Professor Noam Chomsky - What Kind of Creatures are We?

all right well noam chomsky i think it's time for me to turn this over to you i really want to give you a a great big welcome from uh ali u of a and appreciate your coming today and presenting to our organization and uh uh we're truly an honor and a privilege to have you with us and uh i i know that a number of emails have gone out and i don't think any introduction is honestly needed because but uh can you hear me thank you for joining us and uh i'll let you take it over okay thank you very much uh i'd like to have a few words about the nature of language its acquisition origins its use and so far as we understand these matters today and i'd like to approach this within the broader context of the inquiry into language and mind which are the distinctive properties of the human species well a good place to start is at the beginning of our historical records classical greece the delphic oracle which produced the maxim know thyself well the two word aphorism is open to many interpretations the intended interpretation was individual you the individual should know yourself an unexamined life is not worth living the platonic expansion of this but it makes good sense to understand the maxim of the oracle collectively know ourselves the reason is that we're pretty much similar it's a very young species humans have only been around for two to three hundred thousand years that's uh barely an eye blink and evolutionary time it's been mentioned by evolutionary biologists partly as a joke but partly real that if you see two squirrels and a tree outside your window they very likely are more diverse genetically than all of the humans in the world humans have arisen within a very narrow window of evolutionary time and they have unique properties one of these properties is human language there's no analog no similar properties anywhere in the organic world furthermore it's a species property it's not only unique to humans but uniform among humans that means that any infant can acquire any language with equal facility as far as we know so it's a species property virtually definitive of humans there is another one thought at least thought in any form that we can grasp and understand the kind of thought that made it possible to produce the maxim of the oracle to interpret it to reflect on its significance myriad other mental activities that occupy us all day even in sleep uh just again a common human property totally unique to humans well when you have two unique species properties makes good sense to ask what relation holds between them the simplest relation would be simply identity language constructs thought thought is what is constructed by language and that is the conclusion that was reached 2500 years ago classical india classical greece and it's a tradition that runs through millennia late 19th century great american linguist william dwight and whitney simply defined language as audible thought it's basically thought which we happen to externalize through the mouth pick up through the ear we now know that that's too narrow you can use any other sensory motor system so [Music] if i don't know if this is being translated into sign but if it is you would be seeing the same thing in sign language which works very much the same as spoken language in the way it's acquired its structure the way it's used down to small details even the way it's normally stored well you can think of the language as being kind of like a program that's in your laptop so suppose you have a program in your laptop calculator for say multiplying numbers the program works by itself you can attach it to a printer it could be a printer that uses colors and dark letters and one font another font the program doesn't care it'll can be hooked up to sorry try to be louder it can be hooked up is this any better can be hooked up to any printer that you want the program stays the same and language seems very much like that there's an internal system constructing thoughts it can be hooked up to one sensory motor system or another to externalize it but the program seemed to be the same well let me make a side comment here on the history and then return to the mainstream of the discussion uh what i've just been describing is basically the long millennia-long tradition of the study of language and mind that goes right into the 20th century early 20th century it changed the general view in the early 20th century still largely surviving is that language is not primarily an instrument the means of thought but rather it's a system of communication that's sort of the mantra today in philosophy of language much of linguistics cognitive science this i think is a result of the rise of behaviorism in the 20th century which drove scientists [Music] philosophers others to look at what's what's visible what's apparent behavior to keep away from hidden things like what's going on in your mind that was a very sharp departure from the scientific tradition it's very anti-scientific in my view became quite dominant especially in the first half of the 20th century there's also a kind of a primitive version of darwinism which does have its roots in darwin's writings the idea that all changes through evolution must be small changes which operate through natural selection and over time may give large-scale differences it's now known that that's completely false but the residue of that does lead you to the belief the illusion in my view that language somehow must have emerged from simpler systems going back to animal communication the evidence is strongly against that and there's nothing in evolutionary biology that uh leads to such conclusions actually this is notice that this behaviorist turn is a turn away from the effort to understand and explain an effort to describe uh it's basically what you see in silicon valley linguistics the kind that makes the headlines about the things that it's claimed that machines can do associated with this was conception of how language is acquired dominant conception philosophy of language linguistics uh language is acquired through training in instituting habits if there's anything new that said by analogy all completely wrong completely refuted still very common well let's go back to the tradition which i think was on the right course the study of language took a large step forward with the 17th century scientific revolution the primary lesson of the scientific revolution was that you should be puzzled by simple things that you take for granted so a heavy ball of lid plainly falls faster than a light one except that it doesn't as galileo showed with elegant thought experiments take a moving through the sea suppose there's a a lead ball on the top of the mast if it falls it should fall behind the mast because the sailboat is moving forward except that it doesn't fall to the base of the mast again galileo showed this with thought experiments have you done an actual experiment the results would have been all over the place but as in much of science then and since the idea was to construct ideal situations often by thought sometimes by careful experiment if you can that bring out the hidden reality between behind the phenomena that you see which are always complex many variables can't be studied that's why scientists do experiments well if you go back to scholastic science in the 16th century but galileo and his contemporaries were struggling against scholastic science had answers to everything if a bull falls to the ground and steam rises it's because they're going to their natural place if two objects attract or repel each other it's because they have sympathies and antipathies if you perceive a triangle that's because the form of the triangle passes through the air and implants itself in your brain so there was an answer to everything all the answers were wrong they were based on what modern science called occult qualities invented concepts that don't exist and the willingness to be puzzled by things opened the door to the new science the science of the last several hundred years the scholastic science had answers to everything the new science had no answers to anything that's how serious inquiry begins and the same has been true of the study of language in the last 60 or 70 years breaking with the behaviorist structuralist tradition well let's take a look at language there are simple things that we take for granted we talk to each other we understand each other what could be simpler than that well to the new scientists galileo and his contemporaries this seemed an extraordinary puzzle amazing how can it be that with just a few symbols we can produce we can construct in our minds infinitely many thoughts never before expressed never in the history of the language and we can even convey to others who have no access to our minds we can convey to them the innermost working of our minds they regarded this as an amazing phenomenon in fact galileo himself regarded the alphabet as the most spectacular of human inventions because it facilitated this astonishing quality uh this and it is indeed a remarkable fact it's been repeated by others uh one famous case is gutlog frega a couple of centuries later the founder of modern logic and logical philosophy mid-19th century he wrote that he found quoting him he found it astonishing what language accomplishes with a few syllables it expresses a countless number of thoughts and even for a thought grasped for the first time by a human it provides a clothing in which it can be recognized by another to whom it is entirely new it's essentially galileo's insight and it raises very deep questions how can this possibly be the case uh it for uh this initiated a rich study of what was century study of what was called universal and rational grammar universal because it tried to cover all languages rational because it was interested in explanation not description very different from the structural linguistics and philosophy of language of the first half of the 20th century which described itself as a taxonomic science we have procedures of analysis you can apply them to data you come up with an organization of the data it's basically today's silicon valley that's doesn't care about understanding doesn't care about explanation just organization of data a radical change from traditional science including the science of language uh again the old tradition was picked up revived around mid 20th century well it does raise deep problems for descartes in the 17th century it was the basis for his establishment of the classical mind-body problem uh descartes accepted the basic idea of the new science galileo royal society in england later newton others that the world is basically a machine a complex machine the kind of machine that could be built in principle by a skilled artisan and that in their theology in fact was created by a super skilled artisan machine means something built out of ears and levers and pulleys and so on that's the world but descartes recognized that there are certain things that didn't seem to fall within [Music] machine science what was called at the time mechanical philosophy philosophy meant science so there were things that didn't fall within it one of the most striking ones descartes felt correctly is the creative aspect of language use what amazed galileo his contemporaries centuries later freyja how is this possible for a machine it is impossible for a machine and that led descartes a sensible scientist to postulate a new principle the thinking principle why that's the mind-body problem next task for descartes and other scientists was to see how they're unified well somebody's at the door my dogs are unhappy about it the uh this collapsed pretty quickly isaac newton's great discovery was that the world is not a machine there are no machines there are no bodies in the sense any sense of body that we have the world is just not intelligible to us that was the criterion of intelligibility john locke immediately recognized that as he put it within his theological framework just as god has attributed to whatever exists properties that we cannot conceive like interaction without contact so god may have super added to whatever there is the property of thought thought is just a property of organized matter whatever matter turns out to be all of this was worked out extensively through the 18th century then forgotten totally forgotten revived in the mid 20th century it's now become a standard idea after a long period of neglect well that leaves us with serious mysteries how is all of this possible and the answer to that question is we don't know we do not know how it is possible to answer galileo how it is possible for or fraga with a small number of symbols to construct new thoughts new to us maybe new in history to somehow transmit them to others in such a way that they can interpret the inner workings of our minds a lot of this is just um unintelligible to us well there are parts of it that we can understand here we have to make a distinction which was actually discussed by aristotle but then forgotten until the 20th century the distinction is between possession of knowledge and use of knowledge it's a crucial distinction in modern technical terminology for those of you familiar with that that's the distinction between what's called competence and performance what we know and what we use the internal going the language the internal language provides us with a store of an infinite number of possible thoughts in our that's our possession of knowledge our use of knowledge is perception and production when you perceive the noises that i'm making you are accessing your internal store of knowledge to provide an interpretation for these noises that you're hearing when i produce a sentence i'm carrying out two acts the first act is selecting a thought from the store of possible thoughts the second act is implementing it using the various mechanisms of my mind my articulatory system to externalize that internal thought the second part of it we can study and understand the first part of it selecting the thought is completely beyond our understanding we have no idea how this takes place it's part of the general mystery of voluntary action which is basically to us an impenetrable mystery well just keeping to what we can hope to understand uh the system of language that constructs thoughts the formation of the internal language uh keeping to that uh we can discover some quite remarkable properties so following the tradition of early modern science let's take simple cases so take the sentence the bombing of the cities is a crime notice it is a crime not or a crime actually that's puzzling when you think about it why doesn't the verb is agree with the closest noun phrase the closest noun phrase is cities so why don't we say the bombing of the cities are a crime okay let's take the bombings of the city is a crime or a crime the bombings of the city are a crime why don't we say the bombings of the city is a crime again just using the simple problem for property of adjacency when you look at it you see that what humans are doing is taking a is ignoring the simplest possible computational procedure adjacency and instead using a complex procedure finding not looking just at the linear order of words but finding a structure and then finding the central element within that structure turns out to be pretty complicated computation but that's what we do reflexively and in fact that's what every infant does you can show by experiment that this is what infants do down to before the age of two 30 months of age i can't obviously can't and this is true of all constructions in all languages to give another example a little more complicated take the sentence uh the man who fixed the core carefully packed his tools the man who fixed the car packed carefully packed his tools notice it's ambiguous he could fix the car carefully where he could carefully pack his tools now put the adverb carefully in the front of the sentence carefully the man who fixed the core packed his tools unambiguous furthermore it means carefully practice tools and it's unambiguous in a puzzling way carefully looks for a verb but it doesn't look for the closest verb it looks for the more remote one which if you build up the structure is the structurally closest one so what you're doing is what everyone is doing reflexively is ignoring the linear order of words and attending only to structures that you never hear and that your mind creates to give one last example takes a sentence the man who saw bill is young suppose i didn't hear the word bill clearly i say i want to ask who who it is so i say who is the man who saw young gibberish can't say that fine thought you can't say it you have to you can only say you can't use the simplest computational procedure and find the closest occurrence of who and put it in the front it's blocked well this is universal as i say children of 30 months old know it all constructions all languages what it means is the infant all infants are ignoring 100 of what they hear which is worse in linear order they're ignoring very simple computations on linear order and they are reflexively paying attention to something they never hear that their minds create and using computations on those abstract structures it's a pretty remarkable fact when you think about it and that is the kind of thing we discover as soon as you try to answer the galilean challenge this property is called structure dependence in the technical literature pretty much the same is true of learning of words there is a myth standard myth that the way children learn words is they see a cat their mother says cat they make an association between the thing they see and the sound your mother produced and this is repeated over and over again in many circumstances different situations finally the kid figures out that the sound cat goes along with that feline object totally false nothing like that happens it's very careful experimentation on this by now the primary researcher and this is wonderful scientist lyla gleitmann who old friend who passed away a couple of weeks ago what lily lightman showed in her experiments is that on the is the children pick up words and the meaning of words on one or two presentations they at the peak period of language learning around two years old children are picking up words at about one every waking hour which means that they hear them once maybe twice they usually don't pay attention to their mothers just whatever is going on in the circumstances and they know the meanings of the words and the meaning of the word when you look at it is very complex no time to go into this but when you really look at the meanings of words you find out that they are quite rich well actually let me mention a few examples from classical greece where this matter was discussed so take aristotle again he took us his example the word house he said what is a house what's the meaning of the word house and he said well has two components one is what he called matter it's the bricks the timbers and the physical things that the house is made up of the other is what he called form an idea in the mind the characteristic way the object is used the thought and the mind of the architect to construct it the way we conceive of it so something could look exactly like a house physically but it could be a library if that's the way it's used it could be a garage it could be a stable could actually be a paperweight for a giant could be whatever we conceive of it whatever the architect had in mind that as a consequence and that's quite correct it means that a physicist looking at the object couldn't know the meaning of the word that means that the child when it's when it's learning the meaning of the word is using all of this intricate understanding in the mind none of it presented as evidence furthermore that's true of every word in the language let's take an earlier example pre-socratic heraclitus asked a rather profound question how can you cross the same river twice second time you cross it it's a physically different object in fact it could be radically different it could be flowing in the opposite direction it could be not water but the arsenic from an upstream plant it could have been divided into several tributaries it could be like the river i pass on the way to work here in tucson where i live the grelito river which never has any water in it it's a dry bed but it's the relito river because every once in a while you get a trickle of water through it and historically it was a river so it's a river it's all the way we conceived of it you could take the relito river uh start using it for commuting to tucson would be a highway not the river so physically it hasn't changed at all all and that's again true of all the words in our vocabulary one special case of it is called polysemy suppose i say the book is harder to understand than it is to burn what's the book it's simultaneously abstract because you can understand it and concrete because you can burn it the world doesn't contain objects that are both concrete and our abstract but our mind constructs them and that's what the words of language are and these are learned virtually without experience another this one is a real mystery we don't know how that's done but it's a fact that has to be explained well these are among the amazing things that you discover as soon as you begin to look at the facts closely one conclusion we have to draw is that language what we call language has two completely separate components one of them is an internal language that constructs thoughts has no order no linear order just abstract structures the second part is a system that externalizes it into one or another sensory motor medium usually sound could be signed could even be touch notice that that second component is an amalgam of language and sensory motor systems it's not pure language sensory motor systems have nothing to do with language they were in place long before language emerged they have not changed since language was emerged in the evolutionary record so there's something extraneous to language kind of like the printer that your laptop could be attached to uh well and that of course has linear order and does use adjacency and other things well we when you try to find explanations for language you real explanations you face a kind of conundrum uh a child learns language with virtually no evidence actual statistical studies of the data available to children sure that it's very sparse very little and as i mentioned they know things for which they have no evidence at all like structure dependence so that makes it seem as though what's innate internal to the mind must be very rich in order to carry the child from extremely limited data to very rich possession of knowledge so on the one hand the learning the internal structure must be very rich but then there's another problem this system had to evolve well it evolved in a very brief period of time we now know from genomic analysis that human beings began to separate at least 150 000 years ago and they all have the same faculty of language same possession of knowledge of language same possession of knowledge so it was in place before 150 000 years ago well that's not long after humans emerged as i said humans emerged maybe two to three hundred thousand years ago that's a very narrow window in evolutionary time and a click of an eye before humans emerged there's no evidence in the archaeological record for any kind of symbolic activity after humans emerged pretty soon in evolutionary time yet very rich symbolic activity ordinarily assumed plausibly that this is associated with the emergence of language well what that seems to mean is that the basic faculty of language must be quite simple since it emerged so quickly in evolutionary time but now we have a conundrum looking at acquisition looks like the internal innate system must be very rich looking at evolution seems like it must be very simple i should say the first problem acquisition is sometimes called plato's problem plato did raise the question how can we know so much with so little evidence didn't really have an answer the second question is sometimes called dorwan's problem how could we evolve such a system so quickly so it must be both very complex and very simple there's a third problem whatever the faculty of language is that has to be universal has to cover all languages because an infant has equal access to all of them well it's this collection of problems that has guided the inquiry into language for the last 6 years 70 years since the abandonment of the behaviorist structural restrictions and the return to what had been the tradition and it seems that finally in the last few years we are getting to the point where we may have answers to these questions so uh that if it worked to the extent it works out is actually a new era in the uh long millennia-long study of language the first time when we can reach genuine explanations satisfying answering plato's problem darwin's problem and the universality of the theory so let's look first at the universality uh it seems recent research is increasingly attending towards the conclusion that the diversity of language in the externalization in superficial things like choice of flexible the way you uh the sound you associate with the lexical item so it could be uh you know cat in english and french something else in some other language but that's superficial that of course you can learn easily but the concept the meaning of the word that's already inside so one aspect of variety is just association of sound and symbol that's easy there are others there are many the variety of languages looks on the surface very large so there are some languages english happens to be what's called a highly analytic language of lots of simple words from one after the other to make a sentence there are other languages at the opposite extreme a sentence could be one word and other words could be scattered around freely well it's by now been learned over recent years that these apparently radically different languages are almost identical at the deeper level they have the same internals very much the same maybe exactly the same internal structure the internal language the one that generates thoughts looks as if it's pretty much the same and exactly the same no matter how different the superficial appearance actually that's not too surprising the internal language can't be learned we have absolutely almost no evidence for it it's true of the meaning of words through things like structure dependence other major properties of language so it makes sense for it to be identical among people or close to that on the other hand the externalization is not strictly language remember it's an amalgam of language and sensory motor systems which are totally divorced from language well connecting these two things is a fairly complex process two things that have no relation to one another can be done in many different ways it's basically solving a hard cognitive problem uh different communities have solved the problem in different ways so it looks complex that's a complex problem the internal language looks quite simple because it's based on very simple computational procedures so on the one hand we solve the problem of darwin's problem we have a very simple system to solve the problem of the plato's problem because the learning is all in the superficial part the internal part just isn't learned it's just there like structure dependence the ways of interpreting those sentences that i gave and the variety is where you expect it to be in the complex problem of relating the internal language that produces thought to the external system that translates that hands it over to the printer that allows other people to read and see and interpret it if that to the extent that that can work out you have genuine explanations well that carries us on to the next chapter how has it worked out can only give the barest hint of that here so let's take the major principle of language structure dependence pretty remarkable principle wasn't even noticed until fairly recently in all the whole history of language again the fact that all operations of the internal language not externalization all operations of the internal language that generates thought ignore linear order there's no linear order there at all uh just abstract structures created by the mind and somatic interpretations like the ones for river house book every word you can think of which are constructed by the mine uh so that's uh let's take structure dependence how can we account for this we have to show that when language evolved it mother nature hit upon the simplest possible combinatorial operation the basic property of language all languages is that each language is what you know what the knowledge possessed is an infinite array of sentences which are structured expressions hierarchy of structures and can somehow be assigned an interpretation as a thought and by the externalization system mapped onto the printer sensory motor organs that's the basic property it's a property called recursive generation you have to generate an infinite array of expressions well the understanding of how to do that developed really in the early part of the 20th century as the theory of computation was developed by leading mathematicians kurt girdle alan turing alonzo church and we'll post establish the modern theory of computation which shows clearly how a finite object like your laptop or your brain can store within it programs that can yield an infinite array of objects expressions in our case and do it in a way which could be hooked up to a printer that allows us to ask what are the operations that mother nature would have hit upon as soon as some accident took place in the evolutionary record a couple hundred thousand years ago and the property of recursive generation somehow appeared what's the simplest way to deal with it well the simplest computation happens to be what's called binary set formation forming a set out of two elements not changing either of them you repeat that over and over again you get hierarchic structures uh indefinitely so a plausible evolutionary scenario is some accident took place maybe some small mutation out of it came a process of recursive generation mother nature then picked the simplest one binary set formation was called merge in the recent literature and then everything flowed from that structure dependence flows from it immediately many other properties do as well where you can show that that's the case you have a genuine explanation notice incidentally that this is the way evolution works generally evolution goes through all gen evolution goes through three uh i see uh half of the question but i can't read it uh okay could you follow me just hold it for a second we'll come later okay yeah the uh the way evolution works is three steps first some accident takes place could be a mutation cosmic ray passed by and changed the dna slightly uh an accident it could be what's called symbiosis so if you go back a couple hundred million years it turned out that a bacterium there were only bacteria and similar microorganisms at the time one bacterium by accident swallowed another microorganism that led to the development of what are called eukaryotic cells complex cells the basis for complex life our life for example the third step is winnowing of the various things constructed in the second step which ones have better reproductive capacity that's natural selection those are the ones that uh survive so it seems to be the way language developed first came some accident maybe imitation which yielded recursive generation then mother nature comes along tries to find the most elegant solution to the problem of how to deal with these new systems in one case it was eukaryotic cells in our case it was a generation by merge the winning stage apparently never came for language there's no winnowing stage where we stayed the same maybe because time just hasn't been long enough for natural selection to operate maybe it's because at the second elegant stage the system was so closely integrated that you just can't change it or it falls apart that would be a very interesting discovery and research is tending in the direction of showing that it's correct well like time is running out let me stop here we're now reaching what would be the next chapter how does all this work out in detail but i think if this succeeded you have a general picture of how we can hope to meet the galilean challenge the basic challenge in the history of the study of language and mind and do it in a way which can give genuine explanations for some quite remarkable properties of language and thought while leaving identifying other properties as a total mystery which we simply cannot comprehend that seems to me pretty much where we stand today i'll stop there thank you gnome uh can you hear me okay great so um we'll have actually uh ask people to raise their hands for questions i think it'll be easier to moderate them but we did have a couple come in through chat and maybe we can start there leslie bailey if you want to unmute yourself and ask your question go ahead oh great okay i'm afraid this is a really really stupid question gnome i'm sorry um if possible please restate the difference between the behavioral structural approach to language and what replaced it and how that transition occurred i mean i know it's very complicated but if you could just say it in two sentences it's quite let's take a look at how the structural linguists identified their own theory their words structural linguistics was identified mid 20th century a consensus had been reached among the at that time a fairly small number of structural linguists it was a very small field then it's a huge field now the small they reached the consensus that structural linguistics was what they called a taxonomic science there were procedures of analysis i was actually a student in the late 40s studying this you studied the procedures of analysis here's what you do when you have an informant from a say native american community here are the procedures you use if all you have is a text here's the procedures of analysis you use those procedures identify units you organize the units you're done it's a taxonomic science field's over okay so it's descriptive let me continue there was also a theory of learning it was stated explicitly by leonard bloomfield the leading linguist american linguist to the early 20th century language in his words is a matter of training and habit if there's anything new produced it's by analogy same view in philosophy of language take the major figures w v quine lyric wittenstein essentially the same language is a matter of training fabric habit your coin schenerian operant conditioning anything new is analogy the new system that developed starting around the 20th century was essentially a return to the old tradition we want to find explanations for linguistic phenomena we want to find out why they work this way not the other way why do we say the bombing of the cities is a crime not or a crime essentially the 17th century questions as soon as you started on that you had the same transition that happened in the 17th century from knowing everything to knowing nothing as soon as you started asking these questions turns out he had no answer well i should say that the entire tradition was totally unknown completely forgotten wasn't rediscovered until the 1960s after much of the ground had already been laid the structuralist behaviorist era essentially wiped out the history was gone but the new system that i've been discussing is essentially picking up from what the tradition was i don't know if that's clear enough to answer the question i no that's great thank you so much all right so our next question is from david dalton david i've just unmuted you so you should be able to ask your question um listen to our discussion here and i ask is not our consideration a little bit anglo-centric in that english yes it has an alphabet of 26 characters but if you look at mandarin it has an alphabet of 40 000 characters only 2 000 of which are extant today not to mention that mandarin is a tonal language whereas english is more of a literal language and i don't know mandarin myself but i just wonder how is it that we can with such broad brush strokes compare all of language in this discussion i would seem that there would be quite a bit of difference in the semantics of mandarin versus english or say german or french well i think the comment about the alphabet that i made was galileo he was of course thinking of latin and greek but the alphabet is a very superficial matter it's like the printer you can use all sorts of alphabets you can use the mandarin style the korean style hieroglyphs just as you can use any printer turns out that the internal system for mandarin for english for yokuts for potawatomi for tagalog the internal systems are very much the same maybe identical lots of languages or tonal languages but they function essentially that's just a different printer it's uh and i i mentioned just one example when i was talking languages like yogurts is an example where a single word includes the whole sentence it was thought for a long time that those languages must be very different the more we discover the more we find that at the internal level they're the same the internal means for constructing expressions and the semantics is either identical or very close to it doesn't look like that on the surface you're quite right on the surface they look wildly different and it was assumed if you go back 50 years it was assumed that languages can differ in just about every possible way and when you study a particular language you can't use any presuppositions from another language that was something sometimes called the bowasian principle that's what i learned when i was a student seems to be completely wrong it seems that at the core all languages are either identical or very close to it the systems that construct internal expressions in the mind expressions of thought don't seem to differ uh the externalization is very it looks very different they seem arranged differently different sounds so on but all of that seems superficial to go back to the analogy of the laptop in the printer what's the internal program in the laptop seems pretty close to common for all languages that we know including very diverse ones the printer can be very different and the alphabet isn't even the printer it's not the sounds it's a representation of the sound so it's even more remote from language our next question is from fran manley fran if you want to unmute yourself or i've on you did you actually fran you're good to go yeah i got a two-part question professor one is there any evidence as to whether or not other species for example like dogs have a common language which which allows them to communicate even with dogs they don't know and my second question if i may is on children i've been led to believe that children don't use a word unless they've heard it but i got the impression from what you said that's not necessarily so they don't have to hear a word to learn it okay thank you well on the second one children have to hear a word in order to not not to it depends what you mean by learn the word they know the meaning of the word before they hear what the sounds are so if your chinese child child a french child whatever it may be you have the meaning in your head you don't know what sound is going to be associated with it that you have to hear and uh a lot of lightman and her work and others have shown that it takes almost no no it no experience coupled hear it a couple of times and you say okay that's the internal word that i know with all of its complex meaning you do have to hear something to know whether it's pronounced cat or some other way okay how about the animals do they have a language form of language other species do they have more did they have do you like do animals or do other species have does for communication yeah so animals let's say dogs i happen to have two of them right here on the desk at the moment you heard a couple of them before well i've looked at them pretty carefully they're they have a number of ideas in their heads maybe a dozen or so sometimes they can do moderately complicated things like i'll give you my favorite example is a big dog and a small dog if the big dog wants to be taken i hate to say the words because unfortunately you'll hear them but wants to be extricated to the outdoors paraphrase which she doesn't understand when she wants that she has a trick to do it she takes something away from the small dog and the small dog comes over to us to complain and we make the big dog give it say a toy give it back to the small dog and then she runs right to the door and expects us to extricate her okay so it's more of a behavior than a language right it's a small number of things it's an entirely different from human language in every respect every animal that's known including the closest to us apes monkeys symbols like take a vervet monkey has been studied carefully has uh maybe a dozen calls like if the leaves are fluttering in a tree the vervet monkey will come out with a sound which makes all the other monkeys run away and hide we interpret that as a warning sound maybe there's an eagle in the tree or something i don't know what the monkey's thinking but every or can make another symbol which is we interpret as meaning i'm hungry okay uh the point is that each symbol is keyed explicitly to a physically identifiable entity motion of leaves in a tree hormonal changes something like that okay thank you and none of the none of the human symbols are like that that's aristotle's point about house so they're radically different furthermore there's no combinatorial structure in the animal systems it's just a symbol now there have been major efforts to try to train chimpanzees like infants the main study you may have seen it is called the nem study named after me actually by a number of very fine cognitive scientists some of them former students now one of them a colleague and friend they made an extensive effort to raise the chimpanzee nym from infancy pretty much the way you raise an infant extensive efforts to get nim to pick up something like language they used sign because the articulatory system for the chimpanzee doesn't work very well so train it like an infant could learn sign at first they thought they were getting somewhere but when they looked closely it turned out it was nothing it just it's just beyond the capacity of a highly intelligent ape to do anything like what a one and two-year-old infant does they're just built differently there are things that other organisms do that we can't do like i live in the desert in the back of my house there are desert ants who have cognitive capacities that i can't come close to they can navigate in ways which i can't possibly do maybe humans can duplicate it with complicated instruments and they have a brain about the size of a minuscule brain organisms are just different our difference is we have the capacity for language doesn't have any counterpart anywhere no our next question uh is from les i've unmuted you less so go ahead and ask yeah this question is uh it's maybe more philosophical but the idea about uh thought and expression can your thoughts before expression internally or do they rise simultaneously and cannot exist independently so these are things where there's no we can't introspect because it's all beyond the level of introspection and there's no external scientific evidence but the simplest theory the simplest assumption the one that would have to be disprove is that they're just the same thing the expression is the thought there's an internal system which has provided we all have the knowledge we possess includes an infinite array of these thoughts formulated in the linguistic system when i produce a sentence i pick out one of them somehow how is unknown then i implement it so there may be no distinction between thought and expression no first or second just that's what the thoughts are there's a repository of thoughts in linguistic form and we can then implement them and on the receptive side you can reconstruct them maybe not the same way maybe even a different language but something very much like them our next question is joe keller joe i've unneeded you thank you scott professor as the newly conceived human begins to develop her that inborn language structure will her external experiences add to or enhance that structure in any way well here we ought to pay attention to the way built-in structures work generally to take your visual system the visual system does quite complicated things very this time talk about them but very rich things if your visual if your visual system is not stimulated in the early weeks of life it'll degenerate we know this from studies with other animals cats monkeys and so on and basis that we live about the same visual system it's shown that unless a kitten is presented with pattern stimulation in the first couple of weeks of life it'll be blind well that's the way and all instinctive behavior works like that it has to be triggered by something and then it is partially shaped partially shaped by the way by the stimulation so take kittens again if kittens are presented in the early weeks of life only with horizontal lines then in as a mature cat it'll be able to perceive horizontal lines not vertical lines uh that's a little bit like different learning different languages you can shape the internal system marginally you can't train a kitten to recognize uh curves for example in infancy has to be lines in different orientations uh lots of things you can't get it to do and that's the way instinctive behavior works quite generally and it seems to be the same with language as far as we know the system is built in the internal system has to be triggered in order to get started to develop then pretty much develops on its own course there are some respects in which experience outside data shapes the system one is the trivial one of how do you pronounce this uh others or are you going to be like english an analytic language with a lot of separate words that form a sentence or are you going to be like yokuts let's say a language where you have one huge word that includes everything well that requires some stimulation but and a large part of recent research very recent research is to try to sort out which parts are fixed and fundamental which parts are provided by the shaping expect of experience and it's turned out pretty surprising so 40 or 50 years ago the greatest anthropological linguists friends of all of us here ken hale former teachers friends one of the leading anthropological and formal linguists of the past generation he firmly believed that languages differed in what's called a parameter an option of choices some had flat structure with no hierarchy some like english head hierarchy and it looked that's the way the evidence looked over time actually ken ken hale and his students julie legging were able to show that even the most extreme cases of languages that looked like completely free word order or flat no structure even they had the same internal structure as the hierarchy languages these are real discoveries major discoveries over the past couple of decades and we don't know how far it'll go but i think it's a reasonable guess by now that it'll probably go to showing that the internal system that generates thought is either uniform or very close to it and that the apparent variety and complexity and mutability the change from generation to generation all of that's probably in the externalization and the trivial lexical properties like do you pronounce it cat or some other way our next question is miriam burt miriam you are unmuted go ahead thank you i think my question professor probably follows on pretty well to the recent one you the one you've just responded to you said earlier um that around two years to 30 months is this riotous wonderful time when you know they're learning a new word or a news every every waking hours but i wonder if a child if she's deprived of both that visual stimulation and extended language you know the auditory to hear the words from because she's in an orphanage or you know in a very poverty state if you will you know for external stimuli will she be able to catch up if she is subsequently you know uh exposed to the this this stimuli you know internally and extern i mean visually and auditorially or is she always going to be a little behind and not quite up to what she could be thank you that's a very it's a very difficult question because we cannot do experiments you don't do experiments with humans rightly or wrongly we do do them with cats and monkeys that's how we know about the visual system which is pretty much shared with us can't do it with any other organism because no other organism anything will remotely like language so we just have to look at the experiences that we see and they're they're very interesting uh you learn a lot from them so there's a very fine neuroscientist named helen neville oregon state university who done fine work on the brain and language about 30 years ago she got interested in pretty much this question she took children who are what are called low ses socioeconomic status and who seemed to be problems in school they weren't learning they couldn't speak all kind of behavioral and other problems and she began to study them carefully she noticed some interesting things their parents don't talk to them i mean they say get out or go outside or leave me alone and not because they're bad people they're just people are trying to survive on two jobs and miserable circumstances parents never read them stories you know they've never had that experience so she tried a simple experiment trying to bring the parents and the children together and encourage them just in some free time to interact like have a mother read a story to the children things like that turned out they enjoyed it they did more of it they started having much higher cognitive achievements more problems in schools and so on well that's partial answer to your question i'll mention another real experiment from a real event there's a very fine another cognitive neuroscientist who tom and massimo know very well he lives in massachusetts he began working with children who were such extreme behavior problems that they were basically excluded from school they couldn't stay in a classroom had to be excluded we started working with them looking at them carefully turned out he immediately found out that these were kids who came from homes where they didn't eat breakfast they didn't eat breakfast they got on a school bus the bus roamed around for an hour they finally got to the school they were out of their heads they were sent to it maybe an arithmetic class they couldn't pay any attention they ran around the room and so on well he started doing something for one thing when they came to school the first thing he did was give them candy because their glucose levels were low they hadn't eaten then instead of putting them right in an arithmetic class he let them stay outside and run around for half an hour then they brought them into school other things like that by now these kids are performing better considerably better than the average in the public school system well i think that's the kind of evidence that we have about your question it's a humanly very significant question we can't do direct experiments on it we're not going to deprive children and see what happens you take the natural experiments and see what you can find out and i think the basic answer is that if the deprivation is not too severe the child can probably recover normal cognitive capacity if the deprivation is extremely severe and there are such cases a child is simply deprived of external contact maybe a sociopathic father something like that then they may never recover but it's hard to know what's cause there's a famous case genie which you may have read about but it's hard to know what the lack of recovery is from because any kid like that is going to be psychotic if you're brought up tied to a chair until you're 12 years old with somebody throwing food at you you're going to be psychotic if you can even survive so how much is that interfering with what's happening well we don't know and we don't want to do the experiments which will answer it we've got about five minutes left our next question is from linda green linda you are unmuted go ahead um my question has to do with brain research what contemporary brain research is uh shedding light on linguistics well it's it's difficult because brain research altogether is very difficult and brain research on humans is particularly difficult because we do not allow invasive experiments with a cat we allow ourselves to stick an electrode into a neuron and see what it's doing they don't do that with humans so it's all kind of kind of evidence you can pick up by external imaging and so on and you can learn a great deal in fact the most important discovery about this actually has to do with what i call structure dependence the fact that the computational operations of language make no use of a hundred percent of what we hear words in linear order they only deal with abstract structures in the mind they're very ingenious experiments initiated by a friend and colleague of ours andre amoro italian linguist neuroscientist he and his colleagues devised experiments on the following paradigm they present took subjects two groups of subjects who happened to be speakers of german and presented them with invented systems one of them modeled on a japanese language that they'd never heard another using simple rules which violated structure dependency so a rule that says the if you want to negate a sentence take the negative particle say not and make it the third word of the sentence very trivial computation they asked how subjects responded to this well it turned out that when they were given a language modeled on a natural language the areas of the brain that are mostly involved in language processing and use all fired up the way you'd expect them to when they were given the system the artificial system with simple computations on linear order they just got diffuse activity in the brain meaning the brain is treating it as a puzzle not a language okay that's quite an interesting result it's been replicated there's different variants of it and so on but that's the kind of thing that can be done there are other there is other experimental work in neuro linguistics which shows a number of things all right we uh probably have one quick question ralph lofty if you'd like to uh i've got you unmuted ralph if you'd like to ask your question yes uh i just read recently about the multi-language brain and it looks like people with multi-language capabilities that actually perform worse than uh mono language brain and and and how is how could that be when they let you said the languages of the universe so the thoughts should be universal and they should express okay yeah well first i don't know what you saw but i don't believe it uh every child is easily capable of learning many languages in infancy it's very common in many cultures for a child to know three or four languages they don't even know they're different languages this is the way you talk to your mother this is the way you talk to your grandfather this is the way you talk to the kids in the street finally at some age maybe four or five they may realize hey i'm speaking different languages but it's just normal as far as we know every child can do that there's no known difference between there's no even any way of studying it because all children have this multi-language capacity but i'd be very skeptical about the source you looked at frankly okay thank you so much yes well i think that's all the time we have today that professor chomsky we really appreciate having you present to us and i'm grateful to know as i when i was a child and i probably begged for candy for breakfast that it wasn't a bad thing but uh no i we really appreciate you being here and again thank the department of linguistics at the university of arizona for hosting this uh session and course and uh appreciate it so much so do you have any other parting words for us sir i hope some of you will be interested enough to look into the meat of the subject how it's actually done the next chapter we we did have uh a few people inquiring about the publications and things that would be a good service to kind of get a general understanding of this area of study what what are your recommendations well i have a book called what kind of creatures are we which we've been to many of these questions but there's a lot more work they're very i should say they're excellent general introductions to language there's a book by charles yang called the infinite gift which focuses a lot on acquisition but brings up these things there's a recent book by david adger a-d-g-e-r i think it's called something like the infinite mind or something like that which is an excellent introductory discussion at a high level to many of us there's a very fine general introduction to many aspects of language by an outstanding linguist ian roberts it's called the wonders of language which goes into these and many other things there's a lot of very high level introductory uh literature for syntax the kind of thing i'm talking about our colleague here andrew carney has a book simply called syntax which goes into many of these things in a very lucid way but in considerable detail well thank you again so much we really appreciate it and uh i'm gonna i can't unmute everybody apparently um thank you thank you thank you so much thank you so much absolutely outstanding thank you so much thank you professor thank you thank you fabulous thank you thank you you

Channel: OLLI at the University of Arizona

Views: 14,853

Keywords:

Id: p2MTK7amseU

Length: 93min 13sec (5593 seconds)

Published: Tue Jul 19 2022