Pinker (1994:334) points out that:
nonhuman communication systems are based on one of three designs: a finite repertory of calls (one for warnings of predators, one for claims to territory, and so on), a continuous analog signal that registers the magnitude of some state (the liverlier the dance of the bee, the richer the food source that it is telling its hivemates about), or a series of random variations on a theme (a birdsong repeated with a new twist each time: Charlie Parker with feathers)
On the other hand, Hockett (1960, as cited in Harley, 1995:9) listed 16 universal characteristics of human language, which provide a useful framework for thinking about how animal communication systems differ form human language. Three characteristics are picked out to discuss in details for making a clearer distinction between human language and animal communication.
The particular words used to describe entities, actions, and attributes in any language are arbitrary (Gleason & Ratner, 1998). The words of a language are symbols, which are abstract and do not look like what it stands for, but are agreed to abide by any individual within the community. Take “books” for example, although the word has nothing similar to the entity books, like shape, colour and so forth (in fact, “shape” and “colour” are also arbitrary words used by people), every one agrees with its meaning. However, animals may have their arbitrary symbols in their groups, but they lack a sufficiently rich system of symbols and rules for their combination to allow transmission of a large variety of concepts (Gleason & Ratner, 1998).
Languages do not vary infinitely in that there are constraints on the nature of possible linguistic rules that reflect the nature of human cognition (Chomsky, 1981, as cited in Gleason & Ratner, 1988). Gleason and Ratner (1998) propose that language is a rule-governed system of behaviour and the rules are quite arbitrary in nature as words. English, like other language, has conventions for putting and ordering specific words in a sentence with no real reasons. Sentences, for instance, “I like studying English”, are hierarchical structured, with arbitrary words combined by arbitrary grammar. In brief, both vocabulary and grammar of any language represent arbitrary conventions that the users of a language agree to abide by (Gleason & Ratner, 1988). In contrast, it is difficult to analyze the “substructure” of infant or animal cries, not to mention syntactic constructions. Although some substructure may appear to exist in the bee’s dance and the bird’s replication of human speech, such messages lack the infinite productivity of human language (Gleason & Ratner, 1998).
Human language is so much more creative and flexible than the communication system of any other species (Cairns and Cairns, 1976:5). For example, “Internet” is a word created with the development of technology by people. It is true that some of the signing apes (Terrace, 1979; Rumbaugh, 1977; as cited in Gleason & Ratner, 1998) produce brief utterances that relate to their current intents (for example, “Tickle Nim”), but, as Demers (1989, as cited in Gleason & Ratner, 1998) notes, animal communication lack unbounded productivity of new and varying messages compared with human language.
Marshall (1970, as cited in Harley, 1995) points out that language is under our voluntary control, intentionally expressing the full range of speakers’ experiences, even imaginary ones (in remote time and space). For instance, when both speaker and hearer (or writer and reader) share the same rule system, message transmission can be not only creative but also unambiguous. It is very different from animal communication which is typically context or stimulus dependent; vocalizations are likely to occur under narrowly specified conditions (Gleason & Ratner, 1998).
These characteristics of language give it many properties not shared by animal communication or infant cries, as Harley (1995) suggests that language is species-specific, a special faculty unique to humans.
To find out why language is species-specific, language evolution should be concerned with. According to Wikipedia (2010), the evolution of modern human language required both the development of the anatomical apparatus for speech and also neurological changes in the brain to support language itself. Scientists prove that the descent of larynx and lengthening of laryngopharynx enables human adults better to speak. In addition, the seat of human language in the brain is special as well. Pinker (1994:334) proposes that genuine language is seated in the cerebral cortex, primarily the left perisylvian region to support the complex language, while the vocal calls of primates are controlled by phylogenetically older neural structures in the brain stem and limbic system, structures that are heavily involved in emotion. Cairns and Cairns (1976) conclude that humans have a biological basis for speaking language in general.
Furthermore, Noam Chomsky argued that children must innately be equipped with a plan common to the grammars of all languages, a Universal Grammar, that tells them how to distill the syntactic patterns out of the speech of their parents (Pinker, 1994:22). Endorsing the Universal Grammar, Pinker (1994) concludes that language is an instinct, conveying the idea that people know how to talk in more or less the sense that spiders know how to spin webs.
However, the conclusion drawn by Pinker based on Universal Grammar is proved controversial. Tomasello (1995) argued that: in the common understanding of both scientists and laypersons alike an instinct is a behavioural competency, or set of behavioural competencies, that: (a) is relatively stereotyped in its behavioural expression, and (b) would appear in ontogeny even if an individual were raised in isolation from its species-typical set of experiences (Birney & Teevan, 1961, as cited in Tomasello).
But language does not fit the criteria. Firstly, there are a large variety of languages existing in the human species that are fundamentally different from one another, including vocabulary, syntactic construction (grammar), even individual speech personalities. Language is not stereotyped in its behavioural expression, as it varies from one region to another. For example, English is very different from Chinese, from words to grammars. Secondly, an individual human being can acquire any other kinds of language in the condition of several years of linguistic experience with other language speakers. Thirdly, according to feralchildren.com (2010), most of the feral children who lived away from human contact from a very young age cannot speak human language at all when they are found. If language is comprehended as an instinct as spiders’ ability to know how to spin webs, why the feral children cannot speak a word when people found them while spider can spin webs without any instruction from other spiders?
Pinker (1994:18) proposed that
language which develops in the child spontaneously, without conscious effort or formal instruction, is deployed without awareness of its underlying logic, is qualitatively the same in every individual, and is distinct from more general abilities to process information or behave intelligently.
To make clear whether children acquire language spontaneously, the language acquisition process would be discussed further.
According to Taylor (1976:237), there are two opposing theories of the acquisition of language. The rationalistic approach says that there are linguistic universals and children are born with them, which is compatible with Pinker. The empiricist’s approach says that language is learned by establishing stimulus-response connections through conditioning.
Eisenberg (1965, as cited in Taylor, 1976:184) found that babies respond differently to sounds at the speech-sound frequencies than to those at higher frequencies. They respond differently also to patterned, as compared with constant, acoustic stimuli. It seems to agree with the rationalistic approach that children are born to know language. However, Macnamara (1972, as cited in Paivio & Begg, 1981), as well as Piaget, proposed that children develop nonlinguistic cognitive processes before they learn their linguistic signals. He assumed further that these processes are primarily learned: the child does not have a complete set of cognitive structures when he or she begins to learn language, but their developmental onset precedes that of language. It can be a reason accounting for antenatal training. For example, it is generally believed that the music antenatal training plays a very important role in the embryo’s growth. Hearing capacity and the comprehension of music are stimulated right away in the mother’s abdomen. This case illustrates that babies can access to the music in prenatal period, so the likelihood of accessing to human language is drastically high. Antenatal training can be viewed as a developing nonlinguistic cognitive process before they learn their linguistic signals.
Sinclair-deZwart (1973, as cited in Paivio & Begg, 1981) expands on Piaget’s view of cognitive development, which divides cognitive development into several stages. The sequence is roughly as follows:
1. The sensory-motor stage (birth to 2 years), during which the child develops action patterns by acting on the environment. The child starts noticing the existence of objects and knowing that objects are separate from the self, and builds a personal world of permanent objects (Piaget, 1952, as cited in Taylor, 1976). Apparently, feral children also have the capacity of sensory-motor manipulation of the objects, since they share the same biological make-up with normal human children.
2. The next period, representational intelligence, begins toward the end of the second year (the end of the sensory-motor period), after the action structures have been internalized. The child differentiates between an object and its internal representation, a symbol. Symbolic representation has several forms, including symbolic play, imitation, mental images, and drawing. For example, a child will not eat the pebble which is used as a “pretend” candy in a game, because he realizes the symbolic nature of the pebble. However, animal can only tell the object superficially by its shape, colour, smell or other physical characteristics. Feral children also have symbolic function as well, since Piaget (1952, as cited in Taylor, 1976) proposed that the symbolic function originates independently of language.
3. Language is an extension of the representational level after symbols become socialized. Linguistic structures build on the general cognitive structures established during the first two years. Unlike the child’s private signs, words are social signs, and their relationship to significates (objects) is arbitrary and conventional (Piaget, 1952, as cited in Taylor, 1976). Therefore, words are sharply differentiated from the private signs of the child. For instance, a child may call the cat as “mew” at first. It is “imitation”, a sort of symbolic representation proposed by Sinclair-deZwart (1973, as cited in Paivio & Begg, 1981). It shows that language is assimilated to the child’s private symbolism (Piaget, 1952, as cited in Taylor, 1976). When its initiative name “mew” meets with conventional name “cat” that used by all the people around, the child’s symbolic function will be slowly weaned from its initial egocentricity, so the child’s use of language becomes increasingly socialized (Piaget, 1952, as cited in Taylor, 1976). Take feral children into consideration, they have symbolic functions as normal children that is stated above. However, if the feral child is isolated from human society from birth or around 2 years old, the child’s private signs are impossible to be socialized since he or she does not have any contact with the human society and staying with animals who cannot speak human language. As Piaget (1952, as cited in Taylor, 1976) suggests, the use of language in social contexts plays an important part in the child’s weaning from its initial egocentricity. Therefore some feral children having minimal or even no experience been exposed to human society will stay in the stage of symbolic thought in terms of language acquisition.
Being restricted in the stage of symbolic thought, shown by many actual cases according to feralchildren.com (2010), feral children are not able to use human language, since language is a socialized signs that require the use in the social context, human society. Does it mean that feral children are transformed to a wild animal, like wolf, which cannot speak or learn human language any longer?
Reported by feralchildren.com (2010), the feral children found by human are generally taught human language afterward,. Thanks to the human instinct biological basis for language, human language can be taught to feral children, although most of them can only learn some words rather than mastering the syntax like normal children. Those feral children who were captured in their late childhood are proved to lean a modest amount of language with special training (Taylor, 1976). Fromkin and coworkers (1976, as cited in Taylor, 1976) described a 15-year-old girl who was kept in diapers like a baby until she was found at age 11, and could utter only grunts. In a few years she advanced to the speaking ability of a child of 2-2years under a university-backed linguistic program. Although the progress she made is modest, but she did learn to speak, and surpass the learning capacity of any chimpanzees trained by human. Taylor (1976) argued that the slow process of learning language is due to the existence of a critical period for acquiring language. According to Taylor (1976), the concept of a critical period implies that language may be acquired readily during this period but not so readily later. This period may last until the early teens, when the brain completes its functional organization (Taylor, 1976). The feral children cases show that certain capacities are genetically given, so that language is species-specific (Hebb, Lambert & Tucker, 1971, as cited in Paivio & Begg, 1981). However, what can be deduced from feral children’s failure to use language when isolated from human society?
Hebb, Lambett, and Tucker (1971, as cited in Paivio & Begg, 1981) opposed Chomsky’s and McNeill’s nativistic viewpoint with an empiricist position based on perceptual learning. Hebb and its colleagues (1971, as cited in Paivio & Begg, 1981) argued that
language is species-specific and is therefore influenced by hereditary mechanisms peculiar to humans, such as (a) auditory analysis and (b) dealing simultaneously with verbal and nonverbal representational processes (ideas, mediators).
These hereditary factors are innate capacity shared by feral children as well. In addition to the hereditary endowment, Hebb and his colleagues argued that learning is involved in children’s speech development, because the child is born into a language-filled environment with both verbal and nonlinguistic uniformities (Paivio & Begg, 1981). Therefore, Paivio and Begg (1981) conclude that experience plays an important role in language development. Without being exposed to a language-filled environment, the access to experiencing human language is interrupted, and that is why most of the feral children could not speak a word when people found them.
To sum up, language is species-specific, only possessed by human beings, due to the result of language evolution--the development of the anatomical apparatus (the descent of larynx, lengthening of laryngopharynx and special seat of human language in the brain) for speech and also neurological changes in the brain to support language itself. These are all biological make-up innately owned by human beings for speaking language, but it does not mean that language is a Universal Grammar which is innately equipped in the human’s brain. Feral children’s inability to speak language is a counter example of “the language is an instinct” which does not need to learn. Rather, language should be learned in a social human context as experience plays an important role in language development.