Thus the metaphysical assumptions of Gestalt psychology lead us (as Koffka somewhat reluctantly admits [31]) back to the kind of 'faculty psychology' abandoned about a century ago; and ultimately to Plato; whereas Behaviourism leads us ultimately back to the atomism of Democritus combined with the scepticism of Ecclesiastes. Köhler's chimps look at the world through the 'eyes of the soul'; Hull's rats are wired marionettes jerked about by a non-existent puppet player.
Putting Two and Two Together
Contiguity is the Deus ex machina of Behaviourists such as Watson and Guthrie, and the bête noire of Gestaltists, who reject association by contiguity as 'blind' and 'meaningless'.
'There is not a single example', writes Köhler, 'of an effect produced by the interaction of two things or processes quite independently of their properties. Nevertheless this is the character of the classical law of association as we find it stated in most textbooks.' [32] I cannot remember having seen anywhere the law stated in quite that form, and Köhler's description is obviously a caricature; how can two things interact 'independently of their properties'? Does not the term 'interaction' imply that properties of the interacting processes are involved? Contiguity -- the overlapping of two events in spacetime -- of a regular, recurrent kind is the base on which inductive inference can build. This applies, impartially, to the dog learning that the gong is a signal for food, and to Kepler observing that the tides follow the moon. Kepler concluded that there must be a causal connection between the moon and the tides, although his ideas of gravity were of the most erratic kind; and his was a correct approximation, followed by the closer approximations of Newton and Einstein -- though we are perhaps as far as ever from 'grasping the intrinsic, material relations' between bodies acting at a distance. More often than not, the starting point of the scientist's inquiry is 'post hoc, ergo -- let's hope -- propter hoc'. Even in the rare limit case of the Eureka process, contiguity -- the simultaneous activity in the mind of the two matrices which are to be integrated -- provides the link; and the provider may even be 'blind chance' -- a fungus sailing through the window into Fleming's laboratory.
Learning, then, in the most general sense of the word, consists in putting two and two or A and B together. It may be done gradually, by plodding through hypotheses and eliminations; or all of a sudden, following upon a single implicit try. A and B may stand for recurrent features abstracted from a series of perceived events; or A may be a signal-pattern requiring the correct choice of reaction B among other possible choices; A and B may be codes of behaviour or universes of discourse, each complete in itself and adequate for routine tasks; but to solve a certain new task they must be put together. If the subject is ripe for the solution, the putting together can occur in a single flash. But let me repeat that such lightning inductions based on a single case are possible only if both M1 and M2 are well-established, flexible matrices which the subject knows 'inside out'. Only those chimpanzees discovered spontaneously the use of a stick as a rake who had previously played with sticks; those who had not, though of equal intelligence, failed to see the light.
There is a rather striking parallel between the present interpretation of discovery as a bisociative process, and the conclusions which Hebb reached regarding the sudden appearance of new insights:*
The sudden activation of an effective link between two concepts or percepts, at first unrelated, is a simple case of 'insight' . . . [33] Insight, as a sudden perception of new relationships, can result from the simultaneous activity of two conceptual cycles in adult learning. [34] The insightful act is an excellent example of something that is not learned, but still depends on learning. It is not learned, since it can be adequately performed on its first occurrence; it is not perfected through practice in the first place, but appears all at once in recognizable form (further practice, however, may still improve it). On the other hand, the situation must not be completely strange; the animal must have had prior experience with the component parts of the situation, or with other situations that have some similarity to it. . . . All our evidence thus points to the conclusion that a new insight consists of a recombination of pre-existent mediating processes, not the sudden appearance of a wholly new process. [Hebb's italics] Such recombinations must be frequent in man's everyday living, and in a theoretical framework we must consider them to be original and creative. . . . [35]
Hilgard (1958) came to similar conclusions: 'Because all learning is to some extent cognitively controlled, the distinction between blind learning and learning with understanding becomes one of degree.' [36]
NOTES
To p. 581. Köhler made a distinction between one-box experiments and the building of two- or three-storey structures. In the latter additional difficulties arise from what he called the chimp's 'lack of a feeling for statics'. This handicap no doubt adds to the anlmal's perplexity; but Koko's and Sultan's behaviour in experiments of this kind (cf. pp. 47, 118, 122 ff) indicates that not only the problem of balance, but the whole box-building business goes against the chimpanzee's grain.
To p. 585. The importance of visual clues varies of course with different species. Adams' cat immediately saw that the piece of liver can be hauled up by the string to which it was attached. Köhler tried a similar experiment with his dog: a basket, conraining food, was suspended from a rope outside the barred window of the room so that the dog could easily have hoisted it up with her teeth or paws. But she 'did not even attempt this simple method of self-help, and paid no attention to the string which was lying just under her nose -- whilst at the same time she showed the liveIiest interest in the distant basket. Dogs, and probably, for instance, horses as well . . . might easily starve to death in these circumstances which offer hardly any difficulty to human beings -- or to chimpanzees' (Köhler, 1957, p. 31). Far be it from me to suggest that cats are more intelligent than dogs; what the experiment shows is that cats are phylogenetically more ripe for this type of perceptual and manipulative skill than dogs.
To p. 589. Hebb's Organization of Behaviour was published in the same year as Insight and Outlook; his A Textbook of Psychology in 1958.
XIV
LEARNING TO SPEAK
Intending and Saying
Preparing to say something, whether it is a single sentence or a public lecture, is to set a hierarchy in motion.
' . . . And has the reader never asked himself', William James wrote in 1890, 'what kind of a mental fact is his intention of saying a thing before he has said it? It is an entirely definite intention, distinct from all other intentions, an absolutely distinct state of consciousness, therefore; and yet how much of it consists of definite sensorial images, either of words or of things? Hardly anything! Linger, and the words and things come into the mind; the anticipatory intention, the divination is there no more . . . [The intention] has therefore a nature of its own of the most positive sort, and yet what can we say about it without using words that belong to the later mental facts that replace it? The intention to say so and so is the only name it can receive. One may admit that a good third of our psychic life consists in these rapid premonitory perspective views of schemes of thought not yet articulate.' [1]
In other words, before the verbal hierarchy is set into motion, there is an ideational process of a highly conscious character, an intention or active expectation, which itself is not yet verbalized. Consider what is involved in preparing an ex-tempore lecture. The first step is to jot down the principal arguments or themes in key-words -- in 'symbols of the second remove' so to speak. Each theme is then treated as a sub-whole, a flexible matrix of ideas with an invariant code: the logic of the argument to be conveyed. But the ways of putting it across are many: factually, whimsically, by concrete examples. My strategical choice is governed by the lie of the land: the character of the audience; and by feedback from implicit tries: their anticipated reactions. If I have decided on concrete examples, I must search for them in my memory, and then again make strategical choices. The next question is where to start, to decid
e on the sequential order, and the approximate time allotted to each of the various subjects so as to make a balanced whole. In this quasi-architectural planning, the arguments are still treated as sub-wholes or building blocks, whose 'contents are known but are present without adequate verbal designation'. [2] All this is still a long way from the choice of actual words; in fact the hazy intentional situation described by James repeats itself on several levels.
But as we approach the actual formation of sentences, automatisms begin to intrude, indicating that we are nearing the bottom of the hierarchy -- the consummatory acts of language which terminate the appetive behaviour of thought. The sub-codes of grammar and syntax, which now enter into action, are almost wholly automatized; and when we finally arrive at the formation or syntactic units -- individual words patterned into phrases -- there is a good chance that these will be 'fixed action-patterns' -- verbal formulae, clichés, mannerisms, stereotyped turns of phrase, which remind one of the fighting rituals of the stickleback (e.g. 'the evidence tends to show', 'as we have seen before', 'we must bear in mind, however', etc.). Technical papers and bureaucratic utterances are conspicuous by their narrowness of vocabulary and rigidity of phrasing. Fortunately, there exist non-abstractive hierarchies whose criteria of relevance are aesthetic or emotional, which co-determine the tactics of verbal choices and counteract the tendency towards automatization.
Having gone through all these implicit motions, we end up by spelling out the actual sounds or ink marks, vocal patterns or typewritten letters. Yet even on this automated level, hierarchic organization prevails; the phonetic sequences or manual patterns are triggered off as wholes and perceived as wholes; nowhere, in the course of our descent through the hierarchy, do we strike rock-bottom, made of hard 'atoms of behaviour'.
Thus 'verbal behaviour' is initiated by unverbalized intentions at the top of the language hierarchy, and terminates in deverbalized sensory-motor activities; at each level it is governed by rules which elude verbal definition, and modified by extraneous factors acting on the plastic matrices of language. Each time we slice behaviour 'vertically' -- instead of horizontally on a single level -- we arrive at a series which at the top recedes into an elusive blur and at its base vanishes in the twilight of awareness.
The above description may have seemed unnecessarily verbose, since the hierarchic structure of language is so obvious that it need not be stressed. After all, nearly half a century has passed since Watson proclaimed that speech consists in the manipulation of words, and thinking in the sub-vocal manipulation of words. However, if the reader thinks that the cruder forms of Paleo-Behaviourism are a matter of the past, he should turn to the first Note* at the end of this chapter (which will also provide some light relief). It is an excerpt from a textbook for College students, published in 1961, which starts with the sentence: 'For many of you, this is your first encounter with psychology as a science.' One wonders what this first encounter will do to the mind of the trusting student who is here quite literally led to believe that human discourse consists in the chaining of S.-R. units which are best studied by bar-pressing experiments 'under the more ideal conditions in the laboratory'.
The Dawn of Symbol Consciousness
The previous section referred to speech as a performance. Learning to speak proceeds more or less in the opposite direction -- from the bottom to the top of the hierarchy. But only 'more or less', because bit-learning plays here a much lesser part than in the acquisition of mechanical skills like typing. Let us take a closer look at a few characteristic aspects of the process.
The first vocal ventures of the child confirm the now-familiar principle (cf. Chapter II) that spontaneous motor activity precedes sensory control. The child, too, 'acts on the environment before it reacts to it' -- as parents must learn in sleepless nights; cooing and babbling are spontaneous expressions of joie de vivre, in which a surplus of energy is discharged, as in the waving of arms and wriggling of toes.
At the early stages these spontaneous babbling sounds are the same whether they are produced by an American white or a Negro baby (i.e. the frequency spectra of the phonemes are practically identical). [3] At five to six months, however, when syllabic, speech-like sounds become increasingly frequent, the spectrum shifts noticeably towards the sound patterns produced by the adults in the child's environment; and at twenty-four months the resemblance is close. Thus the originally undistinguishable phonetic matrices of infants from different language-groups become differentiated by the imitative repetition of adult sound-patterns. Imitation is at first probably automatic, the auditory-vocal feedback apparatus being excited by the input as in the young singing bird; then shades into a more or less conscious response -- from 'echolalia' to 'metalalia'.
The first correlations between a sound pattern and an object or person are probably 'stamped in' by the parents at an age when the child's nervous system is not yet mature for them. In Book One (pp. 220-3) I have mentioned two examples at opposite extremes: Watson conditioning an infant less than six months old -- i.e. much too early -- to say da each time it was given the bottle; and Helen Keller's sudden and dramatic discovery, at the age of seven -- i.e. when she was overripe for it -- that 'everything had a name and each name gave birth to a new thought'. In the average, normal child the dawn of symbol consciousness seems to be a gradual, cumulative process. From approximately the eighteenth month onward there is a sharp increase in the child's vocabulary (Book One, p. 221) and somewhere around that time it makes that 'most important discovery of its life', that verbal labels can be attached not only to particular things and events, but that everything under the sun has a name. The universe of words, and the universe of things, have become integrated and will remain inseparable. Henceforth, every word must mean something and everything must have a name.
With the emergence of language, we have attained a new level of the cognitive hierarchy, which represents a sharp break in the continuity of learning processes in animals and man. Homologue laws still operate: 'All things have names' may be regarded as yet another case of empirical induction which we have seen operating on all levels; and verbal symbolism may be regarded as an extension of sign situations. Dolphins can be taught to respond to verbal or visual commands; dogs and chimps can make their wishes be known to their masters; cats will learn that bolts and loops mean escape; bees can communicate their experiences by dancing. But such communication by signs extends only to a few particular situations; it is generalized to some extent in the laboratory situation, when it seems to dawn on the cat that 'all these contraptions are means of escape', as it dawns on the child that 'all these words are means of getting one's way'. But at this point the cat has reached the limit of its abstractive capabilities, whereas the child has only got to its first inkling of what words can do for you.*
Their first, obvious advantages to the child are that they can be utilized both as labels, and as levers which make things happen, that they serve both the progressive socialization and internalization of behaviour -- communication and inner discourse. This is well-covered ground which needs no further labouring. But another phenomenon of early language behaviour is rarely emphasized: the appearance of the naming question or 'naming mania'. The child points at anything it sees, asks 'This -- ?' or 'What that?', and is visibly satisfied to learn the answer, without any utilitarian reward. Alternatively, it points at things, calls out their name or, if it has forgotten it, invents a new one. Here is a true paradigm of latent learning, of the exploratory drive, or of l'art pour l'art behaviour -- whatever one wants to call it. To quote Piaget: 'It is . . . no exaggeration to say that sensory-motor intelligence is limited to desiring success or practical adaptation, whereas the function of verbal or conceptual thought is to know and state truth.' [4]
But there is another aspect to the 'name-expectation': henceforth the child's concept of a thing or event will be experienced by it as incomplete if there is no verbal label attached to it. The concept will behave like a molecule with a free valency, a
s it were. We can perhaps recapture an echo of this when we hunt for the forgotten name of a person. Also, when we learn a new language, we suddenly have a whole class of free valencies: we feel actually frustrated when we discover that the French have no word for 'snobbery' or 'understatement', and are tempted to exclaim like a child: 'But surely they must have a name for it?'
The behaviour of the rat exploring a maze or of the monkey fearfully peeping into the box with the snake, could be described as interrogatory, and the same description could be applied to the puzzled expressions and actions of a small child before it has learned to speak. But its first explicit, verbal questions refer to the names of things, prompted by the need to 'saturate' the free valencies in its pre-verbal object-concepts. Is it too speculative to assume that this origin of the questioning habit must influence the whole later development of thought? The fact that each naming question can be readily answered by adults, may implant the implicit belief into the child that all questions are both meaningful and answerable; that the nature of explanation is based on the same kind of simple and direct connections as that between 'thing' and 'name'. This implicit belief seems to have been one of the factors responsible for the aberrations of human thought.
Concepts and Labels
I have talked of (pre-verbal) 'concepts' to which the verbal label becomes attached. The Concise Oxford Dictionary defines a concept as an 'idea of a class of objects; general notion'; Webster as 'a mental image of a thing formed by generalization from particulars'; Hilgard: 'When a symbol stands for a class of objects or events with common properties, we say that it refers to a concept' [5]; Piaget [6] regards the development of 'object concepts' as one of the principal achievements of the 'sensory-motor intelligence' which precedes the rise of verbal intelligence in the child's first two years. He distinguishes six stages in the attainment of stable object-concepts, which he defines as 'a system of perceptual images endowed with a constant spatial form throughout its sequential displacements and constituting an item which can be isolated in the causal series unfolding in time.'[7] It is interesting to compare Piaget's rather abstract formulations with the lurid 'incisiveness' of Wilder Penfield -- a neuro-surgeon. He too starts by securing his flanks with quotes from Webster and Oxford; then he goes on:
The Act of Creation Page 71