From “The Nature of Knowledge” by Henry Plotkin
LANGUAGE
A question that we all consider at some time is, what is it that makes us different from other animals? The use of tools, the realization that others have minds, the quality of mercy and compassion in our treatment of fellow human beings, and culture have all been argued for as what makes us human; and all have been observed or attributed in some usually partial and uncertain way to other species. We do, though, seem to have one ability which is truly unique to ourselves: language.
Certain other species of animal can transmit complex information with considerable precision, the most famous example of which is the dance of bees on returning to the hive, which transmits to other bees the position and richness of food sources or a promising new hive site. And it is now known that some other species have a limited ability to use specific reference in their communication; that is, they have signals that correspond to specific objects in their environments. For example, a common African monkey, the vervet, has different signals for flying predators, ground predators and snakes; and other vervets hearing these signals react in ways appropriate to the danger being signaled.
Impressive as such feats are, they are not language. Language is the ability to use a relatively limited - albeit quite large - set of symbols to generate a virtually infinite number of meaningful combinations to form utterances, each of which has meaning. I can, for example, say, 'I am going to France', and 'It is to France that I am going', and 'I am on my way to France', and 'France here I come', and many others besides, all of which have the same meaning. Each of us has a vocabulary of around 30,000 to 40,000 words, of which the most commonly used number just a few thousand. Yet we are able to generate some 1030 different meaningful sentences of about twenty words or less in length. Now, 1030 is an extremely large number - so large, in fact, that it dwarfs the 1010 seconds which is the approximate length of a normal human life span of threescore years and ten. If we never drew breath but just talked all our lives, we would utter but a tiny fraction of all the sentences we are capable of producing. Of course, many of these would be novel - they will never have been produced by us, or others, before. Indeed, if we look at the previous two or three sentences on this page, or at the sentences following, it is more than likely that I have never produced them in that exact form before in my life. It is this characteristic of human language, its extraordinary creativity that makes our linguistic ability different from anything that any other species of animal has.
There has been a small storm among psychologists and linguists going back about thirty years as to whether apes, especially chimpanzees, can be trained into language use. This work has its beginnings in older, and very curious, experiments in which infant chimpanzees were brought into human households and raised, effectively, as if they were human children, as one of the family. They spent the same amount of time with their human 'parents' as ordinary children would; they were fondled, played with and spoken to as one would fondle, play with and speak to human children. Under these circumstances the ability of such animals to develop any sort of spoken human language was virtually nil. They just did not learn to speak. Now, this may merely have been because chimpanzees lack the kind of vocal cord apparatus, and the neural control of such apparatus, that we humans have. Perhaps these chimpanzees indeed learned to use language, but could do so only in their heads, lacking the appropriate peripheral anatomy by which their language could be expressed vocally.
In other words, these early experiments were bad experiments, and the reason for their failure, perhaps, is a trivial one. This possibility led to the question being asked as to whether exploiting some peripheral apparatus which chimpanzees do have, their hands, might not lead to more interesting and revealing results. Chimpanzees are very good at using their hands. Might not sign language, of the kind used by deaf people in the United States in which signs stand for whole words, be the symbolic medium by which chimpanzees might come to learn and use language, just as deaf people do? A quarter of a century of experiments lead us to the answer 'no'. Such apes certainly do acquire something in the region of 100 to 500 hand signs (symbols), most of which are for objects in the world, a minority being action words (verbs) and a limited number of action modifiers (for example, words like 'more'). But no good evidence has ever been obtained that these animals have ever used these symbols in the structured, organized way (that is, syntactically) that characterizes human language; and no evidence at all exists for the creativity of symbol use, that is the sine qua non for accepting that it is language that one is observing in these animals. Whatever one might think of these experiments, they certainly do not demonstrate the existence of linguistic ability in these creatures.
Given that language is unique to our species, which must mean that some part of that portion of our genetic makeup that is unique to us as a species is a part-determiner of the human ability to function in the realm of language. In other words, we must be genetically predisposed to learn, think in and communicate by language. This uniqueness of language as a human form of know- ing about the world, as well as its apparent multiple roles in human psychology, means that an understanding of language is an essential part of a science of human knowledge. We come to know a language in the sense that language is a part of our environment, it is something 'out there', as important and prominent as any other set of features of the environment. Also, to some extent - and some people think it is to a very considerable extent - we think using language and hence manipulate our knowledge of the world through language. We also communicate our knowledge to one another primarily through language.
This nexus of roles for language means that the relationship between internal organization and external features of the world that characterizes all forms of knowledge takes, in the case of language, a dynamic, double form: a word, or sentence, as external event is matched to internal representations of the meaning of the utterance, and meaning itself may take more than one form; and we also have words, or combinations of words, in our heads which match a very large number of features in the world. It is, of course, this symbolic quality of language that makes it so central to our psychology: a dog and the word 'dog' are both 'things' in the world; and our internal representations also have this dual quality of some kind of image, primarily visual perhaps, of the dog, the thing itself, and the brain state that corresponds to the word 'dog'.
On the other hand, we should be careful not to overstate the power of language in our knowledge of the world. We have already seen how limited we are in dealing with faces using language; and in the next section we will see that there is a general poverty of language when it comes to dealing with emotional states. And yet, language so dominates our lives, especially our interior mental lives, that it is difficult not to think of it as anything other than the central plank of human understanding and knowledge. So, returning to the main thread of the argument, from the simple fact of the uniqueness of language to our species, the case can be made, though not an absolutely watertight case, that we do not come to know language itself, or to know the world through language, in the same way and via the same mechanisms that members of other species come to know, say, the spatial relation- ships of their worlds or how they come to associate temporally related events. Knowledge of, and by way of, language cannot be accounted for by some generalist, non-species-specific mechanism. But can the argument be made that knowledge of a language is also not attributable to some species-specific yet generalist ability
of the human mind? The answer is yes. There is strong evidence for language being understood as a genetically determined and domain-specific organ of mind that goes beyond its merely being unique to humans and the difficulties that we have in describing faces using language or giving adequate verbal expression to our emotions. In other words, once again the evidence seems to stack up against a tabula rasa approach to human knowledge.
Consider, for example, some further remarkable figures about language learning and language use. At one year of age all normal human infants, irrespective of which of the 5,000 -plus languages of the world they are being raised to speak, have a vocabulary of about ten words. Four years later the average vocabulary of five- year-olds is somewhere between 5,000 and 10,000 words. Such an increase requires the learning of about five new words each and every day throughout that period. This is an extraordinary achievement, because it occurs in such an unguided, untutored, unintended way. Adults and older children do not drill younger children into learning words and their meanings. They learn by a process that seems, by its ease of occurrence, analogous to diffusion - words and meanings seem to seep from the linguistically rich environment surrounding the child into the brain of the child. It is, as has been remarked by others before, as if children cannot stop themselves learning a language. It takes very little to stop children learning to read, write or ride a bicycle. But language seems to have the magical quality of largely growing from within the child - not completely, though, for those rare instances of children raised in environments totally devoid of language show us that such children do not have language. But given some minimal linguistic component in a child's world, in a relatively short time that child will come to a mastery of language.
Mastery of language does not just mean having the phonetic know-how for pronouncing words properly and a vocabulary to go with it. It also means speaking with the correct syntax - that is, the rules that govern word order, and the wider grammatical rules governing forms such as word inflections (these are changes in word endings in accordance with tense or to indicate plurals). For example, by five years of age every English-speaking child knows how to say, 'Give the dog a bone.' No normal child would ever produce sentences like 'The bone dog a give' or 'Bone a the dog give'. And all children of that age know how to transform a declarative, such as 'John is at home', into the interrogative form 'Is John at home?' Later transformational tricks will include such linguistic wizardry as effortlessly turning active constructions ('John gave the dog a bone') into passive ones ('The dog was given a bone by John').
What is extraordinary in all such examples is the contrast between the casual immersion into a language environment which the children experience and the highly specific and intricate language structures that emerge. The parents and caretakers of children are usually quite permissive in what they allow young children to say. They may occasionally correct gross errors, such as an incorrectly generalized past-tense rule like 'John gived the dog a bone', but they often let these pass, and it is not explicit lessons in grammar that result in the child becoming linguistically so capable. Specific, disciplined learning simply cannot explain the way in which language develops in the child. Once again, as in the case of bird song discussed briefly in the previous chapter, language learning in the human child is well described by the seeming paradox of the child knowing what it has to learn. All attempts at accounting for linguistic development in the child by using 'generalist' learning processes and mechanisms such as associative conditioning and trials-and-error learning are now wholly discredited. One can no more explain the exquisite structure of a five-year-old's language just by the environment in which it has been raised any more than one can explain the structure of that child's hand just by the meals that it has consumed or the objects that it has grasped. In both cases, language and hand, the structure comes in large part from genetic information that is nurtured and finds expression in a supportive, but not directly instructive, environment.
This contrast between an impoverished and unstructured environment, on the one hand, and the specific and intricate structure that develops in all children, regardless of differences in their upbringing, on the other, is what Noam Chomsky, the American linguist, called 'the argument from the poverty of the stimulus'. The argument is taken as very strong support for the assertion that language is caused by 'innate factors [that] permit the organism to transcend experience, reaching a high level of complexity that does not reflect the limited and degenerate environment', as Chomsky himself put it.
There is yet further evidence for the innate determination of linguistic learning in the stages of language development that occur in all children, again irrespective of the language that they come to speak in time and irrespective of the culture in which they are raised. All children begin to speak with one-word utterances at about ten or twelve months of age. These are made up predominantly of object words (such as 'ball' or 'daddy'), with lesser numbers of action words (for example, 'throw' or 'hug'). All children then produce two-word utterances at about twenty to twenty-four months. These two-word utterances are structured (in English, for instance, the infant might say 'throw ball', almost never 'ball throw'), without functors - that is, without prepositions (like 'on' or 'in'), conjunctions ('and', 'but', 'or'), articles ('a', 'the'), auxiliaries ('have', 'has') and copular verbs ('am', 'is', 'are', 'was') - and without inflections; functors always appear later. This order is almost never violated, and, again it is worth stressing, the universality of the pattern of development occurs irrespective of the child's specific experiences.
There are other universals of human language that are present, irrespective of the language spoken or the learning experience. One such is the oddity known as 'motherese'. 'Motherese' is a misnomer, being a form of speech adopted by all adults when addressing infants: the baby seems to 'release' a special form of slow, high-pitched talking with exaggerated intonation ('koochee- koochee-kooo' or its equivalent in Japanese, Swahili or any other language), a form of speech that adults never normally use when addressing other adults or even quite young children. There is also evidence that infants pay more attention to 'motherese' than to ordinary speech. Less of an oddity, and just as fascinating, are studies carried out on deaf children raised using hand signs rather than spoken language. Such children pass through exactly the same developmental stages as speaking children, and the language that emerges has the same characteristics of grammatical structuring. So language is not confined to vocal speech and the auditory channel. It is a ‘knowing’ part of the human mind that normally attaches itself to tongue and ear, but functions almost as well when operating through the hands.
Finally, linguists have for decades been trying to uncover the universal transformational rules that perhaps govern the structures of all languages, and through which all languages can eventually be described. One rule that does seem to be universal is used in the transformation of a sentence from the declarative to the interrogative form. For example, 'The tiger is alive' becomes 'Is the tiger alive?' The rule is that without regard to any specific language it always involves a syntactic restructuring of transposition with the verb following the noun phrase. So 'The tiger that is stripy is alive' becomes 'Is the tiger that is stripy alive?', and never 'Is the tiger that stripy is alive?' or 'Is stripy is alive the tiger that?'
Since the late 1950S, the work of Noam Chomsky and others has completely changed the way that psychologists understand human language. It has come to epitomize the general approach that human knowledge is innately determined and domain-specific. This Chomskian understanding of language fits well with the picture of the evolution of intelligence developed in Chapter 5. Homo sapiens has a spectacularly evolved capacity for learning and remembering an extraordinary amount of linguistic material, and of creatively generating a prodigious quantity of linguistic output. This is a tool both for dealing with rapid change by thought and communication and for generating change in the form of a rapid flux of ideas and conceptions. What I cannot yet understand, and neither can anyone else, is exactly what the functional origins of language are. If, as argued in the previous chapter, instincts are the mother of intelligence, we are not yet able to understand the nature of that originating instinct. Perhaps it arose as part of the social function of intellect a la Humphrey and Cosmides, with language supplying symbolic form to more concrete types of social exchange and social contract, the symbolism giving rise to possibilities of greater complexity and subtlety of interaction. Or perhaps it arose in response to, or hand in hand with, the requirements of the evolution of human culture, a tertiary heuristic that can operate at even higher-frequency levels of change than the secondary heuristic of individual intelligence, and hence which requires a medium of information-transmission that can match those high frequencies of change.
*AJ, Thanx for the article but I still disagree for the reasons I mentioned. Language is not innate, inherent, genetically human, rather, like consciousness, it is a by-product of complexity - and an *inferior* by-product if there ever was one, I might add. Do ponder on the futility of understanding the expression "I love you." Is there an inherent bucket of love which, on its own, this expression belongs to? To me, it seems safe to say that the human race is a mess because development of this 'inherent' trait called language-ability hasn't kept pace with the speed of evolution. So we must agree to disagree with your proposition. |
