Department of Linguistics
ANIMAL COMMUNICATION AND LANGUAGE
Robert Mannell (1999)
The aim of this lecture (1) is to examine the following questions:-
- How do the forms of communication used by animals differ from human language?
- Can animals be taught to use languages that are analogous to or the same as human language?
Natural Animal Communication
Pearce (1987, p252) cites a definition of animal communication by Slater (1983, see Pearce for reference), which we will also use as a working definition in this lecture:-
Animal communication is "the transmission of a signal from one animal to another such that the sender benefits, on average, from the response of the recipient".
This loose definition permits the inclusion of many types of behaviour and allows "communication" to be applied to a very large range of animals, including some very simple animals.
Natural animal communication can include:-
- Chemical signals (used by some very simple creatures, including protozoa)
- Smell (related to chemical signals, eg. pheromones attract, skunk secretions repel)
- Posture (eg. dogs, geese)
- Facial gestures (eg. dogs snarling)
- Visual signals (eg. feathers)
- Sound (eg. very many vertebrate and invertebrate calls)
Such signals have evolved to:-
- attract (especially mates)
- repel (especially competitors or enemies)
- signal aggression or submission
- advertise species
- warn of predators
- communicate about the environment or the availability of food
Such signals may be:-
- instinctive, that is genetically programmed
- learnt from others
Some linguists (eg Chomsky, 1957, Macphail, 1982, both cited in Pearce, 1987) have argued that language is a unique human behaviour and that animal communication falls short of human language in a number of important ways.
Chomsky (1957) claims that humans possess an innate universal grammar that is not possessed by other species. This can be readily demonstrated, he claims, by the universality of language in human society and by the similarity of their grammars. No natural non-human system of communication shares this common grammar.
Macphail (1982, cited by Pearce, 1987) made the claim that "humans acquire language (and non-humans do not) not because humans are (quantitatively) more intelligent, but because humans possess some species-specific mechanism (or mechanisms) which is a prerequisite of language-acquisition".
Some researchers have provided lists of what they consider to be the criteria that animal communication must meet to be regarded as language.
For this lecture the list devised by Hockett (1960) is utilised, although this list is not the only such list available. Such lists tend to be quite similar and certain elements of the Hockett list are considered particularly important in evaluating the question "can animals be taught language?"
Hockett's thirteen "design-features" for language are as follows:-
- Vocal-auditory channel: sounds emitted from the mouth and perceived by the auditory system. This applies to many animal communication systems, but there are many exceptions. Also, it does not apply to human sign language, which meets all the other 12 requirements. It also does not apply to written language.
- Broadcast transmission and directional reception: this requires that the recipient can tell the direction that the signal comes from and thus the originator of the signal.
- Rapid fading (transitory nature): Signal lasts a short time. This is true of all systems involving sound. It doesn't take into account audio recording technology and is also not true for written language. It tends not to apply to animal signals involving chemicals and smells which often fade slowly.
- Interchangeability: All utterances that are understood can be produced. This is different to some communication systems where, for example, males produce one set of behaviours and females another and they are unable to interchange these messages so that males use the female signal and vice versa.
- Total feedback: The sender of a message also perceives the message. That is, you hear what you say. This is not always true for some kinds of animal displays.
- Specialisation: The signal produced is specialised for communication and is not the side effect of some other behaviour (eg. the panting of a dog incidentally produces the panting sound).
- Semanticity: There is a fixed relationship between a signal and a meaning.
- Arbitrariness: There is an arbitrary relationship between a signal and its meaning. That is, the signal, is related to the meaning by convention or by instinct but has no inherent relationship with the meaning. This can be seen in different words in different languages referring to the same meaning, or to different calls of different sub-species of a single bird species having the same meaning.
- Discreteness: Language can be said to be built up from discrete units (eg. phonemes in human language). Exchanging such discrete units causes a change in the meaning of a signal. This is an abrupt change, rather than a continuous change of meaning (eg. "cat" doesn't gradually change in meaning to "bat", but changes abruptly in meaning at some point. Speech loudness and pitch can, on the other hand be changed continuously without abrupt changes of meaning.
- Displacement: Communicating about things or events that are distant in time or space. Bee dancing is an example of this.
- Productivity: Language is an open system. We can potentially produce an infinite (2) number of different messages by combining the elements differently. This is not a feature of, for example, the calls of gibbons who have a finite number of calls and thus a closed system of communication.
- Traditional transmission: Each generation needs to learn the system of communication from the preceding generation. Many species produce the same uniform calls regardless of where they live in the range (even a range spanning several continents). Such systems can be assumed to be defined by instinct and thus by genetics. Some animals, on the other hand fail to develop the calls of their species when raised in isolation.
- Duality of patterning: Large numbers of meaningful signals (eg. morphemes or words) produced from a small number of meaningless units (eg. phonemes). Human language is very unusual in this respect. Apes, for example, do not share this feature in their natural communication systems.
Click here to see a table that examines the extent to which various communication systems meet these 13 design features.
Teaching Language to Apes
(and other animals)
It seems well established that no animal communication system fulfils all of the criteria outlined by Hockett (1960). This is certainly true for the apes. It is also true for most other species such as parrots and may also be true for animals such as dolphins, who have a complex communication system which involves a complex combination of various sounds.
Why try to teach a human-like language to another species?
Just because a species doesn't have such a communication system in the wild doesn't necessarily prove that they are incapable of using one.
What kind of language should we teach these animals?
We must avoid using features of human language that are physiologically difficult or impossible for the animal to manage.
For example, spoken human language is extremely difficult or impossible for most animals because of the structure of their vocal organs. Apes, for example, can't produce a large proportion of the vowels and would have difficulty with some of the consonants. This may be due not only to the shapes of the vocal organs but also to the limitations of the motor centres in the brain that control these organs. We might attempt, on the other hand, to teach apes language that involves them using their hands (eg. sign language or the manipulation of symbols).
Some birds, such as certain parrots and the Indian Hill Mynah, are able to mimic human speech with great clarity. We could, therefore, attempt to teach such animals spoken human language.
Dolphins cannot be taught either type of language but may be able to understand sounds or gestures and to respond by pressing specially designed levers.
What do we test for?
Animal communication systems generally lack one or (usually) more of the following features:-
Most researchers attempting to teach language to animals are attempting to test for the existence of these features in the "language" use of their subjects.
Projects with Apes
The ape species include gorilla, chimpanzee, bonobo (a distinct species of chimpanzee) and the orangutan. Apart from some very early attempts to teach spoken language to chimpanzees (generally resulting in the production of no more than 3-4 words) language production training has involved the use of the hands, either through the manipulation of symbols or through the use of sign language. Comprehension training has involved these types of language as well as training in the comprehension of spoken language.
Here are some of the most important studies on apes and language:-
- Gardner and Gardner (1969)
American sign language
- Patterson (1978)
- Premack and Premack (1972)
Chimpanzee (Sarah and others)
- Terrace et al (1979)
Chimpanzee (Nim Chimpsky)
- Rumbaugh and Savage-Rumbaugh
Chimpanzee (Sherman and Austin)
Symbols on a keyboard
Bonobo Chimpanzee (Kanzi, Panbanisha)
Understanding spoken language
Symbols on a keyboard
There are some web pages that you might wish to look at. They describe the work of Sue Savage-Rumbaugh and colleagues. They include:-
- Great Ape Trust, Iowa, USA
- A 1995 New York Times article entitled "Chimp Talk Debate: Is it really language?", by George Johnson
- A British newspaper report (July, 1999) describing the use of a speech synthesiser interfaced to the bonobos' keyboards.
Projects with birds usually involve parrots or the Indian Hill Mynah. These birds are selected for their ability to mimic human speech. The African Grey Parrot and the Indian Hill Mynah are generally considered to be the birds with the greatest ability to mimic human speech patterns but a number of other species (mainly parrots such as the budgerigar) can be trained to "speak".
African Grey Parrot (Alex)
Here are a few web pages that discuss the work of Irene Pepperberg and colleagues with Alex the African Grey Parrot.
- The Alex Foundation research page (links to various papers and a short movie)
- "Studies to determine the intelligence of African Grey Parrots", Irene Pepperberg, 1995
Projects with Cetaceans
Cetaceans, such as whales and dolphins, have been shown to be readily trainable to respond to gestures and sometimes to verbal and other acoustic commands. Also, many species have very complex acoustic communication systems. It has been hypothesised that it may be possible to train them to understand language encoded in either gestures or appropriate acoustic signals. Appropriate acoustic signals are assumed to be sounds that are similar to the natural communicative sounds that these animals produce. In the project listed below, one dolphin was trained on gestures and the other with sounds. (refer to chapter 8 of Pearce (1987) for a description of this project).
- Herman, Richards and Wolz (1984)
Dolphins (Akeakamai and Phoenix)
- Bindra, D.; Patterson, F.G.; Terrace, H.S., Petitto, L.A., Sanders, R.J., and Bever, T.G., (1981), "Ape Language", Science, 211, 86-88. (3)
- Brakke, K.E., & Savage-Rumbaugh, E.S., (1996), "The development of language skills in pan - II. Production", Language & Communication, Vol.16, No. 4, pp. 361-380.
- Chomsky, N., (1957), Semantic structures, The Hague: Mouton
- Frisch, K. von, (1962), "Dialects in the language of bees", Scientific American, (also published in:- Wang, W. S-Y., (ed.), (1982), Human communication: Language and its psychobiological bases, Scientific American)
- Gardner, R.A. & Gardner, B.T., (1969), "Teaching sign language to a chimpanzee", Science, 165, 664-672.
- Hockett, C.F., (1960), "The origin of speech", Scientific American, (also published in:- Wang, W. S-Y., (ed.), (1982), Human communication: Language and its psychobiological bases, Scientific American)
- Macphail, E.M., (1982), Brain and intelligence in vertebrates, Oxford: Clarendon
- Patterson, F.G. (1978). "The gestures of a gorilla: language acquisition in another pongid.", Brain and language 5: 72-97.
- Pearce, J.M., An Introduction to Animal Cognition, Lawrence Erlbaum Associates, 1987, Chapter 8, "Communication and Language", pp251-283 (Library call number QL785.P42)
- Premack, A.J., & Premack, D., (1972), "Teaching language to an ape", Scientific American, (also published in:- Wang, W. S-Y., (ed.), (1982), Human communication: Language and its psychobiological bases, Scientific American)
- Savage-Rumbaugh, E.S., Murphy, J., Sevcik, R.A., Brakke, K.E., Williams, S.L., & Rumbaugh, D.M., (1993), "Language Comprehension in Ape and Child", Monographs of the Society for Research in Child Development, Serial No. 233, Vol. 58, Nos. 3-4
- Sparks, J., (1969), Bird behaviour, London: Hamlin.
- Terrace, H.S., Petitto, L.A., Sanders, R.J., & Bever, T.G., (1979), "Can an ape create a sentence?", Science, 200, 891-902.
- Williams, S.L., Brakke, K.E., & Savage-Rumbaugh, E.S., (1997), "Comprehension skills of language-competent and nonlanguage-competent apes", Language & Communication, Vol.17, No. 4, pp. 301-317.
1. This topic was originally presented as a first year linguistics lecture in early November 1999. The content hasn't been significantly updated since then so it's now very likely to be out of date. A continuing effort has been made, however, to keep external links up to date. In some cases, as external links have ceased to exist, local copies of the affected documents have been made available.
2. Click here for a discussion of the claim that a human language can potentially have an infinite number of sentences.