Language and Cognition

What is language? This simple question completely defies a simple answer. For most investigators, language consists of abstract representations of objects, properties, actions or emotions. These abstractions, which are generally vocal or gestural, are ordered according rules (the rules are called syntax) and may be rearranged in many different ways (recursion) according to the desired meaning. This definition comes largely from human experience with our own languages, and may be somewhat limiting when considering possibilities for abstraction in animal communication. Some linguists and biologists argue that elements of language formation and usage are unique to humans (eg Hauser et al. 2002).

A few species exhibit the ability to learn, respond to, and form language following patterns that conform fairly well with human language usage. These include chimpanzees, bonobos, parrots, and dolphins. Humans can, through programs of reinforcement, induce animals in these species to learn to understand human expression (or human generated artificial sets of abstractions) and in some cases to communicate with the human observer. In a few cases, if more than one animal has been taught the abstractions, animals may use them to communicate among themselves. None of these species, though, appears to use these facilities in the wild, and none of them generates the complexity of recombinations of the abstractions seen in human communication. This observation leads to several important questions:

• If they can do it, why don't they seem to do it in the field?
• If they don't seem to be doing it, is it just because we haven't measured it in the right way?
• If they really don't do it, what is there any importance or significance in having these capabilities for these species?

Clearly we don't yet have the answers to these questions. What we learn from training animals in semi-domesticated situations to communicate is that they have the flexibility to use their neural capabilities in ways that, for us, are unexpected.

The ability to recognize and respond to human vocal commands extends to a wide range of species in addition to the stars of animal language learning (chimpanzees, bonobos, parrots and dolphins). Some individual dogs seem to be able to differentiate as many as 100 words or vocal commands. Cats have an impressive but somewhat more limited capacity (at least in terms of what they deign to show us). Ungulates (hooved mammals--horses and the like) can learn to respond to their names and to a number of simple verbal commands. These species, for the most part, can also learn to respond to visual signals, like the hand signals used by handlers of field dogs. All of these animals respond to a variety of signals, but they seem to lack a similar diversity of signals in their own communication. Is this seeming lack of diversity of signals because its not there, or because we don't know how to look for it?

If an imitative capacity is essential for forming language, then non-human primates--the monkeys and apes--are poorly equipped for using language (Hauser et al. 2002). In fact, imitation is found in a far more limited set of animals than is command-following. In many bird species vocal communication is learned by imitation, but the capacity to imitate is limited to a few appropriate sounds and learning can only occur during a short period of time in the bird's life. More open-ended imitation of sounds occurs in a few bird species--parrots and their relatives, starlings--and in dolphins.

In the sense used here, the "dance language" of honey bees isn't a language at all. While some biologists have argued for a cognitive component of the bee dance language, in fact the highly stereotyped patterns used in bees to communicate the distance and direction of food resources includes none of the imitative or recursive elements that come into play in human communication. Some animals are able to specify enemies with different vocalizations (Seyfarth et al. 1980, Fischer 1998). These abstractions are important to the survival of these species, but are not structured with syntax, nor are they recursive (involve large numbers of possible recombinations of the abstractions). These are great examples of communication, but probably not examples of language.

Fischer J 1998 Barbary macaques categorize shrill barks into two call types
ANIM BEHAV 55: 799-807
Hauser MD, Chomsky N, Fitch WT 2002 The faculty of language: What is it, who has it, and how did it evolve? SCIENCE 298 (5598): 1569-1579
Hihara S, Yamada H, Iriki A, et al. 2003 Spontaneous vocal differentiation of coo-calls for tools and food in Japanese monkeys NEUROSCI RES 45 (4): 383-389
Kako E 1999 Elements of syntax in the systems of three language-trained animals ANIM LEARN BEHAV 27 (1): 1-14
Menzel CR, Savage-Rumbaugh ES, Menzel EW 2002 Bonobo (Pan paniscus) spatial memory and communication in a 20-hectare forest INTERNATIONAL JOURNAL OF PRIMATOLOGY 23 (3): 601-619
Pepperberg IM 1999 Rethinking syntax: A commentary on E. Kako's "Elements of syntax in the systems of three language-trained animals" ANIM LEARN BEHAV 27 (1): 15-17
Pepperberg IM, Gardiner LI, Luttrell LJ 1999 Limited contextual vocal learning in the grey parrot (Psittacus erithacus): The effect of interactive co-viewers on videotaped instruction J COMP PSYCHOL 113 (2): 158-172
Pepperberg IM, Sandefer RM, Noel DA 2000 Vocal learning in the grey parrot (Psittacus erithacus): Effects of species identity and number of trainers. J COMP PSYCHOL 114 (4): 371-380
Pepperberg IM, Shive HR 2001 Simultaneous development of vocal and physical object combinations by a grey parrot (Psittacus erithacus): Bottle caps, lids, and labels. J COMP PSYCHOL 115 (4): 376-384Seyfarth R M, Cheney D L, Marler P 1980 Monkey responses to 3 different alarm calls - evidence of predator classification and semantic communication. SCIENCE 210 (4471): 801-803