person walking away does not seem to become a dwarf -- as he
should; a black glove looks just as black in the sunlight as in shadow
-- though it should not; when a coin is held before the eyes in a tilted
position its retinal projection will be a more or less flattened ellipse;
yet we see it as a circle, because we know it to be a circle; and it
takes some effort to see it actually as a squashed oval shape. Seeing is
believing, as the saying goes, but the reverse is also true: knowing is
seeing. 'Even the most elementary perceptions', wrote Bartlett, [7] 'have
the character of inferential constructions.' But the inferential process,
which controls perception, again works unconsciously. Seeing is a skill,
part innate, part acquired in early infancy.* The selective codes in this
case operate on the input, not on the output. The stimuli impinging on the
senses provide only the raw material of our conscious experience -- the
'blooming, buzzing confusion' of William James; before reaching awareness
the input is filtered, processed, distorted, interpreted, and reorganized
in a series of relay stations at various levels of the nervous system;
but the processing itself is not experienced by the person, and the
rules of the game according to which the controls work are unknown to him. The examples I mentioned refer to the so-called 'visual constancies'
which enable us to recognize that the size, brightness, shape of objects
remain the same even though their retinal image changes all the time;
and to 'make sense' out of our sensations. They are shared by all people
with normal vision, and provide the basic structure on which more personal
'frames of perception' can be built. An apple looks different to Picasso
and to the greengrocer because their visual matrices are different.
Let me return once more to verbal thinking. When a person discusses,
say, the problem of capital punishment he may do so 'in terms of'
social utility or religious morality or psychopathology. Each of these
universes of discourse is governed by a complex set of rules, some of
which operate on conscious, others on unconscious levels. The latter are
axiomatic beliefs and prejudices which are taken for granted and implied
in the code. Further implied, hidden in the space between the words, are
the rules of grammar and syntax. These have mostly been learned not from
textbooks but 'by ear', as a young gypsy learns to fiddle without knowing
musical notation. Thus when one is engaged in ordinary conversation, not
only do the codes of grammar and syntax, of courtesy and common-or-garden
logic function unconsciously, but even if consciously bent on doing so
we would find it extremely difficult to define these rules which define
our thinking. For doing that we need the services of specialists -- the
semanticists and logicians of language. In other words, there is less
difference between the routines of thinking and bicycle-riding than our
self-esteem would make us believe. Both are governed by implicit codes
of which we are only dimly aware, and which we are unable to specify.*
Habit and Originality Without these indispensable codes we would fall off the bicycle, and
thought would lose its coherence -- as it does when the codes of normal
reasoning are suspended while we dream. On the other hand, thinking which
remains confined to a single matrix has its obvious limitations. It is
the exercise of a more or less flexible skill, which can perform tasks
only of a kind already encountered in past experience; it is not capable
of original, creative achievement.
We learn by assimilating experiences and grouping them into ordered
schemata, into stable patterns of unity in variety. They enable us to cope
with events and situations by applying the rules of the game appropriate
to them. The matrices which pattern our perceptions, thoughts, and
activities are condensations of learning into habit. The process starts
in infancy and continues to senility; the hierarchy of flexible