen brightness depends upon retinal intensity, and
so on. In these cases, usually regarded as normal, size, brightness,
etc., are found to vary with the properties of local stimulation,
because conditions of surrounding stimulation are practically constant.
For the same reason, by assuming the introspective attitude, we may
find experiences corresponding to local stimulation even in those
other cases where _naïve_ experience shows the contrary, because by an
analytical attitude the influence of surrounding conditions temporarily
can be suppressed to a high degree.
But we do not suppose such isolation to be the normal state of affairs.
If in all the examples given we accept direct experience at its
face-value, our fundamental assumption about the processes underlying
experience and behavior must be opposite to the assumptions of both
introspectionists and behaviorists; i.e., instead of reacting to local
stimuli by local and mutually independent events, the organism reacts
to an actual _constellation_ of stimuli by a total process which, as a
functional whole, is its response to the whole situation. This is the
only viewpoint which can explain how to a given local stimulus there
may correspond altogether different experiences as soon as surrounding
stimulation is changed. We face this dilemma: _Either_ we take the
traditional horn, in which case direct experience and everyday behavior
are irrelevant to our psychological analysis, _or_ we trust all
kinds of experience impartially, in which case we have to accept the
suggested radical change in physiological theory.
“Total process” and “functional whole” are, however, terms which sound
rather vague, or even mystical, to most scientists. It would be well
to present our physiological assumption in more definite detail. This
may be done. And at the same time we shall obtain a more general view
of our problem if we ask ourselves just why the prevailing viewpoint
should have appeared so utterly convincing to the last generations.
The chief reason seems to be that sensory experience is in most cases
quite orderly, and that the behavior, which accompanies it, is no
less orderly. Now since the early days of European science man has
been deeply convinced that the processes of nature, if they are left
to their own “blind” play, will never produce anything like order.
Why should they? Does not the accidental intercourse of inorganic
natural forces around us produce chaos and destruction everywhere?
We have been able to formulate some of the rules according to which
isolated processes occur in nature necessarily, but where many
processes influence each other without control, no reason is apparent
why confusion and general disorder should be avoided or why the
whole complex should develop toward an orderly distribution. _Without
control_--this appears to me to be the decisive point. For as soon
as man began to restrict the possibilities of natural processes by
appropriate rigid arrangements, the same forces which apparently would
have produced chaos, destruction or disorder without control were
guided toward orderly function in the service of mankind. This has been
man’s conception of nature for thousands of years; in modern times the
engineer still enforces in the same manner the functional order in his
machines. The “blind” forces of nature are allowed to make the machines
_move_, but the _order_ of the movement is enforced by those special
arrangements which are the essential traits of the machines and the
pride of their inventors.
From this viewpoint, which prevails even in current theory, a young
science will tend to presuppose the existence of “special arrangements”
wherever the distribution of forces or processes in nature is found to
be orderly. Aristotelian astronomy is a good example. The movements of
stars in the sky maintain a remarkable order, so different from what
one would expect to occur in “free” nature, that something special
must control those celestial events. Obviously, the possibility of
a star running wild or a planet going astray must be excluded by
rigid arrangements. Therefore, in Aristotelian theory, the stars
are fixed upon those famous crystal spheres, the turning of which
forces the stars to move in their apparently regular orbits. There
are even engineers, viz., the stellar deities, who watch and control
the machinery. Three hundred years ago this conception was still
entertained in pious awe. The functional value of its crystal spheres
in the orderly sky, however, was precisely the same as that of any
rigid arrangement enforcing orderly processes in our machines. There
is a deep longing for rest and safety in man. Early in history it
was satisfied by such a primitive belief, the content of which may
appear crude, narrow and absurd to modern eyes. What was so shocking
in Galileo’s astronomical discoveries? That there was so much going
on in the sky and the astronomical order was so much less definite
than one could happily believe before! If the heavens begin to show a
lack of rigid institutions, similar to the flexibilities we observe on
this planet, who can feel secure in his most important beliefs? There
was much exceedingly human, miserable fear in the furious attacks
directed against Galileo by the Aristotelians of his time, and the
fear was evoked by the amount of change and dynamics Galileo found
in the world as the home of mankind. One may well suppose that the
excitement produced by Harvey’s discovery of the circulation of the
blood contained a similar element of fear, because the conception of
man as a rigid structure was suddenly disturbed by the idea of internal
unrest, accompanied by the possibility of its fatal cessation at any
moment.
In any case the same general motive expresses itself in the well-known
tendency of early biological thought to explain all remarkable
properties of organic life, and most of all its orderly course, by
special arrangements enforcing order. In the time of Descartes, his
so-called mechanical interpretation of organic functions may have
been bold enough; still he was absolutely conservative in assuming
without a moment’s hesitation that--apart from the influence of one
engineer, the soul--all delicate, vital processes were enforced by
special arrangements, connections and channels throughout the whole
body of man. Figuratively, the organism was for him what the sky had
been for Aristotle--full of crystal spheres. He did not know the laws
of dynamics, it is true. But though we know much more about them
since his time, the main changes in biological theory seem to have
been refinements of his type of thinking rather than discoveries of
new general concepts about organic order. Even in our own day, though
doubting the validity of that machine-conception as a satisfactory
explanation of _all_ organic order, biologists do not give us any
other clear general principle whereby to understand the orderliness of
organic function.
We shall see at once what the possibilities of alternative explanation
are, if we try to get a more concrete picture of the characteristic
trait in Aristotelian and biological machine theory.
In physical processes we find two sorts of factors determining
events at every moment. In the first class belong the actual forces
of the process itself; they represent the _dynamical_ side of it.
In the second class we have those properties of the system which
may be regarded as _constant conditions_ of its events. So, at the
present time, the elementary charge of one electron seems to be an
unchangeable condition of all physical events. In a conducting wire
the specific conductivity of its material will be regarded as such a
given condition. On the other hand, the electrostatic forces of the
current itself driving an electron along that wire are an example of a
dynamical factor. But among the constant conditions two further classes
may be discriminated. In the case of electric current it is one given
condition that all moving charges have the amount of electronic charge
or multiples of it. Another condition of the current is given by the
mode of the distribution, the connections and the spatial order of the
conducting material, i.e., a network of wires.
Immediately the fact becomes clear that, whereas in all systems of
nature, in addition to their own dynamical forces, processes depend
upon conditions of the first class, the influence of special conditions
of the second class may dominate in one case and be at a minimum in
another. When we consider the play of electrons in an atom or the
integration of atoms forming a molecule, we do not find a special
topography laid down beforehand which determines those processes.
At every moment, in addition to constant properties of the material
(first class of conditions), the actual _dynamical_ situation develops
the next event out of itself. If we wish to refer to the second
class of conditions in such a case, it would be to point out that
the topographical situation consists of physical space without any
special determining or limiting feature in it. On the other hand,
the equilibrated distribution of electric charge on the surface of a
conductor depends upon the rigidly given form of that conductor as a
limiting topographical condition; and in a steam-engine the piston
can move only one definite way, determined by the rigid walls of the
cylinder.
This leads to a classification of physical systems which is decisive
for our problem. In all of them process is to be regarded as
necessarily determined, but among the various cases we find enormous
differences in the relative influence which limiting topographical
conditions, on the one hand, and the play of actual forces, on the
other, exert upon the course of events. Wherever we have given
topographical conditions, preëstablished and not changeable by
the process itself, their existence means the exclusion of some
dynamical possibilities and the restriction of the process to only the
possibilities compatible with those conditions. Electric charges may
move through the conductor in various ways, but they are _prevented_
from leaving it, and so their final distribution depends upon its form.
Again the electric current may have one direction or the other, its
distribution may vary enormously; but, if the wires are surrounded by
isolating material, the dynamics of the current itself will remain
restricted to the interior of the wires. The most extreme case will
consist in a system where preëstablished topographical arrangements
exclude all processes except only one; and an example of this type is
given by the piston’s motion strictly confined between the walls of the
cylinder.
In this case the steam in the cylinder exerts its pressure in all
directions, but, owing to the topographical conditions, it is not
allowed to do work except in _one_ direction, in that, namely, in
which the piston is free to move for a certain distance. Consequently,
nothing but the motion as such is determined dynamically in such a
system, whereas its direction is strictly enforced by topographical
arrangement.
Now this is exactly the relation between dynamics and preëstablished
topographical conditions which we find in typical industrial machines.
The number and forms of special one-way functions which may be enforced
in such systems are enormous and varied. Still the general principle
is everywhere the same. Sometimes a little more than the minimum may
be left to dynamical determination, but at present no one would
construct mechanical systems for industrial purposes where the form and
distribution of process would be to any considerable degree a matter of
dynamics.
It is the same idea again which occurs to Aristotle when he views the
remarkable order of celestial movement. His spheres are topographical
conditions enforcing that order. And since Cartesius, neurologists have
worked with the same concepts wherever they have dealt with orderly
organic function in higher animals and in man. It is not the dynamics
of nervous processes as such which they suppose tend toward coördinated
function. Vitalists may have such a mystical idea! Rather, special
anatomical topography is the only explanation for order; and by it the
dynamics of process are compelled to produce orderly results.
Once more we return to point out that it is the same conception which
forms the common basis of both introspectionism and behaviorism, so
far as their physiological principles are concerned. Take vision, for
example. How many things may be present simultaneously in one actual
field! Still, excluding extreme peripheral regions, there does not seem
to be any confusion for the most part. One object appears separated
from all others, and the sharpness of its contours is evidence of a
high accuracy of function. But the field is not only clear as such; it
also corresponds admirably to the physical realities. Points which
are neighbors in physical space are neighbors also in the visual
field; the center of a circle in physical space appears as the middle
of a symmetrical figure in vision, and so forth. All this order is
as remarkable as it is necessary for our response to the objects
which, in the form of bodily movement, must be adjusted properly to
the physical world. The order of projected images upon the retina
is easily explained by the properties of the pupil, the lens and so
forth. But what about the processes which, streaming from here into
the brain, will eventually determine experience and behavior? Since
experience and behavior show a similar order, this order must have
been enforced or preserved throughout the entire process. Only one
kind of explanation seems, then, to be possible: In the nervous system
we have a topographical arrangement, preventing confusion and mixture
during conduction. Indeed, if from each point of the retina local
processes are conducted on definite and isolated paths toward their
final cortical termini, and if the totality of these termini somehow
reproduces the geometry of local retinal processes, then the dynamics
of the process are excluded completely from the determination of its
own direction and distribution. In terms of such a construction of the
facts we can feel secure that the most important property of vision,
i.e., its accurate order, is guaranteed by so trustworthy a factor as
preëstablished anatomical conditions.
Similar considerations would lead to similar results in the case of
touch and hearing. But what about learning and habit formation?