Critique of the linguistic theory of Noam Chomsky
What is behind the theory that language structures in the mind are inborn?
Noam Chomsky, the American professor of linguistics, is well known in progressive circles for the courageous stand he took against US imperialism on the question of the Vietnam war, and which he continues to take on the question of peace in the middle east.
Because he sets a fine example of courage and principle on these issues, it might be thought that he was also progressive as regards his philosophic thought. This, unfortunately, is not at all the case. Even on other political issues, such as, for example, his analysis of the Spanish civil war, which takes the same anti-working class, anticommunist, line as the MI5 informer George Orwell, his understanding is unreliable.
And when we look at his philosophical views, we can see why he is prone to make political mistakes, notwithstanding the finer points of his character.
For Chomsky’s philosophical views are influenced by assumptions taken from the positivist, empirio-criticist stable, thoroughly criticised and denounced by Lenin in his Materialism and Empirio-Criticism, a book Lenin felt constrained to write because he found that comrades who could not properly distinguish between positivism and materialism, besides getting hopelessly muddled, were in danger of losing their way and ending up in the enemy camp, siding with the reactionary classes rather than the revolutionary proletariat. (1908)
Before embarking on explaining just how Chomsky’s philosophical views are wrong, and demonstrating their relationship to positivism, let us briefly recap the essence of materialist philosophy, to which positivism is opposed.
According to materialism, which Marxism embraces, all our ideas have but one source – the material world. If there were no material world, we would have no ideas.
If, on the other hand, humans were not around with their ideas, this would not affect the existence of the material world one jot. This is what is meant when materialists say that matter is primary and ideas are secondary.
In this, materialism is opposed by idealism, which considers that, on the contrary, what we think of as the natural, material world is but a creation of our own ideas. Standing ‘between’ materialism and idealism (or rather wavering between materialism and idealism and not knowing which to accept) are various philosophers who have tried to ‘reconcile’ these two opposite conceptions by suggesting that knowledge (truth) is made up partly of ideas and partly of the material world – that it is a fusion of both.
Lenin’s above-mentioned pamphlet shows that these would-be compromisers, who think they have bridged the unbridgeable gap between materialism and idealism, end up logically 100 percent in the idealist camp, since they are forced to concede that until humanity came on the scene, the world could not have existed, because humanity was not there to contribute its ideas to that part of ‘reality’ which necessitated, according to their theories, the human touch.
Idealists object to materialism by pointing out that people are well able to have ideas that have no equivalent in nature. For instance, we can imagine a purple sun, although we have never seen any such thing. We also have ideas we believe to be correct, but in fact turn out to be wrong – which obviously do not reflect the material world.
If ideas were merely a faithful reproduction in the human mind of what is happening in the material world, then everybody’s ideas would be the same, and none of them would ever be wrong.
But materialists would never claim that people’s ideas represent at any given time a fully accurate, fully detailed model of the material world that has given rise to them. They claim that the human organism strives (automatically and unconsciously) to reproduce within the brain models (which it does by the million) of the phenomena it encounters in the material world.
These models are our ‘ideas’, our ‘conceptions’ of the world. They are quite like computer programs, except that they are self-loading. For instance, when I find myself wanting to tie my shoelaces, my shoelace-tying model automatically makes itself available, enabling me to tie my laces without even consciously thinking about it, and then makes way for other programs as soon as the job is over.
If I have not used a program for a long time – for instance, my how-to-drive-a-car-with-an-automatic-gearbox program – it may be harder to retrieve, and may not click automatically, leaving me sitting at the driver’s wheel wondering what to do next. But gradually, as I touch the controls, or with a few words of reminder from somebody, it ‘all comes back’.
In putting together these models, there is plenty of scope for error. In the first place, the information we obtain about the phenomenon we are unconsciously building a model of may be very scanty, because of the limitations of our data-input system – ie, our five senses. We put together the model as best we can from the limited data available to us, filling the gaps (still unconsciously and automatically) from what we know about other things.
For instance, if I encounter a new species of animal that I have never heard of before and see this animal standing in front of me, not knowing what it is, in the mental model I automatically construct of it, the animal’s blood will be red – not because I have seen its blood but because the blood of similar animals is red.
A model may also be deficient because, although I have perceived a particular characteristic of a phenomenon through my sense data, I have not incorporated it in my model because it did not strike me as important. The building brick was there, but it was discarded.
And finally a model may be fantastic, in the sense that it is put together from bits and pieces of other models but has no counterpart in the material world. When we imagine a purple sun, which we have never actually seen, we are making use of our working memory to pick out characteristics of things that exist in nature and fitting them together in different combinations.
It is a process rather like cutting up pictures of different animals and then reassembling pictures that have different heads on all the bodies than those that belong to them. To be able to imagine a purple sun, however, we do have to have obtained from the material world the idea of purple and the idea of sun. Otherwise, we could not conjure up the idea of a purple sun.
Since it is not an idea that is of any particular use to us, it will be nothing more than a passing fancy, forgotten almost as soon as it is conjured up.
Still, humans’ ability to contemplate combinations of data other than those presented to them in nature is the basis for human beings being able to intervene physically in the material world to cause these new imaginary combinations to come about in reality – if they believe there may be some advantage to be gained from doing so.
For instance, a group of people seeking shelter could be presented with a forest full of trees. They can imagine if the trees were arranged differently, they would provide shelter. They can then intervene to change the arrangement of the trees to accord with the imagined possibility – cut them down and bunch them together, laying them horizontally instead of vertically, to provide the necessary shelter.
From the scanty data we have about the way the human mind works, what kind of model of the mind are we able to piece together – assuming that we are consistently materialist and do not wish to ascribe to any theory that attributes ideas to divine intervention?
Some of the important characteristics of a working model of the human mind would be as follows:
1. The ability to recognise patterns of sense stimulation arising from a human being’s contact with nature. For instance, I recognise places and people I know because they ‘match’ the mental models I have built up through regular contact.
2. The ability automatically to incorporate data from other mental models into new mental models that are being constructed, if those data appear to be useful in getting the model to work.
3. The ability consciously to incorporate into mental models both old and new information gained from other people who can help me build better models:
(i) by drawing my attention to typical errors I should avoid (that other people have discovered by experimentation when they made the same erroneous assumptions I had been making – eg, an assumption that light things fall more slowly than heavy things); (ii) by drawing my attention to significant items that I would have been inclined to overlook; (iii) by providing me with information about what other people have been able to prove exists but is beyond the scope of my sense perceptions (eg, infra-red or ultra-violet).
4. The best mental models are multidimensional, built up as a result of the person in question actively interacting with the material world, which gradually whittles away errors and misconceptions. Hence a person who has little understanding of a particular process will gain understanding if he involves himself in it.
This is why it is often useful to get into a problem armed only with formulas or routine procedures that one does not fully understand. As one is involved, nature gradually shapes our mind so that it acquires a reasonably accurate mental model of the process we are involved in and we thus learn to understand it better.*
5. The models in use discard unnecessary detail. My arithmetical model, which tells me that 5+5=10, may be based on five fingers plus five fingers being equal to ten fingers, and five wooden blocks plus five wooden blocks also being equal to ten wooden blocks, and five steps plus five steps also being equal to 10 steps. But since 5+5=10 every time, there is no need to incorporate in my arithmetical model the steps, the fingers, the blocks or anything else.
6. This process of pattern recognition is quite automatic (although it can be made even more efficient by conscious procedures). Chomsky cites the example of a slave boy whom Socrates discovered ‘knew’ principles of geometry: he knew what a triangle was and some of its characteristics; he knew about straight lines, squares, circles and some of their characteristics, despite never having been to school. Plato thought this fact ‘proved’ the transmigration of souls (the slave boy must have received his education in an earlier life), and Chomsky thinks it ‘proves’ that some knowledge is genetically acquired.
What both fail to appreciate is that although the slave boy has not been to school to receive support for the conscious improvement of his mental models, nature has been forming models in his mind since he was born.
The concept of a triangle is in his mind because in his life he is constantly encountering triangles and his mind has picked out the pattern ‘triangle’ from a million different concrete manifestations in nature of triangles. Besides seeing triangles, he will have handled triangular objects, and thus learnt a great deal about triangles, unbeknownst to himself, to Plato or to Chomsky.
An example of the sheer power of human beings’ pattern recognition faculties is provided by language acquisition.
Language is primarily an association of different sounds with the different phenomena of the material world. The human larynx is physically able to produce an infinite range of different sounds, which is what renders language possible.
But the material world is not just made up of concrete objects. The objects have distinguishing features even among objects of the same kind. Furthermore, everything in the universe is in a state of perpetual motion, with everything reacting and responding to everything around it.
Hence human languages over the millennia developed words to stand in not only for concrete things, but also for their characteristics and for their actions, and the characteristics of those actions. Concrete things are represented by nouns (cat, tree, sky, universe), distinguishing characteristics of concrete things by adjectives (green, smelly, extraordinary), action words by verbs (run, speak, sit, sleep) and distinguishing features of actions by adverbs (quietly, fast, roughly).
Since everything in nature exists in time and space, there have also developed words to express the location of things and activities in time and space and in relation to one other. All developed languages have these features, because all languages have as their purpose to enable one human being to communicate to another the ideas in his mind, which in turn are all models (more or less accurate, as we have seen) of the material world.
Humanity has developed languages over the millennia on the principle that when humanity perceives that some thing, process or relationship exists in the material world, it devises a word for it in order to be able to communicate about that thing, process or relationship, as the case may be.
Modern philosophers of language are prone to suggest that things can only exist if we have a word to describe them. Once our language fails us there is nothing we can say.
Yet somehow we do manage to find the words, or adapt existing words, to include the new phenomenon we need to describe: eg, the computer ‘mouse’. New words flood the market.
The development and use of language involves, as has been said, the use of humanity’s subconscious recognition of repeated patterns and an ability to extract those patterns for generalised use.
For instance, the existence and use of the word ‘house’ covers a huge range of concrete manifestations of house. All language involves a subconscious process of abstraction of this kind.
In addition, all languages have grammatical patterns applying different patterns of construction to nouns, adjectives, verbs and other parts of speech. And it is here where some of the most striking features of human pattern recognition ability can be demonstrated.
For without anybody teaching a child the rules of grammar, the child is able to apply them automatically without even realising that she is doing so. It becomes apparent that a child is applying these grammatical rules which she has subconsciously identified and incorporated in a mental model of her native language when she makes mistakes such as: “When I went to Granny’s I seed her friend.”
So efficient is the child’s pattern recognition ability that in learning language she only needs conscious instruction on specific items that do not fit the usual patterns.
Incidentally, it was apparently experimentally established many years ago that a baby only a few days old is able to distinguish his mother’s language from other languages. This can only go to show that the processing of sound patterns begins while the child is still in the womb and is astonishingly efficient.
It has, incidentally, also been shown that babies only a few days old whose mothers watch the television regularly are able to recognise TV advertising jingles. One assumes that this must be a manifestation of the same pattern recognition ability, rather than any genes bearing Pepsodent advertisements.
And so to Chomsky and his theory of universal grammar.
Noam Chomsky asserts that language structure (nouns, adjectives, verbs, etc) must be innate, that is, that it must be genetically encoded.
His main argument in favour of this hypothesis is that when you dissect the brain you find that those areas where language processing activities are carried out are in man more complex and developed than is the case in the corresponding areas in the brains of animals. He concludes that this must mean that those areas of the brain contain an innate grammatical structure.
This conclusion does not at all follow, however, from the anatomical observations (the accuracy of which we would not question). These observations are every bit as consistent with the fact that human beings have a huge capacity to associate sounds with the millions of phenomena of the material world, using a different sound for lots of different things.
The English language contains about 20,000 words, all of which have several different meanings, dependent on context. These words have patterns of use reflecting phenomena of nature. To store all that information does require a well-developed area of the mind.
Chomsky would admit that in a dog the parts of the brain where the sense of smell is located are more developed than they are in humans. Because of this, a dog can smell a huge range of things that a human being cannot.
The dog gives evidence of this ability in practice by being able to find things by following scent in a way that no human being could. This does not mean that the source of the smell is in the dog’s brain; that the dog’s brain has some innate grammar of smells enabling it to categorise minute quantities of molecules passing up its nostrils.
All it means is to say it has more receptive sensors to smell than humans have and can smell things that human beings can’t; just as humans can say things that dogs cannot, since, apart from all else, barks, whines and yelps do not lend themselves to very sophisticated communication, although they do produce a limited language, which can and is used to enable the dog to communicate to a very limited extent.
It is obvious from what has been said that Chomsky belongs to that school of thought which will not accept that the source of all our ideas is the material world. He seeks to promote the theory that some of our ideas are inborn, in the form of language structures in the mind.
This is all in the positivist tradition, although it has to be said there is no precise definition of positivism. Almost every person whom a Marxist would consider to be following an essentially positivist approach would vehemently deny being a positivist – mainly because of differences of detail with other positivist philosophers.
In fact, many of them would claim to have spent their whole lives proving the falsity of positivism! From a materialist point of view, however, they have merely spent their whole lives trying to rescue positivism from the absurdities into which it drives those who embrace it.
The essence of positivism is that it denies that the material world is the sole source of knowledge and claims that the human mind contributes Something to the equation. The differences between positivists centre round what that Something is.
The problem positivists especially sought to solve was that of scientific truth. How can one distinguish between what is true and what is false?
A materialist answers this question by reference to the material world. Does your idea work when you put it into practice? If it does it must be true (up to a point at least): your model of reality is serviceable. If the idea does not work when you put it into practice, there must be something wrong with it.
It may be a question of detail, or the model may be totally misconceived. But there is an error somewhere. Practice (ie, the material world) is the sole criterion of truth, since it is the only source of knowledge.
But if you are a positivist, you do not accept this. You are at best agnostic (undecided) about whether your senses convey to you information about the way the material world really is.
Yet you can hardly deny that all humanity has concepts of true and false, and that these concepts are absolutely central to the work of scientists. Scientists’ whole existence is devoted to the pursuit of truth. (It is a pity that philosophers do not all follow their example!) But if the criterion of truth is not practice, material reality, then how do you distinguish truth from falsehood?
The positivists’ answers to this question are many and varied. The philosophers most associated with positivism today, the logical positivists (Ayer, Russell, the Vienna circle) say that a statement is true if logically consistent, the assumption being that logic is some kind of genetically-inbuilt litmus test, rather than a set of patterns actually occurring very regularly in the material world, obscured by a mass of detail, but detected by philosophical geniuses, especially Aristotle.
Once Aristotle had abstracted the principles of logic from the mass of detail in which they concretely manifest themselves, he provided humanity with a powerful tool for understanding the material world, and all kinds of happenings which had previously been mysterious become ‘only logical’, and therefore predictable.
However, logic can only be a useful tool for predicting outcomes to the extent that it is based on premises (assumptions) that correctly reflect material reality.
Indirectly, under the guise of ‘logic’ (because ‘logic’ is derived from matter), the positivists were to some extent capable of conducting a struggle against theologians trying to defend a divine order that was in fact merely a reflection of the ‘logic’ of feudal relationships.
But as humanity advances the scope of its investigations from the terrestrial sphere into outer space (where the ‘logic’ of Newtonian mathematics no longer prevails), and as it is faced with the pressing necessity of terminating the exploitation of man by man, which has lain at the basis of society for thousands of years shaping people’s ideas of what was ‘logical’, the old ‘logic’ is time after time shown no longer to be universally valid, leaving logical positivism rather obviously high and dry.
And leaving the philosophers who still stubbornly refuse to accept materialism dashing around everywhere trying to find a substitute for logic as their criterion of truth.
One strain of bourgeois philosophy has decided that there is no such thing as truth, it is just something we are linguistically programmed to talk about and therefore believe in.
With the downfall of ‘logic’ as the criterion of truth (a downfall brought about by the material world itself refusing to behave as insufficient human ideology believes it should), many bourgeois philosophers have turned to ‘language’ to play the role in an essentially positivist approach that previously was played by ‘logic’.
Hence Chomsky’s thesis that language is to some extent genetically programmed, and the frantic attempts by philosophers to discover exactly to what extent language is programmed in our minds, which now take the place of the frantic endeavours of early 20th century philosophers to find out exactly what logical rules were genetically programmed in our minds, and even to what extent those genetic programs can be relied upon!
To the extent that all ideas ultimately reflect matter, the fact that positivist philosophy is seeking in its way to explain the world – even though it is denying that this is something that can be done – does produce valuable observations.
By focussing on language and linguistics, even if for the wrong reasons, important discoveries have been made and lessons learnt on these subjects. But we should not be tempted, as a result, to imagine that we have to accept the positivists’ basic premise that we can know only our own ideas and not the world itself.
If we do not accept material reality as the ultimate arbiter of the correctness of our ideas, we have no yardstick by which to judge the validity of our ideas and no means to correct mistakes, leaving us helplessly entangled in those mistakes.
Surprising evidence recently emerged that completely discredits the theory of universal grammar, although this has not been as widely publicised as it deserves.
A tribe, the Pirahã was discovered in Brazil, living deep in the jungle, completely cut off from other human beings, knowing neither herding of animals nor agriculture, and which had not therefore advanced to class society or civilisation.
Their language is truly prehistoric and was found not to follow the alleged rules of ‘inborn’ grammatical structures identified by Chomsky at all. It has, for example, no past or future tenses, does not distinguish male and female, has no number system, etc, etc.
This language reflects a society that has not yet developed a need to express these concepts in practice and whose language has, therefore, not evolved to enable people to do so.
The missionary who learned the Pirahã’s language had formerly been a disciple of the Chomsky linguistic school, but was forced through his own experience to repudiate the theory. He also came to repudiate the christian missionary work he had been sent to carry out (translating the Bible into Pirahã), realising that the Pirahã would be better off without it! (Daniel Everett, Don’t Sleep. There are Snakes, 2008)
_____________________
* When ideas are expressed through language [originally done for the purpose of communication], it also makes it easier to use language for adjusting those ideas. This is the process of taking an idea that has previously existed only in the subconscious [eg, the rules of grammar in a small child using its mother tongue] and forcing it into the conscious mind.
This process requires language; it can’t be done otherwise. Once in the conscious part of the mind, if the idea is a wrong one, it is easier to identify what’s wrong and adjust it, or to consider possible alternatives.
And, of course, language is susceptible to being written down – which hugely expands the possibility of being able to critique ideas with a view to adjustment. In other words, language is a fantastic tool for problem solving.
In the sense the word ‘idea’ is being used in this article, it can exist without language. If an idea is a reflection of the material world then the reflection can be picked up in the mind without the intervention of language. To say otherwise necessarily would make a huge concession to Chomsky-type positivism.
However, theories [ideas we have of which we are conscious] can only exist linguistically. [There may be an exception here in the case of problem-solving among animals, but animals’ ability to solve problems is so very poor that if there is an exception, it is but one that proves the rule].
Language reflects ideas. What is certainly true is that very early on in life, our ideas become associated with words, so that it becomes difficult for us to evoke an idea without at the same time evoking the word associated with that idea in our minds.
But we do tie shoelaces, for example, in accordance with our well-established ‘idea’ of how this process works, without consciously thinking “right over left” every time we do so. The only time we would resort to language would be if we set about the process in the usual ‘mindless’ way and much to our amazement things didn’t work out as per expectation predicated by our mental model. Then language would come into the process of working out what went wrong and how to put it right – in addition, of course, to expressing our frustration!
