Noam Chomsky – The biolinguistic turn lecture notes – part eleven

The phenomenal properties of these artifacts result from the interaction of invariant principles of the initial state (=the faculty of language) with a finite number of parameters fixed in one or another way. It would incidentally follow that there are only finitely many possible human languages apart from idiosyncrasy and choice of lexical items. And even these are sharply constrained. That means that the problem of unfeasible search is eliminate, its a major conclusion if correct. The conception has now been applied to typologically different languages of just about every known kind.

It lead to many discoveries, a host of new questions that where never before contemplated, sometimes suggested answers. This princibles and parameters approach is an approach. Its not a theory. Within the general approach there are many differse theories. There is actually a very good introduction to the topic by just published by Mark Baker – Atoms of language. He himself made major contributions to the approach. He is been working primarily on languages that appear to be at opposite ends of the spectrum of typological possibilities. Picking that on purpose of course. Mohawk and English, thats the pair he studies most intensifly trying to show that altough they are about as different phenomenally as two languages can be they are in reality virtuall identical apart from very small changes in a view parameters. Thake say a Marsian observer who views Humans as we do other organisms would conclude that they are essentialy identical. Dialectical varience of the same language.

There is been extensive work of a similar character carried out worldwide with quiet revealing results. One major programm funded by the European Union is studying the vast number of languages in Europe. Missleadingly called things like German and Italien and so on thoe the where totaly different languages. Included by this characterisations. And its beeing done elsewhere as well. I dont wanna suggest that the approach has been established. That is very far from true but it has been very succsessful as a research program. As a stimulus to empirical and theoretical inquiry. Progress towards the goals of descriptive and explanatory adequacy has far surpass anything that preceeds. Not only in depth of analysis of particular languages but also in the range of typological different languages that have been investigated and also new areas of linguistic structure that had barrely bin explored before.

Related field, such as the study of language aquisition, have also been completely revitalized within a similar framework. They now look totaly unlike anything that was around 20-30 years ago. There are some important steps towards convergence altho its certainly gonna be a long and difficult course even the approach turns out to be on the right track. We are far from having a clear idea of what the principles and parameters actually are. But I think it is fair to say that the study of language, in the last 20 years, has moved to an entirely new plane.

I wanna pick up these topics tomorrow and then move on to the issues, particlary the third factor – general properties of organisms. And then to move on the questions of intentionality. That is the question of how language now understood within the biolinguistic framework relates to the rest of the world.

Noam Chomsky – The biolinguistic turn lecture notes – part nine

The long term goal of investigating the third sector – that is the role of general properties of organisms in determining the faculty of language and the states it can attain (internal languages) – was actually formulated in the early days of the biolinguistic term but put aside as unfeasible. Attention focused on the first two factors, experience and the initial state in technical terminology = the problems of descriptive and explanetory adequacy.

The latter is how the initial state enters into determining the transition to the final state – the state attained. The earliest attempts, 50 years ago, where to replace traditional or structualist accounts of language by generative rule systems revealed very quickly that very little was known about the sound, meaning and structure of language and that huge problems had been unwittingly swept under the rug. Rather as in the days when it was assumed that bodies fall to their natural place, as has often been the case, one of hardest steps in development of the scientist is the first one. Namely to be puzzled by what seems so natural and obvious and to gain some realistic sense of what had been overlooked was an enourmous taks in itself.

Even more so in the light of the recognition that the apparent complexity and diversity of languages that was very soon discovered just had to be an illusion. The reason for that conclusion is a standard one in biology. Namely as in the case of other organs of the body, experience can play only a very limited role in determing the state thats attained. In this case the attained I – language even a young child has mastered a rich and highly articulary system of sound and meaning and structural properties that goes far beyond any evidence available and its shared with others who have different but also highly restricted experience. So it has to be the case that the initial state plays an overwhelming role in determining the language that the child attains in all of its aspects. Experience surly has a role in triggering and shaping role as in the case of other organs. But it has to be limited one.

So there is no reason to suppose that language and other higher mental faculties depart radically known in the biological world. The task was to show the apparent richness and complexity and diversity is in fact an illusion. That all languages are cast to the same mold and that experience serves only to set options within a fixed system of princibles all determend by the initial state. Which is the case of other biological systems.

Well. Great deal of research of the past 40 years in this areas has been driven by a kind of tension between descriptive and explanatory adequacy. That is the tension between the search for true theories of i-languages, the attained state on one hand, and the true theorie of the invariant initial state of the language organ on the other. The invariant initial state is the topic of whats come to be called “universal grammar”. It is adapting an traditional notion to a quiet a new context. The search for descriptive adequacy, like a true theory of Hungarian, that leads to complex intricate of particular construction and particular languages different from one another. In contrast the search for explanatory adequacy seeks to find the common ground from which the existing languages arise given the data that are structured as experience by the operations of the initial states. Again in some unknown manner.

The first proposals from the 1950s suggested that the initial state – the topic of universal gramma – provides a kind of a format for rule systems and organisations and a proceedure for selecting one instantiation of the format over another in terms of its succsess in capturing authentic linguistic generalizations and empirical notion that incorporates also a kind of a theory internal version of standard best theory considerations. The rules themselves, at the beginning, where adaptations of informal traditional notions which had proven to be utterly inadequate when they where subjected to close examination. So that meant, rules for forming relative clauses in Hungarian, or passives in japanese, or causatives in the romans languages.

The general approach did offer a kind of solution to the core problem of the study of language. Sometimes called in the literature the logical problem of language acquisition. That is how does the initial state map constructive experience to the final state. But as was emphasized, that solution holds only in princible. Because in practice the conception was unfeasible because of the astronomical compuational demands. Well, from about 40 years ago attempts where made to reduce the scale of the problem by seeking valid general prinicbles that can be abstracted from particular grammas and attributed to universal gramma, meaning to the initial state of the language faculty. Leaving a residue that might be more manageable.

Actually some of those proposals where kind of proposals that where then beeing explored and I reviewed in lectures here 35 years ago. After that time the considerable process took off but it still left the tension unresolved. That is the general picture was somehow fundamentally defective. There was no true solution, no feasible solution to the logical problem of language acquisition. A possible resolution of that tension was reached after a good deal of effort about 20 years ago with the crystallization of a picture of language. It marked a very sharp break from a long and rich tradition, tracing back to classical india and greece. Sometimes called the Prinicbles and Parameters approach that dispenses entirely with the core notions of traditional gramma, notions like gramatical construction, or grammatical rule. From this point of view, categories such as a relative clause or passive construction are understood to be real enough but only as taxonomic artifact. So, f.e aquatic organisms – which would include say dolphins, trouts, eels, and some bacteria. Its a category but not a biological category.

Noam Chomsky – The biolinguistic turn lecture notes – part eight

Notice again that that shift still leads us a long way from the problems of actions. Thats a vastly different matter. I have myself often quoted Wilhelm von Humboldt’s aphorism that the core problem of the study of language is the “infinite use of finite means”. It was a leading concern of Cartesian philosophy before him and a problem that really could not be posed until the mid 21th century when the concept of recursive generative proceedures was fully clarified. These proceedures constitue the finite means that are put to infinite use. But its important to be aware, I don’t think I stressed this enough, that despite quiet a lot of progress in understanding the means, means that are employed for infinite use, the question of how they are used is scarcely even addressed. And it was that question, that was the fundamental one for Descart, Humbold and other fairly modern figures. And again those questions are not even addressed for insects led alone humans.

Its reasonably clear that the human capacity for languages, whats called a species property, that is biologically isolated in essential respects and close to uniform accros the species. That actually seems less suprising today than it did not long ago in the light of very recent discoveries about the very limited genetic variation among humans as compared with other primates suggestes that we have all descended from a very small breething group maybe a hundred thousands years ago. Humans are basically identical from the point of view an outside Biologist looking at us. The biolinguistic approach adopted from the start what has been called, I quote the recently published encyclopedia of cognitive neuro science, “the norm these days in neuroscience, the modular view of learning, that is the conclusion that in all animals learning is based on specialised mechanisms, instincts to learn in specific ways” Randy Gallistel again. These organs within the brain perform specific kinds of computation in accordance with specific design appart from extremely hostil environments. The organs change state under the triggering and shaping effect of external factors. They do so more or less reflexively and in accordance with internal design. Thats the process of learning although growth might be a more appropriate term, avoiding misleading connotations of the term learning. The language organ, the faculty of language fits that normal pattern. According to the best theories we have, each attainable state of the system (i language) is a computational system that determines, generates in a technical sense infinately many expressions.

Each of this expressions is a store of information about sound and meaning which is accessed by performance systems. The properties of the I-language resold from the interplay of several factors. One factor is individual experience which selects among the options that are permitted by the initial state. A second factor is the initial state itself which is the product of evolution. And a third factor is general properties of organic systems. In this case computational systems incorporating and its reasonable to expect princibles of efficient computation.

The general picture involving crucially the third factor is familiar in the study of organic systems generally. The classic work of D’Arcy Thompson and Alan Turing on organic form and morphogenesis is an illustration topic currently in contemporary biology. One current example might be suggestive in the present context. There is recent work by Christopher Cherniak, Mathematical Biologist in Meryland, whos been exploring the idea that minimization of wire length – as in microchip design- shall best produce the best of all possible brains. And he has tried to explain in this terms the neuroanatomy of nematode – one of the simplest and best studied organisms. And also various pervasive properties of nervous systems. Such as the fact that the brain is as far forward as possible on the body axis. He wants to try to show thats just a property of efficient computation based on wire length minimization.

Well, one can trace interest in this third factor -general properties of organisms- back to a Galilean intuition, namely his concept that “nature is perfect” from the tide to the flight of bird. And its the task of the scientist to uncover in just what sense this is true. Newtons confidence that Nature must be very simple reflects the same intuition. However obscure it may be that intuition about what Ernst Haekel “Natures drive for the beautiful” has been a driving theme of modern science since its modern origin with the Galilean Revolution perhaps its defining characteristic.

It is hard to say exactly what it is, but that its a guiding intuition is not in doubt. Biologist however have tended to think rather differently about the objects of their inquiry. Very commonly they adopt what Francois Jacob, Nobel Laureate, image of nature is what he called a tinker – which does the best it can with the material at hand. Often a pretty rotten job as human intelligence seems to be keen on demonstrating about itself.

One well known contemporary Biologist, Gabriel Dover, British geneticist. He concludes in a recent book that “biology is a strange and messy business and perfection is the last word one can use to describe how organisms work particulary anything produced by natural selection.” Doe of course produced only in part by natural selection as he emphasizes, and this any biologist knows, and to an extend that cannot be quantified by available tools.

Well, we just dont know which of these conflicting intuition is more accurate – the galilean intuition or say Jacob’s intuition. And we will not know until we know the answer. And they seem very remote as answers. The same author, Gabriel Dover, writes that “we are nowhere near relieving our deepest ignorance about the biological world around us” he goes on to reserve his sharpest words ” for those who seek scientific respectability to complex behavioral phenomena in humans that we cannot even begin to investigate seriously”. He calls that “a sign of intellectual laziness at best and shameless ignorance at worst” when confronting issues of massive complexity which far exceeds the reach of contemporary science. He gives some exambles but for charity I ignore them.

Noam Chomsky – The biolinguistic turn lecture notes – part three

Well with the biolinguistic approach in place we want to discover the relationship between psychological states and the world as described in other terms. We want to know how computational states are related to neurophysiological states or represented in one terminology. We also want to find out how those mental states relate to the organism external world. As for example when the motions and noises produced by our forager bee direct others to a distanced flower or when I talked about a recent trip to india. Or when I say that I recently read Darwin’s Decent of Men but “Men” referring to a book. All of this is called intentionality in philosophical jargon.

The broad issuses were raised permanently at the end of the decade of the brain which brought the last millenia to a close. The American Academy of Arts and Sciences at the end of the millenium in the year 2000 published a volume to mark the occasion. It summarized the current state of understanding in these areas. The guiding theme of the volume was formulated by a distinguished neuroscientist Vernon Benjamin Mountcastle in the introduction to the collection. It is in his words “the thesis that things mental indeed minds are emergent properties of brains while these emergencies are not regarded as irreducible but are produced by principles that control the interaction between lower level events, principles we do not yet understand.” That same thesis has been put forth in recent years as a “astonishing hypothesis of the new biology” a “radical new idea in the philosophy of mind” the “bold assertion that mental phenomena are entirely natural and caused by neuro-physiological activities of the brain” opening the door to new and promising inquiry and so on.

Contributors to the American Academy Volume where for the most part quiet optimistic about the prospects about the remaining gaps between psychological and physiological accounts. Mountcastle’s phrase “we do not yet understand” reflect that optimism. Suggests we will soon understand. He wrote that “researchers speak confidently of a coming solution to the brain-mind problem” similar confidence has been expressed for half a century including announcements by prominent scientists, nobel price winner in one case, that the brain-mind problem has already been solved.

We may recall usefully similar optimism. Shortly before the unification of chemistry and physics, in 1929 Bertrand Russels who new the sciences well, he wrote that “chemical laws cannot at present be reduced to physical laws”. In his phrase “at present” like Mountcastle’s “yet” expresses the expectation that the reduction should take place in the course of scientific progress perhaps soon. Now in the case of physics and chemistry it never did take place. What happend was something different and totally unexpected, namely unification of a virtually unchanged chemistry with a radically revised physics. And its hardly necessary to stress the fact that the state of understanding and achievment in these areas, 50 – 80 years ago, was far beyond anything that can be claimed for the brain and cognitive sciences today. With outh to give us pause.

The American Academy Volume reviews many important discoveries but the leading thesis should arouse our skepticism. Not only for the reason that i just mention. Another reason is that the thesis is by no means new. In fact it was formulated in virtually the same words two centuries ago, late 18th century, by the eminent chemist Joseph Priestley. He wrote that “properties of mind arise from the organisation of the nervous-system itself and those properties termed mental are the result of the organic structure of the brain”. Just as matter is possessed of powers of attraction and repulsion that act as a distance contrary to the founding princibles of the modern scientific revolution from Galileo and Newton and beyond.

Noam Chomsky – The biolinguistic turn lecture notes – part two

Assuming that, I turn to some things that ought be obvious. It can scarcely be denied that some internal state is responsible for the fact that I speak and understand some variety of whats loosly called english but not say, hindi or korean. To borrow and in fact adapt a traditional term we can call this state, wherever it is, thats internal to me a state of the human faculty of language, primarily a state of the brain. We can call each such state an internalized language in technical literatur often called an I language. For simplicity I call it that. It should also be uncontroversial that the faculty of language has an initial state, part of our biological endownment, which permits a certain range of options – the attainable I languages.

The faculty of language then is a special property that enables my granddaughter but not her pet kitten or chimpanzee to attain a specific I language on exposure to appropriate data, data which her mind in some obscure way is able to extract from the glooming buzzing confusion and interpret as linguistic experience. That is no slight task. Nobody knows how its done, but it obviously is. More accurately every infant acquires a complex of such states, thats a complication error, but I put that aside. The expectation that language is like everything else in the organic world and therfor is based on a genetically determined initial state that distinguishes, say my granddaughter from my pets. That assumption has been called the innateness hypothesis. There is a substantial literature debating the validity of the innateness hypothesis. The literatur has a curious character. There are lots of condemnations of the hypothesis but its never formulated. And nobody defends it. Its alleged advocates, of whom I am one, have no idea what the hypothesis is. Everyone has some innateness hypothesis concerning language, at least everyone who is interested in the difference between an infant and say her pets.

Furthermore the invented term -innateness hypotheses- is completely meaningless. There is no specific innateness hypothesis rather there are various hypothesis about what might be the initial genetically determined state. These hypothesis are of course constantly changing as more is learned. That all should be obvious. Confusion about this matters has reached such extreme levels that it is becoming hard even to unravel, but I put this aside.

The biolinguistic approach takes mental faculties to be states of the organisms. In particular internal languages (I languges) are states of the faculty of language. I focus on language but most of what follows should hold as well for other cognitive faculties and in fact for far simpler organisms (bee communication or navigation). Well, when we adopt this approach several questions arise at once.

The central one is to determin the nature of the initial and attained states. And tho the matter appears to be controversal I know of no serious alternative to the thesis that these are in substantial measure computational states wether we have in mind insect navigation or what you and I are doing right now. Again, thats held to be controversal but since there is no alternative ideas I dont understand why. Its held to be controversal for humans. Its not held to be controversal for say insect navigation but the question is about the same.

Investigation of the brain in these terms is sometimes called psychological and its contrasted with investigation in terms of cells, chemical processes, electrical activity and so on that is called physiological. These are again terms of convenience, they dont have any sharp boundaries. Chemistry and Physics where distinguished in pretty much the similar way not very long ago. The formular involving complex molecules that we now study in school. These where pretty recently considered to be “merely classificatory symbols that summaries the observed course of the reaction. The ultimate nature of the molecular groupings was held to be unsolvable and the actual arrangements within a molecule, if this means anything, was never to be read into the formular”.

Kekulé whos structural chemistry paved the way to eventual unification of chemistry and physics. He doubted that absolute constitution of organic molecules could ever be given. His own models, his analysis of valency and so on where to have only an instrumental interpretation as calculating devices. Large parts of physics where understood in the same way by prominent scientists including the Molecular theory of gases, even Bohr’s modell of the atom. In fact, only a few years before physics and chemistry where united in Linus Pauling account on the chemical bond, Americas first nobel price winning chemist dismissed talk about the real nature of chemical bond as in his therms “metaphysical twaddle, this was nothing more than a very crueld method of representing certain known facts about chemical reactions, a mode of representation only” just a calculating device. The rejection of this skepticism by a few leading scientists, whos views where incidentally condemned as a conceptual absurdity, paved the way to the eventual unification.

This very recent debates, talking about the 1920s’ in the hard sciences, I think have considerable relevance for todays controversies in computational theories of cognitive capacity – thats from insects to humans. Important topic, one that I discussed little bit elsewhere, that deserves more attention than it recieves.

Noam Chomsky – The biolinguistic turn lecture notes – part one

Almost exactly 35 years ago I had the opportunity to give several lectures here, the same auditorium I think on the topic “language and mind”. And quiet a lot has been learned in the intervening years about language and the brain hence the mind, in the sense I used the term then, the term mind, mental and such terms.

Using these terms as just descriptive terms for certain aspects of the world. Pretty much on a par with such descriptive terms as chemical or optical, electrical and so on. These are terms used to focus attention on particular aspects of the world that seem to have a rather integrated character and to be worth considering for special investigation. But without any illusions that they “cut nature at the joints“.

In those earlier lectures I took for granted that human language can reasonable be studied as part of the world. Specifically as a property of the human organism, mostly the brain, and for convenience I keep to that. Both then and now, I am adopting what Lyle Jenkins called the BIOLINGUISTIC PERSPECTIVE. Thats the framework whithin which the approach to language that I am considering developed about 50 years ago. Also for convenience I use the term language to refer to human language. Thats a specific biological system. There is no meaningful question as to wether the communication system of bees or what might be taught to aps or mathematics or music are languages or wether airplanes really fly or submarines really swim or other wether computers think or translate languages or other comparably meaningless questions many of them based on a missinterpretation of an important paper by Alan Turing in 1950. Which respond a large and mostly misguided literature, despite Turings very explicit warning not to pursue that direction which has apparently been overlooked.

From the Biolinguistic perspective language is a component of human biology more or less a par with mammalian vision or insect navigation and other systems for which the best theories that have been deviced attribute computational capacity of some kind. Whats in informal usage sometimes called rule following f.e. a contemporary text on vision descripes the so called rigidity princible (was formulated about 50 years ago) as follows: “If possible the rules permit interpret image motions as projections of rigid motions in three dimensions.” In this case, later work provided substantial insights into the mental computations that seem to be involved when the visual system follows these rules in informal terminology. But even for simple organisms thats no slight task. Great many issues remain unresolved in these areas which are quiet obscure even for insects.

The decision to study language as part of the world, in this sense, should be in my view uncontroversial but it has not been. On the contrary. The assumption that this is legitimite enterprise was pretty forcefully rejected and continues to be rejected. Virtually all of contemporary philosophy of language and mind is based on rejection of this assumption. The same is true for what is called the computer model of mind. That underlays a good deal of theoretical cognitive science denied in this case not only for language but for mental faculties generally. Its explicitly denied in the technical, linguistic literature in what I call platonistic account of language and also in a different way denied in the conceptualism that is deviced by the same authors inaccurately attributed to many linguists including me.

It is also apparently denied by many sociolinguists, its incompatible with structural-behavioral approaches to language. Its, little to my surprise, rejected by current studies on language by leading neuro sciences. Most notably Terrence Deacon in recent work which has been favorably received by eminent biologist. The approach therfor seems to be controversal but I think the appearances are missleading. A more carefull look will show I think, that the basic assumptions are tacitly adopted even by those who strenuously reject them and indeed have to be adopted even for coherence.

I am put aside this interesting topic of contemporary intellectual history and I simply assume that lanugage can be studied as part of the world. I continue in other words to pursue the biolingusitic approach that took shape half a century ago, heavenly influenced by ethology, comparative psychology and intensifily pursued than along quiet a few different paths including much of the work that claims to reject the approach.