Experience, Art and
Understanding
- Classical Empiricism’s Passive
Account of Perception Unsatisfactory
- Kantian Theory of Mind as Active
- Experience is not a “Given,” but
rather Constituted by Active Mind
- Gestalt Principles of Perception
A Brief Detour: Locke’s Causal Theory of Perception
John Lock (1632 -1704)
was one of the three “British Empiricists” of the Enlightenment period.
In classical
empiricism, the mind is seen as passive:
John
Locke: the mind is a “tabula rasa”. (Blank Slate)
David Hume (1711 – 1776) refers to mental items as
“impressions.”
Bertrand
Russell (1872 – 1970) talks of “bare particulars”
Note:
the passive
metaphors they use. This is not the
first nor the last time that an unhelpful metaphor for understanding impeded
scientific progress. The impression (humor intended here) that this leaves is
that the mind “simply” receives information during
experience, that perception is a matter of just “seeing what’s really
there.” Knowledge, then, is a matter of
clearly and objectively as possible perceiving the “given.”
You
cannot get much more passive than being an inert rock. So, Locke’s metaphor for mind, a tabula rasa,
is as passive as you can get.
So
then for Locke (and Classical Empiricism generally):
•
The
only way to come to know the world is through sensory experience.
•
We
start life with a blank slate, "tabula rasa."
•
Points
out that
1.
there
is the world … and …
2.
there
are our ideas about the world.
(This put him, somewhat, in league with
Descartes. It raises some of the same
skeptical worries as it did for Descartes, about how reliably our perceptions provide
to us accurate information about the world “as it really is.” …Think “Matrix” type worries here.[1])
·
The world causes our ideas about/ perception
of it.
But
on this model, the mind contributes nothing to this “received”
information. The experience comes
pre-structured and prepackaged. It is a
“given” and clear accurate perception (knowledge) depends upon having an ”innocent
eye,” seeing the objective truth; seeing all and only what is “really there.”[2]
But…
As
widespread as this view of experience became and as influential as this view
was and continues to be, is it accurate?
Immanuel Kant and post-Kantian philosophers suggest the answer is a
definitive, “No.”
All experience is
mediated by Active Mind
This
was not appreciated until relatively recently.
(The myth of the “given” and the “innocent eye” still
persist today.)
Immanuel Kant and
Active Mind
Immanuel
Kant (1724 – 1804) marks an important development in Philosophy and conceptions
of “Mind.” At this point in the history
of Western philosophy, two great opposing traditions had come to an impasse of
sorts: Rationalism and
Empiricism
had both seemed inadequate to account for human knowledge. Rationalism seemed unable to account for
knowledge of our world of experience.
Conversely, Empiricism seemed unable to account for the necessary truths
of math and geometry or even the universality of the laws of nature. Taken to their logical extremes, both epistemological
orientations seemed to end in skepticism (either Descartes’s or Hume’s). Kant’s solution to the impasse was to
revision the very nature of knowledge and experience. The key Kantian insight is that the mind does
not merely receive information in the act of perception; rather the mind shapes
that information and constructs experience out of the raw
sense data it receives. This is
sometimes referred to as Kant’s “Copernican Revolution” in Epistemology. Rather than asking “Is knowledge/
understanding possible?” Kant asks “How is
it that knowledge/ understanding is possible?” Rather than asking “How does
knowledge impress itself onto mind?” (passive
metaphor) Kant asks “How does mind construct knowledge?”
To
accurately account for what is going on in perception it is necessary to see
human experiences as having different content, but a consistent “form.” If we were to abstract all content from human
experience we would arrive at the pure form of
experience. Think of it as blank
template into which mind pours all sensory information and thus arrives at a
coherent experience. Alternatively think
of my (very old, MS DOS based) Maillist program that
can organize records according to one and only one pattern. No matter what data it receives, it will
always organize them in the same fashion according to a single basic template.
In this case it was:
Field 1: First Name, Field 2: Last Name, Field 3: Telephone Number, Field
4: Street Address. Whether the data are
for my mom, my sister, the guy I knew from high school, etc., the record was always be: First Name, Last
Name, Telephone Number, Street Address. Thus I have
knowledge of how my 100th record and any other record will look (in broad
outline) in advance of actually reviewing the 100th record.
That
is, I have a priori knowledge of the 100th record. I know what the first field of my 100th
record is for instance: First Name. My knowledge is not grounded in the particular experience of my 100th record, though it
is grounded in experience in general. After
I get used to my program, and I have seen record after record after record, I
can see the records as having a consisting form with differing content. Now, I cannot make this distinction until I
have looked at the program’s records.
In fact, I cannot make this distinction until after I have look at quite
a few. So, my knowledge of the 100th
record arises from my experience with the program, though it is not grounded in
the experience of my 100th record per se. Further, I do not, nor
could I know what the CONTENT of the record is; but I do know its form. I know the form because when I am referring
to this program’s “records,” I am referring to products of its
organizing function which does not/ cannot change.
In
like manner, Kant claims that claims at after long experience we can make a
similar separation between the varying content of our experiences and their
common form. All our experiences of the world will be of an “external”
“extended: reality that obeys the laws of Euclidian 3-dimensional space. (E.g., the shortest distance between two
points is a straight line.) All our experiences
of the world will be temporally sequential as well. These necessary and universal features of our
experiences are so, according to Kant, not because “That’s how the world really
is,” but rather because that’s the only way human mind organizes experience.[3]
Another
illustration of what Kant has in mind here can be seen in those “Magic Eye”
posters.[4] At one moment they look like flat two-dimensional
images. The next they look like a three-dimensional
image. What is different from one moment
to the next? Is the poster giving you
something different when it looks two dimensional from what it is giving you
when it looks 3 dimensional? No. What is different is what YOU are doing, the activity
of you mind in perception. But this is just a more noticeable example of
what the mind does constantly. The
reason you see reality as three-dimensional is NOT because “that’s how the
world really is,” but rather because that’s how your mind (and every other
normal, healthy human mind) is shaping its sense data. Theoretical physicists might talk of reality
in multiple dimensions, but even they don’t perceive it that way. They come to that understanding purely
theoretically. Perhaps aliens from outer
space perceive reality in more dimensions.
Perhaps God perceives it that way.
But not humans. Not now, not ever,
says Kant. We will always only “image”
the world as three dimensional. And the
same thing goes for unidirectional time.
Kant
is very specific about what these forms and categories of experience are, but
I’ll only refer to a few for illustration purposes.
Space and Time are the
two pure forms of experience according to Kant.
All
human experience will/ must conform to 3-dimensional Euclidian Space.
All
human experience will/ must conform to unidirectional time. (Past to present to future).
Opens the door to
Radical Relativism:
Kant
believed that our (human) empirical knowledge was universal (NOT RELATIVE)
because the pure forms of experience and the categories of thought were
universal for all humans. Therefore, he was certain that what is true for one
human is true for all humans.[5]
BUT....one
might object to Kant’s view.
For
instance, what if we do NOT all put the world together in basically the same
way (e.g., woman according to a female template, men according to a male
template)? If “Men are from Mars and
women are from Venus” then we are not experiencing the same worlds because we are
building our worlds, shaping our experience according to different templates. We are, in a very real sense, living in
different worlds, and truth must be relativized to groups of cognizers who possess the same template. Rather than univalent, truth becomes bivalent
or, perhaps, multivalent. Truth is
potentially as multifaceted as there are minds, and no basis would exist for
claiming that any particular worldview was privileged
among the plurality.[6]
20th
Century Philosopher and Art Historian Applies to Art
Both
Ernest Gombrich (1909 - 2001) and Nelson Goodman (1906-1998)
represent 20th century thinkers who echo the key Kantian Insight:
Mind is NOT passive in experience, but rather active. This insight is crucial to understand what is
going on in art, among a great many other things, they claim. Nelson Goodman, like Kant, denies the
simple/passive view of perception. (There are no “bare particulars;” there are
no “innocent eyes.”) The mind is active
from the onset. The shape and content of experience/perception depend on
the work the mind does (focusing, past experience,
expectations). Perception is a matter of
focusing, ignoring, overlooking, interpreting. Thus, there is no difference between perceiving
and interpreting according to Goodman.
Ernst Gombrich “no innocent eye”
“The
eye comes always ancient to its work, obsessed by its own past and by old and
new insinuations of the ear, nose, tongue, fingers, heart, and brain.”[7]
It does not much mirror the
world as take and make.
Goodman
argued against the persistent myth of the “Pure Empirical Fact.” Classical empiricism would claim that
perception is “automatic,” non-cognitive, or minimally so, merely “biological.” But Goodman insists that perception is NOT
non-cognitive and that what we “see” is a direct result of what and how we’ve
been trained to “read” the world. His
model for perception can clearly be seen in other modes of perception. Consider wine tasting, X-ray reading, hearing
bird calls, and art. Despite the fact that perception
may (seem to) be automatic, there is nothing basic or given in perception.
Nelson
Goodman admits that perception may indeed be automatic, but that does not mean
passive. It does not mean inevitable. Consider the act of hearing a sentence
spoken in your native language. You
cannot help but hear it as a sequence of intelligible (meaningful) sounds.
Though this can be automatic, this is a learned, cognitive skill. You
are interpreting those phonemes whether your realize
it or not. And for those with poor
command of the language, it takes effort.
But for those with good command over the language it is “second
nature.” Automatic perhaps, but nevertheless learned and acquired
actively.
For
another example, think of visually ambiguous images, specifically the
“Duck/Rabbit.”
This phenomena shows why the old model of passive perception
is inadequate for understanding how perception works. For Locke, he thought it was enough to talk
about the object, the perceiving organ and the
perception. On this view the object
“impresses” itself on the mind and the mind is simply the inert passive
recipient of the information.
We
can complicate this simple model a bit by talking about the object, the organ,
the retinal image, and the perception.
But again, on this model, the perception is understood as the inevitable
product of the retinal image. Mind plays
no active role.
But
in the case of ambiguous images, the poverty of this view is revealed. In the case of the “Duck/Rabbit” image, I can
see a duck or I can see a rabbit, that is I can have
the duck-perception or the rabbit-perception and which perception I have cannot
be explained in terms of the object, the organ or the retinal image. When I have the duck-perception, the object,
the perceiving organ and the retinal image are the same as when I have the rabbit-perception. There must be some other factor that
explains the difference in perception, and that factor is the activity
of mind. The old Lockeian model of mind simply cannot
account for the phenomena.
After
Kant theorists argue against passive accounts of mind and knowledge and
emphasizes the role of the faculty imagination in the formation of experience.
This opened the door to the position that there is not merely one right way to
experience or to know. If our experience
of the world is a product of what the world gives us and what we do with it (how
we “image” it) then it follows that there may be as many different experiences
of the world as there are “imagers” and no one is in a privileged position with
respect to its rivals.
This
realization gave rise to the Post-modernists’ notion
that there is no one point of view from which Truth can be determined. Imagine
two groups of people, one who could only see the duck and one that could only
see the rabbit. Which group is seeing
what is “really there” and which group is wrong? Well of course we see in this example that there
is no reason to think that either group is privileged here. Further, the only reason we could have to say
that one of them has truth and the other has falsehood and is not seeing the
world “correctly” would be to advance a political, economic
or social agenda.
Perception: 8 Gestalt
Principles and Bottom-Up verses Top-Down Processing
1.
Distal and Proximal Stimuli vs. Percepts
The root of Distal is clearly
related to the English word distance, and should, I hope communicate
the idea that this it the "outside" stimulus.
Proximal essentially means "near" or "close to", and that root is also used in the English word approximate, which basically means "to make or come close to".
The term Percept comes from
the same root that has a concrete meaning of "to catch" or
"caught". Pretty clearly, it's what we end up with, an identity that
we can use in further processing.
Distal
and Proximal Stimuli vs. Percepts
When
you see somebody walking down a hallway away from you, you don't think that the
person is shrinking, despite the fact that the image
on your retina is doing precisely that. Further, if the person were
shrinking, the image on your retina would be pretty much the same, but you
would likely misinterpret it as the person is moving away for you.
Why?
Well,
one explanation is that because we have adapted to live in the real world and
in the real world that sort of visual simulation is far more often tied to
objects retreating from you rather than staying in a fixed position and
shrinking. This, in effect, is a
perceptual heuristic.
Heuristic:
a
strategy for organizing a great deal of information quickly via a “rule of
thumb” where examining each datum individually is not practical. These are used to control and facilitate
problem solving (in human beings and machines).
Our
visual systems must be able to construct depth information from the 2-d image
that falls on our retina. This requires a perceptual heuristic which Gestalt
theory refers to as “size constancy.” We find depth “cues” and use them to
interpret the retinal image and come up with a depth-independent size. We automatically judge the object of our
perception to be remaining the same size.
This “assumption” is what allows us to further judge its relative speed,
location and trajectory.
These
heuristics a valuable under “normal" life circumstances, but they allow us
to set up all sort of peculiar illusions under “abnormal” circumstances.
For
instance, given two circles of the same size, one surrounded by large circles
and the other by small circles, the circle surrounded by small circles will
appear to be larger. This is because our
visual system uses the size of the surrounding "context" circles as a relative depth cue. Because the circle
with smaller surrounding circles is judged farther away, the size of the circle
itself must be larger.
A
similar situation happens with the so-called Ponzo
Illusion:
The
presence of a “vanishing point” is suggested by an acute angle. It can be enough to make us perceive a size
difference, even between lines of equal length when they differ in distance
from the angle.
In
both cases, the sizes of the retinal
images for the shapes involved are identical (the proximal stimuli
are identical) but our visual system gives us the percept
that the two shapes differ in size.
Since under normal circumstances the presumed distal stimulus
in a three-dimensional scene is further presumed to be larger.
2.
Gestalt Principles (Seven)
1.
Similarity (Perceptual Grouping)
Similar
objects tend to be perceived as “grouped together.” “Similar”
is understood across one or more dimensions (such as color, shape, orientation,
and other simple visual features).
Note that this image is seen as columns of circles and
squares (rather than rows of
“circle, squares”).
To explain this we must refer to
what we are doing with the image, not merely “what it there.”
2.
Proximity (Perceptual Grouping)
Objects
that are closer together are perceived as forming a “group.”
Here
this is seen as rows, not columns.
Interesting
(aesthetically?) things happen, however, when similarity and proximity work
against each other.
In
the last of the three, the images strikes us as
neither/both rows or columns.
3. Good Continuation (Perceptual Grouping)
Objects
or pieces of objects (especially edges and other line segments) are grouped
together to the extent that we can "trace" a good continuation
between the pieces. The Kaniza triangle illusion is a
classic example of this.
Kaniza Illusion
|
|
Name
the beginning and ending points of the lines here.
More
than likely you named a/b and c/d NOT a/c or a/d.
4.
Closure (Perceived Object Shape)
Essentially
a form of Good Continuation; there is
a tendency to "close up" the boundaries of objects even when there
are small gaps or openings.
Here
we tend to see three broken rectangles (and a lonely shape on the far left)
rather than three 'girder' profiles (and a lonely shape on the right). In this
case the principle of closure cuts across the principle of proximity, since if
we remove the bracket shapes, we return to an image used earlier to illustrate
proximity...
5.
Symmetry (Perceived Object Shape)
Everything
else being equal, we prefer to group objects so as to
make symmetrical forms rather than asymmetrical forms. The symmetrical areas tend to be seen as figures against
asymmetrical backgrounds.
6.
Smallness
Smaller
areas tend to be seen as figures against a larger background. In the figure below
we are more likely to see a black cross rather than a white cross within the circle
because of this principle.
As
an illustration of this Gestalt principle, Coren,
Ward and Enns (1994, 377) argue that it is easier to see Rubin's vase when the
area it occupies is smaller. The lower portion of the illustration below offers
negative image versions in case this may play a part. To avoid implicating the surroundedness principle I have removed the
conventional broad borders from the four versions. The Gestalt principle of
smallness would suggest that it should be easier to see the vase rather than
the faces in the two versions on the left below.
Then
there is the principle of surroundedness,
according to which areas which can be seen as surrounded by others tend to
be perceived as figures.
Now
that we are in this frame of mind, interpreting the image shown above should
not be too difficult. What tends to confuse observers initially is that they interpret
the white area as the ground rather than the figure. If you couldn't before,
you should now be able to discern the word 'TIE'.
7.
Common Fate (Perceptual Grouping and Perceived Object Shape)
Objects
are grouped together to the extent that they are perceived as
"traveling" together. This is most literally true when they are all
moving in the same direction or along some common path.
8. Pragnanz The over-riding law of Gestalt Grouping is that we group objects to obtain the "best" or most parsimonious interpretation. Unfortunately, this term is very hard to translate from the German.
Top-down vs. Bottom-up
processing
We
must get clear on the distinction between top-down and bottom-up
processing.
The
notion of "bottom-up” implies effects
seen early in processing, where our previous experience or acquired conceptual
frameworks about the world plays little role.
For
abilities like object recognition, word recognition, and
face processing, however, our previous experience and acquired arsenal of
concepts is not only very important, but influences
our perception; This then is top-down processing.
Interestingly,
top-down effects are most clearly seen when they are absent. The very same
arrangements of chess pieces on a game board (or even the pattern of Xs and Os
in a game of tic-tac-toe) are actually either
perceived as meaningful units (for an expert who has background
learning which admits a meaningful interpretation) or as random (for one who
lack such a background).
Because
of the role that the Gestalt plays, knowledge, perception
and memory are affected. Notice that
knowledge, perception and memory for meaningful
configurations are superior to that for random ones.
Similar
effects can be measured for the ability to recognize crudely scrawled letters
or mumbled phonemes when they occur in the context of a word or “meaningful”
sentence. This is why
people were able to “hear” all sort of things when music recordings were played
in reverse (back-masking). It misses the
point to ask, “But is that really what’s there?” since the
implication of the research is that it’s never “there,” but only up here
(I’m pointing to my cranium now.)