The Uniqueness of Human Language

Previously published in the Midweek Review, The Island of Wednesday, 8^th April 2009

Although living in social groups is not unique to humans, among the myriad forms of animal life on earth only the human species can be said to have undergone the evolutionary process known as civilization. The term civilization refers to the general social development of the human race over the millennia with steadily growing levels of knowledge, skills, mastery of natural forces through science and technology, commerce, government, art, religion, etc. In the final analysis, this never-ending activity is based on the creation, dissemination, refinement, preservation, and transmission down the ages, of ideas. But for language none of these would have been achieved by the humankind. And language itself is a concomitant of the exceptionally advanced brain power of the humans.

What language (especially human language) itself is strikes me as a daunting thing to form an idea of, and I, with my essentially limited knowledge, will not presume to attempt a definition of it in this essay or most likely anywhere else. However, there is hardly any difficulty in identifying human language when we come across an example of it. It is a thing we cannot afford to ignore; it is as common and as precious (nevertheless) in the human intellectual domain as the air we breathe in our physical environment. I will be content here to point to certain characteristics of human language in support of the thesis of my essay – the species-specificity of human language (i.e. language being a distinguishing characteristic of the human species).

Extensive research done over the past eighty years or so into the capacity of animals to learn human language seems to have drawn a blank except for the demonstration of some intelligent apes’ potential to deal with the “barest rudiments of language”. (Noam Chomsky denied even this rudimentary linguistic capacity for animals.)

It was in the 1930’s that two scientists, Luella and Winthrop Kellog, raised an infant chimpanzee called Gua together with their baby son. It was reported that Gua could understand about a hundred words, but did not articulate any of them. Another chimpanzee, Viki, in the 1940’s, brought up under similar conditions by another scientist couple (Catherine and Keith Hayes), after five years of training, was eventually able to produce rather poorly articulated versions of ‘mama’, ‘papa’, and ‘cup’. This, actually, is a noteworthy achievement in view of the fact that non-human primates’ vocal tracts are not physically structured for producing human speech sounds.

In 1966 Beatrix and Allen Gardner, recognizing that chimpanzees are not physically equipped for producing human speech, started teaching a chimpanzee which they named Washoe a version of American Sign Language (ASL). ASL has all the distinctive features of human language. Three and a half years of training gave Washoe the ability to use signs for more than a hundred words including ‘airplane’, ‘baby’, ‘window’ and you’. She was even able to combine some of these to produce sentences of the kind ‘more fruit’, ‘open fruit drink’, which demonstrated some inchoate linguistic productivity. Some of these forms appeared to be her own inventions, eg., ‘water bird’ for ‘swan’. It was seen that she could understand many more signs than she could produce. Washoe was said to be capable of having rudimentary conversations mainly of the question-answer type.

About the same time, Ann and David Premack experimented with a chimpanzee called Sarah. She was taught to manipulate a set of shapes for communicating with humans. The shapes represented words. These could be arranged into ‘sentences’. Sarah was trained to link these shapes to objects or actions. After learning to use a large number of these plastic shapes, she was able to select a blue triangle to get an apple. There is no natural connection between a blue plastic triangle and an apple. Therefore this symbol was arbitrary. Sarah was able to make such sentences as ‘Mary give chocolate Sarah’.

In 1973, Duane and Sue Rumbaugh subjected three chimpanzees (Lana, Sherman, and Austin) to a training technique with a similar artificial language, which they called Yerkish. A set of symbols on a large keyboard linked to a computer was available for Lana and others to manipulate. For example, they had to press four symbols in the correct order to produce the message ‘Please machine give water’.

However, many greeted these claims with skepticism. They argued that Lana and the other chimpanzees used the word ‘please’ without understanding its meaning. The chimpanzees did not know that they could produce a meaningful utterance by omitting ‘please’. The sign for ‘please’ was merely the equivalent of a button on a vending machine; the ability to use it did not need any language knowledge.

It was the psychologist Herbert Terrace who produced the most convincing arguments against claims of chimpanzees’ linguistic achievements. He used a chimpanzee called Nim Chimpsky (a deliberate play on the name of Noam Chomsky, who had asserted that language is an innate ability unique to the human species). This experiment began in 1973.

Nim was taught American Sign Language under controlled conditions. Nim’s classroom activities were carefully recorded and videotaped. He, like Washoe before him, seemed to be developing a linguistic ability much like human children. However, close inspection of the videotapes showed that this impression was wrong: Nim was not actually capable of synthesizing more complex structures out of simpler ones in the way that human children do besides, Nim responded to his teachers’ signing, but rarely initiated any signing by himself. This observation led Terrace to reinvestigate the video records of Washoe’s use of sign language. He argued that both Nim and Washoe only appeared to use signs as language. Terrace concluded that these chimpanzees were clever animals which produced a certain kind of behaviour (signing) in order to be rewarded; their signing was not linguistic at all.

It may appear that animal communication and human communication have certain features in common. Both animals and humans use ‘signs’, either visual or vocal to communicate. For example, bees are known to pass on information about available sources of nectar to their colleagues through a kind of dance, while we choose to convey some information solely through gestures in certain situations without using language at all. However, since language is basically dependent on the manipulation of vocal symbols, it would be relevant here to consider how linguistically close animal communication could be to human communication in terms of ‘articulated’ sounds.

One feature common for both animals and humans is the use of the vocal- auditory channel. Human language is typically generated through the vocal organs and perceived through the ears , but it can also be transmitted without sound, through writing or visual signs. Many other species (eg. monkeys, elephants, dolphins, birds)use the vocal-auditory channel. Another common element is reciprocity (i.e. the sender of a signal can also be the receiver of a similar signal). A third shared feature is specialization, which means that linguistic signals do not serve any other purpose such as breathing. A fourth is that for both humans and animals vocal signals are non-directional (linguistic signals can be picked up by any individual within hearing range). Finally, rapid fade characterizes vocal signals, whether animal or human, which means that these disappear as soon as they are produced. Thus, in certain merely ‘mechanical’ ways animal vocal communication and human language show some close resemblance.

On the other hand, there are infinitely more significant features which are almost entirely unique to human language, while being only vestigially present in animal communication. These characteristics are connected with man’s uniquely advanced ability to think, which is impossible without language.

A human can talk not only about what is present to his or her senses at the time of speaking, but also about things which are yet to happen, or happened in the past, or are merely imagined. This property of human language is known as displacement. Displacement is generally absent in animal communication. Bees are able to indicate sources of food to other worker bees by means of a dance. This may suggest that bee communication has the quality of displacement, but the important thing is that this is extremely limited. Bees, for example, cannot refer to ‘the garden we visited yesterday’.

The second distinctive quality of human language is arbitrariness. That is, there is no natural connection between a linguistic form and its meaning. The English word ‘dog’ has no intrinsic relationship with the animal it denotes. The onomatopoeic theory of language origin may seem to contradict this: words such as boom, clash, crash, cuckoo, and slurp suggest by their sound the objects or activities they name. The truth is that such onomatopoeic words are actually not numerous enough to account for the infinitely vast and complex phenomenon of language.

In the case of animal communication there is a clear relationship between one sound and its intended message. Squirrels, for instance, have one call for mating, one for sharing food, one for warning of danger, etc., (and it is my personal experience that they have a special call to signal the presence of a snake in the vicinity). Arbitrariness, therefore, is not present in animal communication; but it is a distinctive quality of human speech.

The next characteristic feature of all human languages is that new sentences are made all the time. Children who are learning a language make sentences which they have never heard before; adults also do the same in dealing with novel ideas and new situations. This property of productivity is not shared by non-human signifying systems. Vervet monkeys are said to have thirty-six vocal calls, and cicadas only four. These sets of signals are fixed in range. Animals do not produce new signals or combinations of signals for new experiences or events. They have fixed reference (that is, each signal is specific to a particular object or occasion).

Cultural transmission is the way human language is passed on from one generation to the next. Although babies possess an innate capacity for language, they do not genetically inherit their parents’ language. What they acquire is the language that they are exposed to. On the other hand, a dog, for instance, need not necessarily listen to its parents’ barking before it can do the same. The signals used for animal communication are not usually culturally transmitted. In their case, transmission is biological.. Animals do not learn their communication systems; they acquire them instinctively.

Human linguistic sounds are discrete, i.e. they are meaningfully distinct. For example, the sounds represented by the letters f and v in English are not very different from each other, but the use of one rather than the other in a word is meaningful, eg. fine/vine, fat/vat. The calls of animals cannot be analysed into such discrete units that can recur in other combinations.

Yet another distinguishing property of human language is its duality, which means that it is organized at two levels simultaneously. At one level we can produce the individual sounds such as a, p, and t. They have no specific meaning. We can also produce these sounds in a particular combination (eg. apt, tap, pat), which is meaningful. This means that at one level of articulation we have distinct sounds, and at the other distinct meanings. This feature, known as duality or double articulation, makes it possible for us to produce a very large number of sound combinations using a limited set of distinct sounds. A dog’s ‘woof’ cannot be shown to be the combination of three distinct sounds w, oo, and f; neither can a dog be said to put these sounds in different patterns to convey different meanings. Animal communication, therefore, does not demonstrate duality.

The most distinctive feature of human language is that it is rule-governed (structured). When children learn a language, they invariably learn its grammar. The grammar of a language comprises its phonological, morphological, syntactic, and semantic rules, along with its vocabulary. The remarkable thing is that no one teaches them these; they just ‘pick them up’. For example, a child who is acquiring English as his or her mother tongue will learn the process of sentence construction, notions of tenses, forming questions, transformations, nominalizations and other complex manoeuvres without overt teaching. This crucial grammar-based nature of human language is not even minimally present in animal communication. No animal signifying system can achieve the infinite productivity and complexity of human language.

The human being is the most ‘dominant’ creature on earth and the only ‘civilized’ one. Among animals, the humans govern the earth and have already attained enough scientific knowledge and technological expertise to make their presence felt even beyond the solar system if there are any other similarly advanced intelligent beings out there in our celestial neighbourhood! All this has been possible thanks to the humans’ unique intellectual capacity, which goes hand in hand with language.

Advanced rational thinking is probably the most characteristic activity that sets the humans apart from other animals, though there are many instances of animal behaviour that demonstrate something like rational thought (My dog Ralph, a Doberman pinscher, leaves me in no about this). Human thinking is done through language, and would be impossible without it. Language is the means by which human beings assimilate existing knowledge, create new knowledge, disseminate it to their fellow humans, preserve it , and pass it on to posterity. Human beings experience their world, and live their lives through language to an extent inconceivable in the case of animals. True, animals have their own communication systems, but these are narrowly geared to their biological need of survival.

The universal ability that human babies are equipped with to acquire any particular language gradually in a relatively short period of time without being subjected to a course of instruction, but through mere exposure to it, and the extremely limited results of animal language research show that human language is a biologically determined innate faculty specific to the human species. So, neither the many instances of animal communication that we commonly observe nor the scientists’ attempts to train animals to speak or use language like human beings can be said to provide enough evidence for us to challenge the claim that language is a unique human possession.

Rohana R. Wasala