The Articulate Mammal (7 page)

These are the topics examined in this chapter. First, animal communication systems are compared with human language to see if animals can be said to ‘talk’ in any real sense. Second, various attempts to teach language to animals are considered. The overall purpose behind such inquiries is to find out whether humans alone have the power of speech. Are we biologically singled out as ‘articulate mammals’ or not?

Of course, if we discover that animals
do
talk, then we shall not have learned anything useful, just as the fact that we can do the breast stroke does not tell us anything about a frog’s innate swimming ability. Or, as three prominent psychologists acidly noted, ‘The fact that a dog can be trained to walk on its hind legs does not prejudice the claim that bipedal gait is genetically coded in humans. The fact that we can learn to whistle like a lark does not prejudice the species-specificity of birdsong’ (Fodor
et al.
1974: 451). If on the other hand, we find that animals do
not
talk, this will provide some support for the claim that language is restricted to the human race. We are not merely indulging in a neurotic desire to verify that humans are still superior to other species, as has sometimes been suggested. The purpose of this chapter is a more serious one. Some animals, such as dolphins and chimpanzees, have a high level of intelligence. If, in spite of this, we find that language is beyond their capability, then we may have found some indication that language is a genetically programmed activity which is largely separate from general intelligence.

DO ANIMALS TALK NATURALLY?

A first task is to find out whether any animals naturally have a true ‘language’. In order to answer this question, we must compare human language with animal communication. But such a comparison presents a number of perhaps unsolvable problems. Two in particular need to be discussed before we can give a coherent reply to the query, ‘Do animals talk naturally?’

The first problem is this: are we comparing systems which differ quantitatively or qualitatively? On the one hand, human language may have gradually evolved from a more primitive animal means of communication in a continuous line of growth – a viewpoint sometimes known as a ‘continuity’ theory. On the other hand, human language may be something quite different from our basic animal heritage, and superimposed on it. This is a ‘discontinuity’ theory.

Supporters of continuity theories suggest that language grew out of a primate call system, like the ones used by apes today. They assume that humans started out with a simple set of cries in which each one meant something different, such as, ‘Danger!’ or ‘Follow me!’ or ‘Don’t touch that female, she’s mine!’ These cries gradually became more elaborate, and eventually evolved into language. A possible intermediate stage is seen in the cries of the
vervet monkey. This monkey has several alarm calls which distinguish between different types of danger (Struhsaker 1967). The
chutter
announces that a puff adder or cobra is around. The
rraup
gives warning of an eagle. A
chirp
is used for lions and leopards. A less panic-stricken utterance, the
uh!
, signals the presence of a spotted hyena or Masai tribesman. According to some, it is a very short step from an alarm call warning of a poisonous snake to using the
chutter
as a ‘word’ symbolizing a poisonous snake.

Another interpretation of these signals is possible. The monkeys could merely be distinguishing between the
intensity
of different types of danger. They may be more frightened of puff adders than eagles – or vice versa. This plausible explanation has been ruled out by an experiment in which a concealed loudspeaker played recordings of the various alarm calls. When they heard a
chutter
, the vervets stood on their hind legs and looked around for a snake. At the sound of a
rraup
they dived into the vegetation as if hiding from an eagle. And at the lion–leopard
chirp
, they hastily climbed up a tree (Seyfarth
et al.
1980a, 1980b; Cheney and Seyfarth 1990). So the monkeys clearly have a special signal for each type of enemy.

Yet the danger cries of monkeys are still far from human language. They are a mix of a shriek of fear and a warning to others, and are only partly a symbol. The huge gulf between these calls and ‘real’ speech has led many people to argue for a discontinuity theory. Proponents of discontinuity theories claim that humans still retain their basic set of animal cries, which exist alongside language. Yelps of pain, shrieks of fear, and the different types of crying observed in babies may be closely related to the call systems of monkeys. If this view is correct then it is fairly difficult to compare human and animal means of communication. It may be like comparing two things as different as the Chinese language and a set of traffic lights. But a continuity versus discontinuity divide may be over-simple. Language is a complex mosaic in which some features are continuous, and some discontinuous with animal communication. Exactly which is which is still under discussion.

The second major problem we face is that it is not always easy to decide what counts as communication in animals. As one researcher notes:

Students of animal behaviour have often noted the extreme difficulty of restricting the notion of communication to anything less than every potential interaction between an organism and its environment.

(Marshall 1970: 231)

So that, at the very least, sticklebacks mating, cats spitting and rabbits thumping their back legs must be taken into consideration – and it isn’t at all clear where to stop. It is sometimes suggested that this problem could be solved by concentrating on examples where the animal is
intentionally
trying to convey
information. But such distinctions are difficult to draw, both in humans and animals. If a man smoothes down his hair when an attractive woman walks into the room, is this an unconscious response? Or is he doing it intentionally in the hope of catching her attention? In the sea, so-called ‘snapping shrimps’ can produce loud cracks by closing their claws sharply. Since the cracks can upset naval sonar devices, marine biologists have attempted to discover the circumstances which lead the shrimps to produce them. But no one has yet discovered the significance of the snaps. They may be informative – but they may not. There is no way in which we can be sure about making the right decision when it comes to interpreting such a phenomenon.

Having outlined these fundamental problems – which show that any conclusions we draw are only tentative – we can now return to our main theme: a comparison of human language and animal communication. How should we set about this?

A useful first step might be to attempt to define ‘language’. This is not as easy as it sounds. Many definitions found in elementary textbooks are too wide. For example: ‘A language is a system of arbitrary vocal symbols by means of which a social group cooperates’ (Bloch and Trager 1942: 5). This definition might equally well apply to a pack of wolves howling in chorus.

A superficially promising approach was that suggested by the linguist Charles Hockett in the 1960s. In a series of articles stretching over ten years he attempted to itemize the various ‘design features’ which characterize language. For example: ‘
Interchangeability
: Adult members of any speech community are interchangeably transmitters and receivers of linguistic signals’; ‘
Complete Feedback
: The transmitter of a linguistic signal himself receives the message’ (Hockett 1963: 9). Of course, such an approach is not perfect. A list of features may even be misleading, since it represents a random set of observations which do not cohere in any obvious way. To use this list to define language is like trying to define a man by noting that he has two arms, two legs, a head, a belly button, he bleeds if you scratch him and shrieks if you tread on his toe. Or, as a more recent researcher noted: ‘Any checklist … is almost certain to be superficial. Consider the definition of humans as “featherless bipeds”. Who would be content with that as an expression of the nature of humanness …?’ (Anderson 2004: 57). A major problem is that such a list does not indicate which features are the most important, or how they might be linked to one another. But in spite of this, a definition of language based on design features or ‘essential characteristics’ may be a useful first step.

But how many characteristics should be considered? Two? Ten? A hundred? The number of design features Hockett considered important changed over the years. His longest list contained sixteen (Hockett and Altmann 1968). Perhaps most people would consider that maybe eleven features capture the essential nature of language, not all of which are mentioned by Hockett.
These are:
use of the vocal–auditory channel, arbitrariness, semanticity, cultural transmission, spontaneous usage, turn-taking, duality, displacement, structure-dependence, creativity, ability to read intentions
. Some of these features are fairly general and occur widely in the animal world. Others are more specialized.

Let us discuss each of these features in turn, and see whether it is present in animal communication. If any animal naturally possesses
all
the design features of human language, then clearly that animal can talk.

The use of the
vocal–auditory channel
is perhaps the most obvious characteristic of language. Sounds are made with the vocal organs, and a hearing mechanism receives them – a phenomenon which is neither rare nor particularly surprising. The use of sound is widespread as a means of animal communication. One obvious advantage is that messages can be sent or received in the dark or in a dense forest. Not all sound signals are vocal–woodpeckers tap on wood, and rattlesnakes have a rattle apparatus on their tail. But vocal–auditory signals are common and are used by birds, cows, apes and foxes, to name just a few. The advantages of this method of producing the sound are that it leaves the body free to carry on other activities at the same time, and also requires relatively little physical energy. But this design feature is clearly neither unique to humans, nor all-important, since language can be transferred without loss to visual symbols (as in sign language, or writing) and to tactile symbols (as in Braille). Patients who have had their vocal cords removed, and communicate mainly by writing, have not lost their language ability. It follows that this characteristic is of little use in an attempt to distinguish animal from human communication. So let us proceed to the second feature, arbitrariness.

Arbitrariness
means that human languages use neutral symbols. There is no connection between the word DOG and the four-legged animal it symbolizes. It can equally be called UN CHIEN (French), EIN HUND (German), or CANIS (Latin). GÜL (Turkish) and RHODON (Greek) are equally satisfactory names for a ‘rose’. As Juliet famously noted:

What’s in a name? that which we call a rose

By any other name would smell as sweet.

(Shakespeare)

Onomatopoeic words such as CUCKOO, POP, BANG, SLURP and SQUISH are exceptions to this. But there are relatively few of these in any language. On the other hand, it is normal for animals to have a strong link between the message they are sending and the signal they use to convey it. A crab that wishes to convey extreme aggression will extend a large claw. A less angry crab will merely raise a leg: ‘Extending a major chaliped is more effective than raising a single ambulatory leg in causing the second crab to retreat or duck back
into its shell’ (Marshall 1970: 231). However, arbitrary symbols are not unique to humans. Gulls, for example, sometimes indicate aggression by turning away from their opponent and uprooting beakfuls of grass. So we conclude that arbitrariness cannot be regarded as a critical distinction between human and animal communication.

Semanticity
, the third suggested test for language ability, is the use of symbols to ‘mean’ or refer to objects and actions. To a human, a CHAIR ‘means’ a fourlegged contraption you can sit on. Humans can generalize by applying this name to all types of chairs, not just one in particular. Furthermore, semanticity applies to actions as well as objects. For example, to JUMP ‘means’ the act of leaping in the air. Some writers have claimed that semanticity is exclusively human. Animals may be able to communicate only about a total situation. A hen who utters ‘danger’ cries when a fox is nearby is possibly conveying the message ‘Beware! Beware! There is terrible danger about!’ rather than using the sound to ‘mean’ FOX. But, as was shown by the call of the vervet monkey who might mean ‘snake’ when it chutters, it is difficult to be certain. We must remain agnostic about whether this feature is present in animal communication.

Cultural transmission
or
tradition
indicates that human beings hand their languages down from one generation to another. The role played by teaching in animal communication is unclear and varies from animal to animal – and even with species. Among birds, it is claimed that the song thrush’s song is largely innate, but can be slightly modified by learning, whereas the skylark’s song is almost wholly learned. Birds such as the chaffinch are particularly interesting: the basic pattern of the song seems to be innate, but all the finer detail and much of the pitch and rhythm have to be acquired by learning (Thorpe 1961, 1963). However, although the distinction between humans and animals is not clear-cut as regards this feature, it seems that a far greater proportion of communication is genetically inbuilt in animals than in humans. A child brought up in isolation away from human beings, does not acquire language. In contrast, birds reared in isolation sing songs that are sometimes recognizable, though almost always abnormal.

The fifth and sixth features are social ones, in that they relate to the way in which language is used.
Spontaneous usage
indicates that humans initiate speech freely. Speaking is not something which they do under duress, like a dog that will stand on its hind legs only when a biscuit is held above its nose. This feature is certainly not restricted to humans, and many animals use their natural communication systems freely. The other social feature,
turn-taking
, means exactly what it says: we take it in turns to speak. In the majority of conversations, we do not talk while other people are talking, nor do we compete with them. Instead, we politely wait our turn, as shown in a brief conversation between two characters in P.G. Wodehouse’s
Carry on Jeeves
: