Conceptualizing language

[This piece was published at The Electric Agora earlier this year. Chomsky's ideas on linguistics are very polarising and even my qualified endorsement of some of his central ideas prompted some animated exchanges.]

Complex language is a precondition not only for the kinds of interaction which characterize human societies but also for many kinds of thinking. It is both social and biological. A language only develops in a context of social continuity over an extended period of time, though it is typically learned very quickly by infants who are exposed to it. There is still controversy about the extent to which natural language is shaped and constrained by the structure and physiology of the human brain, but it is clear that the advent of complex human language was associated with genetic changes which impacted on various aspects of human physiology (including brain function).

What, then, is language? How should we conceptualize it? The approach I am outlining (and recommending) here is strongly idiolectal.

The term “idiolect” can be understood in different ways and taken more or less seriously in the study of language. Taken in a strong sense, it inclines us to see the individual rather than the language or linguistic community as the primary focus of study. As I see it, language only exists insofar as it is used (or instantiated) by individuals. A social context is a given. But speaking and writing and listening and reading and the thinking (or cognitive processing) which supports these activities or which impinges in some other way on linguistic forms or structures are all things which are done by, or (in the case of cognitive processing) occur within the brains of, individuals.

I don’t deny having been greatly influenced by Noam Chomsky’s ideas in my thinking about language. I was first introduced to linguistics by a former student of Chomsky’s who followed a broadly (but by no means doctrinaire) Chomskyan approach. This general approach appealed to me. Chomsky put the focus firmly on what he originally called (linguistic) “competence” (the individual speaker’s internal intuitions about grammar etc.) rather than “performance” (as a behaviorist might). This distinction was developed over time into one between I-language and E-language. For Chomsky the focus was on the former and consequently on idiolects rather than languages.

The entry for “idiolects” in the Stanford Encyclopedia of Philosophy (credited to Alex Barber and Eduardo Garcia Ramirez) highlights the philosophical implications of the concept and some of the confusions which surround it. Those who claim that idiolects (in any strong sense of the term) do not exist or that such a notion is useless or incoherent “are nonetheless happy to use the word ‘idiolect’ to describe a person’s partial grasp of, or their pattern of deviance from, a language that is irreducibly social in nature.”

But nobody is denying the social dimension of language. Of course a language is a social product, but “partial grasp”? Of what exactly? And (as I see it) any attempt to define idiolects in terms of patterns of deviance from a norm is likely to be arbitrary or trivial unless the norm itself is defined in terms of idiolects.

In what sense does a language exist as distinct from particular instances of language use? Spoken words and written texts are generally assignable to this language or that, but precise boundaries are impossible to draw. Grammars and dictionaries try to do this but they can never reflect the constantly shifting contours of actual linguistic practice which always depend on the knowledge and behavior of individual speakers. In the final analysis, then, what we have is a set of unique and (to a greater or lesser extent) overlapping idiolects. We find it convenient, however, to group sets of idiolects into what we call dialects or languages.

“The substantial debate,” Barber and Ramirez explain, “is not over how to define [the word ‘idiolect’]. It turns, rather, on whether an idiolectal perspective on language is to be preferred to a non-idiolectal one. Someone taking an idiolectal perspective on language treats idiolects […] as having ontological or investigative priority and [sees languages as] nothing but more-or-less overlapping idiolects. […] At issue, then, is what we should take languages to be.”

They go on to explain that Chomsky does not deny that language is at least in part a social product. But he is skeptical of E-language-based approaches. The term “E-language” is used by Chomsky to refer to those things (whatever they might be) that are the target of study for those who take languages and their properties to be external to the mind.

“Chomsky’s case for introducing and using the notion of an I-language is, in the end, indistinguishable from his case for a cognitivist approach to the study of language as a natural phenomenon. And his case against E-languages is that there is no scientifically coherent project to which they belong as posits.”

Chomsky does not deny the existence of some linguistic arbitrariness (emphasized by Ferdinand de Saussure and David Lewis, for example). But he sees the core aspects of language as being constrained by the specifics of our biological nature and the (undoubted) arbitrary and contingent aspects of language as operating within these constraints.

The facts of first language acquisition arguably demonstrate this. It is clear that language learning in infants represents a special kind of learning. Infants are not like little scientists observing and inferring the linguistic conventions prevailing among adult users. And even if they were, even if they were masterminds, they would still be unable, on the basis of the fragmentary, flawed and often inconsistent evidence which the typical linguistic environment provides, to zero in on an appropriate grammar. Logically speaking, there would be countless possible languages which would be compatible with the data. (This is the “poverty of stimulus” argument.) What we see in fact is very rapid, and apparently effortless, linguistic progress. And it calls for an explanation.

According to Chomsky, language acquisition can be thought of as a series of brain states, developing from an initial state, S0 [S zero], through intermediate states to a relatively stable mature state, SM.

From the SEP: "S0 is the initial state common to all humans, idealizing away from individual linguistic impairments and the like. Subsequent states arise through exposure to a particular linguistic environment. Nothing said so far requires that these states be thought of as representational states we could call “knowing a language”. […] [L]anguage acquisition can be described—usefully—as a matter of children evolving through various stages of knowledge en route to acquiring adult competence. This description is useful because the empiricist/nativist debate can now be couched as a debate over what linguistic information must already be known by someone in S0 if information supplied by the linguistic environment is to culminate in knowledge of the mature language M. Empiricists claim that nothing much is needed, that S0 is a “blank slate” to be filled in using environmental data. Nativists claim that plenty of information must already be provided, in the form of innate knowledge of a language dubbed Universal Grammar (UG) by Chomsky. We each come predisposed to acquire only certain languages, the humanly possible ones that can grow out of UG."

Despite the controversies surrounding the notion of Universal Grammar, I tend to agree with Barber and Ramirez that nothing much is added to this account, as an account of language learning, by describing it as development towards the learning of an externalistically specified social language (as opposed to some specific mature linguistic state (SM) of an individual). On this view, the primary target of investigation is the human language faculty, its nature and limits. Of course, other approaches to language are possible but, to the extent that they have scientific pretensions, they will probably be in tension with an idiolectal approach.

"Because there is considerable variety […] in the underlying conceptions of languages, Chomsky’s criticisms can seem sweeping, but the underlying thought is that, because E-languages are less “real” than I-languages, the concept [of an E-language] appears to play no role in the theory of language. […] Linguistic behaviour is the product of both the language faculty on the one hand and external influences—performance systems in the mind/brain of the individual and social factors—on the other. At issue is not whether anything at all can ever be said, usefully, about these “downstream” effects, but whether the notion of an E-language has any pivotal explanatory role to play in saying it (save as a useful shorthand)."

This is well put.

It is worth noting also that Barber and Ramirez explicitly recognize the challenges that idiolectal (or I-language-focused) approaches pose to traditional approaches in the philosophy of language.

"One apparent corollary of [Chomsky’s view of language] is significant for those many philosophers of language who have agonized over how to construct a theory of meaning for English. A common thought is that such a theory ought to take the form of a statement of the referential properties of the expressions of English—a link between words and objects in the world—from which the truth conditions of all English sentences can be derived (e.g., [Donald] Davidson, [Richard] Montague). Echoing P. F. Strawson, Chomsky suggests that referring is something people do. They use words in doing so, it is true, but referring is not something that words somehow do by themselves, through some fantastical medium, English. If referential properties of expressions amount to anything, rather than being relational properties between expressions and external objects (or “word-world” relations) they should be thought of as embodying instructions to the individual’s conceptual system, one of the performance systems with which the language faculty interfaces. If Chomsky is right, a great deal of the philosophy of language is either radically off beam or needs considerable re-interpretation."

I have barely scratched the surface here and don’t have a fully worked out position. But I am convinced that an idiolectal perspective has been and will continue to be extremely useful in the quest to develop a truer and more parsimonious account of human language.

The two Noam Chomskys

Noam Chomsky must enjoy making himself unpopular. His extreme (and extremely polarizing) political views are well-known.

I once sat down with one of his political tracts with a genuinely open mind, prepared to give it a go. But all I saw after devoting several hours to the book was anger and rhetorical posturing. I just couldn't figure out where he was coming from.

As I learned later, his attitudes are at least in part explained by his family background. Chomsky grew up in left-Zionist circles. His father was a distinguished Hebrew scholar and both his parents (his mother was more radical than his father, actually) were followers of the views of the essayist Asher Ginsberg. Writing under the name Ahad Ha'am, Ginsberg rejected purely political Zionism and promoted the idea of Jewish cultural and spiritual rebirth. By his early teens, Chomsky had embraced anarchism. He later identified with anarcho-syndicalism, working as an activist for various radical causes, often in association with radical Christians, whose spiritual and moral motivations for political action were similar to his own.

But what of the pioneering linguist? This side of Chomsky interests me, largely because I was taught by one of his students, and adopted many of his linguistic ideas. It matters to me whether (or to what extent) these ideas reflect reality.

So I have begun doing a bit of reading to see what Chomsky is currently saying and how this relates to the current state of research. This interview/article by Yarden Katz is a good place to start, though Chomsky's broad-brush references to intellectual history (Galileo is a big favorite of his) are not totally convincing. Linguistics is not physics, and it's conceivable that there is nothing there to understand in the way early physicists came to understand the principles of classical mechanics.

People generally draw a clear distinction between the two Chomskys, the political activist and the linguist. The former is generally characterized as progressive and radical and the latter, once radical, is now seen as conservative or even reactionary.

In fact Chomsky's scathing attacks on the trend to base research projects in linguistics and the cognitive sciences on Bayesian probability do make him sound like a bit of an intellectual reactionary. But the real issue is whether there is truth in his criticisms.

Bayesian probability is a topic I don't know enough about to write about, but this piece by a graduate student working with Bayesian methods in conjunction with traditional syntactic theory, seems very balanced and makes Chomsky's strictures on Bayesian approaches look a bit simplistic.

The main point I want to make, however, is that the two Chomskys may have more in common than meets the eye. One can see similarities in patterns of argumentation and thought between the political thinker and the linguist. One may also be able to trace some of Chomsky's basic convictions regarding the nature of human thought and language (as well as his political convictions) back to childhood influences.

If you read reports of his talks to linguists*, it's clear that Chomsky is deeply involved in the academic politics of research funding and concerned with the survival of linguistics as a distinct academic discipline, as well as with defending his status and reputation. These academic-political preoccupations (like any political preoccupations) encourage polarized thinking. What counts in the end is one's own side winning, not objective truth. (After all, the winners write the history books, including the intellectual histories.)

Whatever his motivations, Chomsky certainly exhibits a tendency to see things in terms of dichotomies, and is something of a past master of the straw man approach to dealing with challenges.

What, though, of the ideas that are being fought over? This, after all, is where the real interest lies. Does Chomsky's general view of life impinge on (and perhaps distort) his ideas on language and the mind?

The drivers of our thinking are always deep and obscure. Chomsky's longstanding moral (and, indeed, spiritual) preoccupations would, in my view, be likely to have had a profound influence on the way he sees the human mind, as well as reinforcing his views on the status of reason and intellectual intuition.**

I am really only starting to explore Chomsky's cultural and spiritual background, and I may return to these themes in the future and try to make a stronger case. There is much that remains obscure (the extent and nature of his secularism, for example).

I also need to do a bit of homework on some of the topics discussed in the interview. Frankly, I have sometimes found Chomsky's writings on language and thought, and Chomskyan linguistics in general, to be somewhat unclear or opaque, almost arbitrary in fact. I think this probably reflects Chomsky's commitment to a form of rationalism which is quite at odds with my fairly mundane empirical assumptions.

The interview is usually an easy form of discourse to follow and understand, ideal for introducing difficult thinkers to a wider audience, but Katz's interview with Chomsky remains – to me at least – obscure in parts. And I don't think Katz is to blame.

At first, I was confused by Chomsky's comments on Mendel. On the face of it, the case of Mendel argues for the power of statistical approaches, especially at a time when the basic science is undeveloped. But Chomsky's point – essentially that Mendel was aspiring to a deep understanding, and sought significance in the patterns he observed – is a fair one.

However, his arguments in favor of unification but against reduction in the sciences are less clear to me.

Chomsky's allusion to the case of chemistry not reducing to an older physics because the older physics was wrong seems – in the context of what he is arguing – a bit puzzling. Would not this example argue for having more tolerance for statistical and practical approaches which at least are dealing with reality rather than relying on prematurely postulated grand explanatory theories?

Chomsky himself says that cognitive science is at a primitive stage. 'Pre-Galilean', he calls it, but, as I said, I doubt that the comparison with classical mechanics is all that useful.

A more appropriate comparison for what Chomsky has been trying to do these past decades might be Einstein's doomed attempt during the last decades of his life to create a grand unified theory of physics.

Chomsky's thoughts on the origins of human language are very speculative. In fact, his account of a hypothetical individual in a group of non-thinking individuals 'getting language' (through a genetic mutation), and so being able to think, sounds quite far-fetched. (Chomsky, reasoning in a strangely a priori manner, sees language as an internal thing rather than being essentially communicational.)

There is, however, a lot of truth in what he says about science and intellectual fashion, and, yes, about language also.

I am aware that there are deep and serious questions about word order and context-free grammars and so on at issue here about which I have said nothing. Chomsky has made significant contributions to the application of formal language theory to linguistics, and influenced research directions profoundly. Just because other approaches may currently be in vogue does not mean that the work he inspired was misguided.

I suspect that, as our understanding of natural languages (and natural language processing) improves, many of the principles and insights developed by linguists working in the tradition he pioneered will be vindicated (and incorporated, one way or another, into truly effective natural language processing algorithms). But many of the key questions, both philosophical and practical, appear at this stage to remain unresolved.

Finally, a few thoughts on science and history.

Chomsky was asked by Katz about the importance of the philosophy of science and said it may be an interesting area but it doesn't contribute to science. What he considers valuable is the history of science. And he tries, as we have seen, to apply lessons from the history of science to emerging disciplines such as the cognitive sciences.

Though I am skeptical of some of the lessons he purports to derive, it's clear that a knowledge of the history of one's discipline – and the history of ideas in general – can allow one to put current research and current ideas into some kind of perspective.

Such knowledge is a part of the general culture a scientist might have, rather than a core component of his or her expertise. It's an optional extra, scientifically speaking.

Some people are just more interested in fitting their knowledge into a narrative than others, even sometimes preferring to learn their mathematics, physics, psychology or whatever in part as history.

Others have no interest in history or historical approaches which they see as a waste of time. And so they would be for them.

People have different ways of seeing things, that's all: different strengths, different ways of learning, different aspirations for understanding.

But the odd thing is, for all his talk about history, Chomsky strikes me as a basically – and profoundly – ahistorical thinker.

His fundamental insights within linguistics focus almost exclusively on the synchronic rather than the diachronic aspects of language, and aspire, in effect, to an abstract rationalism.

And look at the nature of his political work, which is perceived as extreme not just by conservatives but also by mainstream progressive thinkers. A true historical sense would have at least mitigated the free-floating and self-generating logic of his polemics.

In fact, you could make a case that his main concern with history is to mine it for debating points in order to advance his causes, defend his theories, and, by extension, to cement his own position in the narrative of science.

But it can't be denied that Chomsky still retains a certain aura, a certain iconic status. This is due, I think, not just to his achievements, but to an unrelenting seriousness, to a rare combination of intellect and passion.



* This hostile account of Chomsky's performance at an invitation-only event in London last year (by the distinguished linguist Geoffrey Pullum) is very revealing.

** It's worth mentioning in this context that, unlike most social scientists and, perhaps, curiously (given the rigorously abstract and scientistic tenor of his work in syntax and related areas), Chomsky takes literary art seriously and respects the value of the writer's insight into society's moral and psychological complexities. But, again, this becomes much less surprising in the light of his early exposure to (and continuing interest in and respect for) Hebrew literature.

Williams syndrome, language and the brain

In recent posts I have made a number of claims about language and the brain. Allow me to clarify and develop a couple of points.

I don't really want to buy into the debate about various versions of modularity or other theories of mental functioning. For one thing, I don't know the science well enough. I don't have a theory, but I don't know that I need one either.

Which is not to say that it is not important to have a basic understanding of how our minds work. My point is that such an understanding needn't take the form of a theory. It may simply develop from a general (or specialist) knowledge of pertinent disciplines (such as psychology or linguistics), and as a considered response to various kinds of evidence. I am particularly interested in the evidence provided by injuries and genetic disorders which affect cognitive and emotional functioning.

Certain genetically-caused disorders and brain injuries seem to provide evidence that language is in some sense a distinct system - or rather a set of systems - even if it interacts (as it obviously does) with non-linguistic processes. How else can you account for people who have a language deficit but can think well in other respects, or, conversely, who may be seriously cognitively impaired and yet maintain excellent language abilities?

Take Williams syndrome, for instance. It is a genetic disorder characterized by a range of medical problems, developmental delays and learning disabilities. Children with this condition seek interaction with others but are very vulnerable as they lack normal caution and social understanding. They are typically unable to cope with numbers and abstract reasoning. They also have impaired gross and fine motor skills.

On the positive side, they often have an affinity for music (and perfect pitch). And they also tend to do well linguistically, at least in certain respects.

Williams syndrome, like so many other conditions which impact on brain function, is selective in its effects. If specific aspects of thinking are adversely affected while other specific aspects are not affected or are enhanced, then this certainly supports the view that the brain consists of many (interacting) systems and sub-systems.

Linguists, of course, see language from various points of view corresponding to various sub-disciplines: phonetics (where the focus is on the actual sounds of language), phonology (more abstract), morphology and syntax, semantics, pragmatics, etc. In other words, language has many aspects, so it is misleading to talk about language ability without specifying exactly what one is talking about.

Likewise, it is not particularly helpful to talk about the brain's capacity for language per se. Better to focus on the particular processes which language use requires, like hearing (or seeing in the case of reading); interpreting the raw data (identifying phonemes and lexemes, parsing, etc.) and so understanding; or speaking (which involves not only mentation but also a very complex sequence of fine motor processes).

Children with Williams syndrome are typically slow to start speaking. This is presumably related at least in part to their fine motor problems. Most reference sources say that older children and adults with WS speak fluently and grammatically and have a good concrete, practical vocabulary (though abstract vocabulary remains deficient).

I picked Williams syndrome to focus on in this post because of an anecdotal report I remembered reading about a profoundly retarded girl with WS who nonetheless had an unusually extensive vocabulary and was able to invent strikingly original stories and fantasies. But the more I read about Williams syndrome the more complicated - and equivocal - the picture looks.

For example, consider this (from a recent research report* abstract): 'Williams syndrome (WS) is a neurodevelopmental genetic disorder, often referred [to] as being characterized by dissociation between verbal and non-verbal abilities, although a number of studies disputing this proposal is emerging.'

And in their own study the researchers found significantly more speech disfluencies (hesitations, repetitions, pauses) in the WS group than in a typically-developing group.

So the lesson of my story is that everything concerning the human brain is likely to be more complicated than it seems, and that only scientific findings - rather than models or theories - can give specific answers to specific questions. Of course, science requires its models and theories, but they are always provisional, a means to an end.

And, in the context of such reflections, it is hardly surprising that I find myself becoming more and more skeptical about certain Chomskian assumptions which have been part of my mental furniture since I took a linguistics course taught by one of the Master's protégés a couple of decades ago.



* Rossi, N.F. et al. 'Analysis of speech fluency in Williams syndrome.' Res. Dev. Disabil. 32(6) (2011): 2957-62.