to summary of reasons
Why the Chomskians Are Mistaken
(Sergio Navega's reasons begin at this point.)
28. Is grammar wired in our DNA?
Most linguists today would say that language clearly has genetic origins. They are confortable with the idea that the brain is somewhat "wired" for grammar and that there's a specific portion of our DNA that codes all the necessary information to generate such domain-specific language-producing organs in our brains. We humans would, according to such views, share this genetic endowment and so it can be usefully considered to be the site of the "universal grammar," not available in any other species in this world. Yet, this very idea is receiving constant criticisms from scientists who try to explain how this grammar could possibly be encoded in genes, and why it appears to be unique among humans.
In what follows, I shall attempt to list some of these criticisms and the more sound approach which considers language as just another learned trait. It is necessary to point out that there's a great advantage for linguists to proclaim the existence of universal grammars: the hard question of explaining how such a thing evolved is not their problem, this being a work to be completed by evolutionary biologists. By throwing this hot potato into other hands, they are mindlessly using a dogmatic point of view, which is incongruent with modern scientific methodologies and may impair the development of new ideas in linguistics.
A Related Question: What are the Origins of Language?
This is one of the still unresolved issues plaguing scientists all around the world. Instead of adding more hypotheses to the question, let's see what is really common to all humans in this world. Is it a universal language of some mathematically rigorous form as proposed by Chomsky?
Before trying to find a place for a universal grammar, let us look for something that we undoubtedly have in common with each other: we inhabit the same world. We share the same physical laws, we have similar sensory apparatuses, we are required to look for the same things (food, shelter, company, etc).
We have mainly the same kinds of cognitive problems: object perception, motor action coordination, planning, decision-making, etc. If I show you a specific pattern of vertical and horizontal lines forming a unique grid, I bet you'll be able to recognize this same pattern, no matter if it is reduced or enlarged, rotated or seen from a perspective.
This is not a human-specific ability: a cat does this and much more when walking along walls and roofs, showing impressive motor coordination. Amazingly, the processing of these visual scenes requires the use of sophisticated perceptual operations that, cognitive scientists are discovering, are very similar to the kind of recursive and embedded behavior typical of languages.
Language processing is neither more complex nor more special than the processing and integration of visual and auditory signals that cats, dogs, apes, lions, tigers and a huge number of mammals perform daily. So instead of being exclusively used in language, features such as embedding, recursion and perception of invariances are traits that have counterparts in our cognition, crossing into other senses.
Isn't this supposed to be obvious? We do indeed all have a single kind of brain, with basically the same kind of neurons, no matter how differently they seem to associate. Recent studies with ferrets have discovered, for instance, that the neural organization of the visual cortex emerges spontaneously in the auditory cortex, when the brains of these animals are surgically patched to redirect visual signals to the auditory region.
Why should we feed the hypothesis that language has a specific mechanism, when a general mechanism can do the same job and also explain vision and audition?
29. Is language a human-specific ability?
How tenable is the hypothesis that language (to a certain degree of grammatical sophistication) is specific to humans? Well, one can certainly say that nobody encounters a bird writing a poem, nor a monkey lecturing to fur-covered fellows, nor are there frequent meetings of the Whale Philosophical Society. It is important to notice, however, that there exists a common confusion between language and external symbolic storage (like words on paper).
One can have language without symbolic storage. Do we understand what dolphins are doing when they engage in those strange noisy exchanges? Their behaviors often suggest that they ought to be talking with each other, because one frequently catches them in some exquisite collaborative activities. These activities are typical of cooperation in communities of linguistically-oriented intelligent agents, and there are quite a number of computer models of spontaneous language emergence in such environments (see for instance Simon Kirby and Luc Steels).
Formal Grammars and Natural Languages:
There is a subtle difference between formal descriptions of grammar and natural languages as spoken by people. Formal languages are rigid, inflexible, don't tolerate noise anddue to recursioninfinitely extendable. To a Turing Machine, a formal language with embedded constructs is the shortest path to the nowhere land.
But nobody would say that phrases in a natural language could be embedded inside one another indefinitely. They seem to possess this recursive characteristic, but there is a clear limit: we can interpret sentences only up to a certain level of recursion, we have not enough memory to do it indefinitely. If this is a clear human limit, why is it so difficult for linguists to accept that Bonobo apes also display linguistic abilities, although in a constrained way? If an alien comes to Earth and if he (she? it?) is able to understand sentences with phrases embedded up to 50 levels, would we humans look like non-linguistic animals to them?
30. Language and Natural Selection:
No one can dispute that languages are the result of some sort of natural selection. Languages emerge spontaneously (as in the case of the Pidgin/Creole transformation and the case of the sign language developed by Nicaraguan children). What remains to be assessed is if "modules of language" in our brain could emerge by natural selection. The nativists say that some genetic alterations and further refinement by natural selection were enough to provide us with "language organs," and consequently with innate grammars. Their problem is to explain convincingly how this can be done in just some tens of thousands of years.
The point is that language, considered as a whole, is indeed a highly complex system requiring specific perceptual adeptness and adequate memory processing abilities. Such sophistication may perhaps be comparable to the complexity of an eye. Eyes evolved during millions of years and are common among most animals. Memory and perception are also common among animals. One could say that most preconditions to linguistic performance are already present (although in smaller degrees) in most mammals.
All these complex preconditions were not the result of a specific period of development. The more complex the feature is, the more "spread" in time its emergence must have been. Mechanisms for learning and performing language require characteristics that had to evolve in several hundreds of thousands of years, and were probably not directly linked to communication (in this light Arbib's paper about language and motor systems is particularly interesting).
It is more probable that language is the result of a "side effect", the same kind of situation that happened with wings of birds. Wings are complex structures that could not have emerged from one or few steps and not for the purpose of flight.
Everything in evolution is done through small adaptations. Wings could not have emerged in a "single" mutation, they were "designed" by evolution initially as better thermal insulation and later (a side effect) were used to support flight. Flight, obviously, is an ability that endows the animal with strong evolutionary advantages (it enables faster escapes from predators, it fosters migration, it allows hunting of prey more effectively, etc). However, instead of being concentrated in a specific kind of animal, wings are commonplace not only among birds but also among most insects.
It is more reasonable to suppose that most of the required processing traits that a brain must have in order to generate/understand language are relatively well distributed among closely related primates and most mammals. Only small details (such as the size of short-term memory or the autocatalytic effect of culture in our society) might separate us from them in terms of remaining linguistic capacity. There's no need here either for any universal grammar, or for language organs.
The issue becomes muddled a bit because of complexity. It is not only the ability of language that is important, it is also the society in which it is used. Language demands large and complex brains, but it also demands a community of users and speakers. In order for this to happen, it is also necessary to develop phonologically sophisticated apparatuses, able to utter enough variations to support complex linguistic expressions. All technical preconditions (perception, generative ability, etc) may be latent in other species, but lack of one of these last features may impair its natural emergence.
to the next reason
to the previous reason
to the summary of reasons
This piece is Copyright © 2000
by Alexander Gross, with specified
portions Copyright © 2000 by
Sergio Navega. It may be
reproduced for individuals and for
educational purposes only. It may
not be used for any commercial (i.e.,
money-making) purpose without
written permission from the authors.
All Rights Reserved.
to summary of reasons
to linguistics menu