Mind Your Language: Thought, Metaphor and Imagination

[Music] [Music] [Music] that was the big bang well of course uh a visual and audio illustration of the big bang and the term big bang that is just a human metaphor a term that we invent to try to describe something that none of us has or ever will experience and and that's a strategy that we employ all the time i mean i'm a physicist but also a writer and one of the great challenges about writing about physics is finding the right metaphor the the right analogy the right language describes some aspect of cutting edge research and i can't tell you the number of times that readers of my books will come up to me and they'll say hey you you use a rubber sheet to describe the fabric of space you use bread rising as it cooks to describe the expansion of space you talk about jitters being infused into the microscopic realm to talk about quantum mechanics then they say but but how do you how do you brian green actually think about these ideas my answer often is i think about it in precisely those terms i use these metaphors and thinking about the actual science doing the actual scientific research and this capacity to use language to describe things that we haven't seen is something that we then go further on using the tools of mathematics but could mathematics itself just be another language one that is particularly well suited to articulating pattern more generally still is it the case that language so thoroughly suffuses human thought that it can be viewed as the substrate of human cognition and finally how can it be that swirls on a page or sounds that we invent can carry such power we die that may be the meaning of life but we do language that may be the measure of our lives i can see that in the midst of death life persists in the midst of untruth truth persists in the midst of darkness light persists if nothing else say this is true that even as we grieved we grew that even as we hurt we hoped that even as we tired we tried that we'll forever be tied together victorious not because we will never again know defeat but because we will never again sow division now where does this power come from was there some singular rewiring of the human brain in the ancestral past that gave rise to language is language something that we can examine at work within the human brain allowing us to understand its relationship to other qualities of human cognition is language and adaptively selected tool like domesticated fire or the hand acts one that builds communities and allows us to achieve things as a collective that would be unavailable to us acting as individuals and could it be that that most iconic quality of language metaphor using a term that originated in one context to tell us something about a vastly different context could that capacity be the root of human intelligence could that capacity explain how we prevailed how we came to dominate the planet these are just some of the questions that we are going to be investigating here tonight with some of the leading thinkers on the subject gnome chomsky evelena federenco daniel dorr and stephen pinker all right let's begin with noam chomsky six decades ago noam chomsky revolutionized the field of linguistics with the contention that there was a genetic basis for our capacity to acquire use and understand language and that all languages have an underlying common structure his theories helped make professor chomsky one of the most widely cited scholars in history and his ideas are still the subject of vigorous debate he joins us now from his home in arizona welcome professor chomsky very pleased to be with you thank you very much let's jump right in famously one of the less celebrated co-discoverers of evolution by natural selection alfred russell wallace he was deeply skeptical of the possibility of explaining human intelligence culture language in particular through the use of evolution by natural selection instead he turned to a higher power as the ultimate explanation what was he missing where did he go wrong in that thinking well remember that was before anything had been discovered about genetics about mutation about the way in which a genome can give instructions to a production system which is not very well understood even now but it does yield complex structures he also was missing uh the discoveries of the theory of computation girdle turing other great mathematicians of the 20th century established how a finite device can have infinite scope you go back to galileo say they were galileo and his contemporaries were quite properly amazed by the fact that with a small number of symbols we can create infinitely many thoughts uh and even convey to others who have no access to our minds the inner workings of our minds they were wondered how this uh achievement could be realized that galileo himself thought that the alphabet was the most stupendous invention uh ever made because it was somehow able to record this capacity but we now are beginning to grasp how a very simple restructuring of the brain maybe a small mutation could have led to the richness and complexity of the capacity to produce thoughts indefinitely and optionally to convey them to others you can get a grasp on that idea so what do you feel is the is the strongest evidence that there was this mental rewiring when it comes to language that there was something that happened inside the brain of our forebears that made them able to have this capacity to acquire and to produce language well the strongest evidence is what we're doing it is a species property it's now known that this genomic analysis showing that humans who began to separate at least 150 000 years ago that's not very long an evolutionary time after the appearance of homo sapiens maybe two or three hundred thousand years ago which is a flick of an eye before the appearance of homo sapiens there's no evidence in the archaeological record of any significant symbolic activity maybe a cl you know a line on a bone or something but none of the complex symbolic activity that you get with the appearance of homo sapiens so here we have a species property common to human the splits the groups that split have the same language faculty so that means it was in place prior to the separation doesn't seem to have been in place anywhere prior to homo sapiens so it looks as if some small change took place well that's side evidence the main evidence is that we can now uh show at least for significant parts of language that you can account for them on the basis of the assumption that nature picked the simplest possible solution when a generative process appeared uh other there's no analog in other organisms so you can try very hard has been tried to teach an ape uh something like language but it's completely impossible and and are you convinced or even surprised if you are convinced that this happened once and that then ceded this genetic change throughout the species should it perhaps happen multiple times is it a very specific one-off chance event that just endowed us with this capacity well if you look at evolutionary history lots of things have happened only once by accident so why aren't we all bacteria because by accident one bacterium happened to swallow another microorganism which led to structural changes because of physical law which led to eukaryotic cells which means we have complex organisms uh a human the appearance of humans after almost four billion life years on earth is an enormous accident um it could very well be that there's no thought anywhere in the universe other than this one accident we have no evidence for it anywhere else and by now there's work suggesting that the probability of anything like complex intelligence even life is extremely small within the accessible universe physically accessible so for all we know we're totally alone it's just some accident and it occurred in the last instant of evolution on earth so even if even if life turned out to be relatively commonplace throughout the universe which is a a possibility we look at the number of stars and the number of planetary systems and the basic biochemistry necessary for life if it comes to intelligent life and if it comes to intelligent life that harnesses the power of language that in your perspective could be a singular event in the cosmos we could be the only life form that uses this mode of thought and this mode of communication that's uh quite possible in fact the grand old man of uh american biology ernst meyer had an interesting debate which you may have seen with carl sagan on the likelihood of finding intelligent life and sagan astrophysicists gave basically your argument all these stars all these planets why shouldn't it appear everywhere that's the fair famous fermi paradox where are they uh meyer arguing on the basis of as a biologist said let's look at the one case we have uh humans that life's been on earth for for roughly four billion years if you look at the actual evolution of humans you find accident after accident just happened i mentioned a couple like an asteroid hit the earth and wiped out the dinosaurs left a little bit of space for small mammals to develop okay ends up being us uh and there's case after case like that they said the chances of this happening are extremely slight but now there's even work show indicating that the complexity of rna sequences is extremely unlikely to appear uh just by straight statistical um measures so maybe there's nothing and maybe it just happened and that's the answer to fermi they're not there okay so deeply lonely a cosmic loneliness that uh may well be how reality is constructed um so focusing back here on on planet earth one question that people naturally ask when it comes to the thought that human language is so singular they point to other other species now you mentioned the difficulty in teaching other animals language but when it comes to you know whale songs or or dolphins or you know gesticulation among some of our primate cousins you see that kind of communication as radically different from what we're doing right now well first of all that's communication uh every species noun to bacteria has modes of communication but language is not just we use it for communication of course but it's basically an instrument of thought uh that was [Music] a long tradition i mentioned galileo but many others just assumed that language is what was sometimes called horrible thought it's the instrument of thought now if you think about our own like language in use that's almost all we do is language is to construct thoughts in our mind sometimes we externalize them sometimes we communicate but that's a kind of a side property of language now by now we can pin that down much more precisely when you look at the actual design and structure of language what you find systematically is that if when there are instances where there are conflicts between communicative efficiency and computational efficiency computational efficiency always wins uh there are very simple cases like uh simple structural ambiguities so if i say flying planes can be dangerous it's structurally ambiguous means the act of flying or the planes that fly well the point is if you just let the rules run freely computational efficiency you get things like this but they harm communication because means can't understand what somebody's saying do they mean to fly planes or planes that fly and there are many things like this some of them pretty serious which impede communication but mother nature didn't care when evolution moves along some change takes place you get the simplest solution simplest most elegant solution may happen to be dysfunctional not mother nature's problem you've got it and that's the way it seems as language develop but just so i can completely understand when you speak of language as a as a tool of thought are you referring to the inner voice that we all hear inside of our heads or are you thinking a more abstract version of language in the thought process well the what's called the inner voice is actually an outer voice if you think about when you're introspecting and using language in your head you can tell how long the sentences are you can tell whether they rhyme which means that you're hearing the outers you're hearing external speech with the articulatory organs disabled but it's not what's inside there's things going on in the mind constructing thoughts which we can't consciously access but which enter into everything we do like interpreting that sentence the way we interpret it and this you find all over what's called inner speech is really just outer speech where you haven't bothered to articulate it the inner speech you can't access and that's going on we can only see its effects and from its effects you can determine what it must be much like anywhere else in the sciences they do we we can't communicate the way as well as other animals can like you and i couldn't be taught to imitate the waggle dance of bees we're just not constituted to do that bees do it instinctively it's a complex system so it's not a matter of more advanced less advanced just different paths and these are modes of communication but what's unique to humans is the capacity for thought so if but if one can introspect one's way to the recognition of language as a tool of thought and if the only thing that we have access to as third-party investigators is the externalization of language i don't mean this as a facetious question but how do we know about the other animals only thing that we really know is that they don't externalize in the way that we do i mean of course i i agree with you i don't think that my dog is thinking about the general theory of relativity or proust or anything like that but but if it's only the externalization that differentiates us from a third party perspective how sure are we about that well in the empirical world we're never sure there's no certainty there's only best theories construct the best theory again so i have a dog two dogs sitting under the desk and i can study the kinds of things they do sometimes moderately complex but if i study their repertoire carefully i think they have maybe a dozen things that they react to there's no indication of anything else going on maybe there's something going on that we have no indication of but you could say that about an atom sure no maybe thinking maybe they have thoughts too no no i i and they're actually of course there's some philosophers who think that atoms do have have thoughts well what we do is scientists just keep to the best simplest theory we can find people want to speculate about other things okay but we have no reason to believe it but if if language then is is fundamentally a tool of thought and that's the basis of our thought did wittgenstein have it right is the limit of our language the limit of our world one issue is whether our mental capacities crucially language and thought but other mental capacities whether they have limits so is it the case as many believe in including some great scientists and mathematicians like hilbert that everything is within our capacity every possible program problem we have a way to solve what's a possibility the other possibility is we're organic creatures like other organic creatures our capacities have scope and limits and that means there might well be problems that we simply can't solve and there might be uh problems that we can't even formulate because we don't have the right cognitive apparatus well those are two points of view there's a separate issue uh which actually wittgenstein probably had in mind it's the sometimes called the superior warf hypothesis that the nature of our language determines the way that the particular language say english not hopi determines the way we perceive the world that's a second separate hypothesis the second one has been investigated empirically for about 70 years by now and there's very little evidence if you look closely it turns out that the there seems to be deep uniformity in the way people think even if they talk about it differently so on the second one i think it's very dubious that there's any significant impact on the of the particular language the first one is a much deeper problem and it's very hard to answer how can we within our framework of thought determine whether there are problems we can't even formulate that's not simple i just want to quickly turn for the last part of the conversation here to human creativity a lot of people would look at that as the iconic quality of what our species is able to do come up with novel solutions innovate ingenuity when facing challenges to what extent do you consider language to be the basis of that creativity well first of all there's traditionally hundreds of years ago there's two different kinds of creativity distinguished one of them is the normal creativity uh taking place at every moment of our lives so just talking to each other every time we're talking to each other talking to a friend at a bus stop we're constructing new expressions maybe never before appeared in the history of the language the other person understands them that's galileo's problem the constant creativity and innovation in norman normal human life exhibited in the construction of thought by language and the occasional interchange with them that's one kind of creativity then there's the kind you're talking about now what has uh maybe aesthetic value or new insights radical new insights into the nature of the world that second kind of creativity nobody knows anything about uh the first kind we can get a partial grasp about but notice we don't really have an explanation for how it is that you you and i are doing what we're now doing how can we can constantly construct new thoughts sometimes express them appropriate to situations not caused by situations that was the problem that led descartes to establish a postulate a second substance race cogitons which he connected to thinking speech but connected to thinking nobody understands that but uncertainty the higher form of creativity we can gesture at but there's no real understanding of it but bear in mind that we don't even understand the most elementary elements of voluntary action like if i decide to lift my finger why am i deciding uh there's plenty of work in the neurosciences on the basis for voluntary action but can't get this far but then when you think of language as a tool of thought are you thinking of it as the exclusive tool of thought or one of many tools that we use i mean when beethoven was writing the ninth symphony was language driving that when einstein was working out general relativity is language driving that well here there's a comment of wittgenstein's which is a posit i think uh he says that uh in his usual aphoristic style he says uh humans think maybe dolls and spirits meaning uh the word thinking the way we use it it's something humans do in language uh sometimes there's kind of open texture we are sometimes willing to metaphorically extend what humans do so maybe dolls and spirits actually turing had something to say about this too in his famous paper on which launched artificial intelligence can machines think uh he opens the paper by saying the question whether machines can think is too meaningless to deserve discussion because we have no concept of thought except for human language and its expression of thought so if you ask when i'm planning to drive to work and i pick this path not another path there's something going on but whether you want to call it thought is like asking whether submarine swim you want to call that swimming okay but it's not a substantive question so if i was to ask you whether artificial systems can or will ever think would you give me turing's response i would give you touring's response if if in fact he went on to say look if we by machine of course event program if we can construct a program that would have many of the properties of what humans do we might be willing to extend the metaphor and say yes they're thinking the way we are but to ask the question whether the program thinks is just doesn't mean anything we don't have a concept of thinking that's independent of what humans do uh with the use of language so it's kind of there's a lot of philosophical controversy about it but it's idle it's like having a controversy about whether airplanes really fly in some languages you say they fly some languages say they glide and some doesn't but there's no substantive question it's a question about what metaphor was willing to accept and and would you even apply that and perhaps you would if i was to ask do you think it's possible for some artificial system that we create to have some kind of inner world not even applying the word think per se but just some kind of inner experience in a world does that fall into the same category of questions that it's kind of meaningless to answer from your perspective it's not meaningless but it's based on discoveries which haven't been made what is it that gives us experience nobody knows at some level of complexity the brain does have the capacity to give us an awareness of things around us conscious awareness nobody has any idea what this is even where it is i mean neuroscientists assume it's something about neural nets uh people working in deep learning use that model but it may be completely the wrong model uh there is uh randy gallistell's one outstanding neuroscientist cognitive neuroscientist who's argued i think pretty persuasively that neural nets in principle do not have the elementary element required for a touring machine computational device so you have to look somewhere else and they've now even found evidence that internal to cells you get computational properties the the cerebellum has huge cells perkinj cells and there's some evidence that internal to those cells you're finding computation going on conditioning going on and so on and it could be and of course there's massive computational capacity once you get into cellular structure so maybe neuroscience is just looking in the wrong place now we'll have to have some totally new direction here but we're moving into areas where you know there's a lot of scattered evidence but no deep understanding so one final question as we reach the end here you've had a long career with profound impact in so many areas your own field linguistics but cognitive science political thought governmental structures and so on as you look to the future do you have a sense of optimism about where we as a language wielding thinking species are headed or a sense of pessimism well we're a thinking species maybe the only one but we're not thinking very well let's face the situation that now exists there is a major crisis coming there's an environmental crisis which is marching forward we know we have a grasp about what can be done to arrest cataclysm but we're not taking the steps we know what they are we're not taking the steps in fact that's even true on narrower questions one of the things if some extraterrestrial intelligence if there is one is watching us would think we're insane i mean take something as simple as the covid crisis every state every country developed countries scientists leaders understand very well that unless they provide vaccines to africa and asia poor countries in latin america the virus is going to mutate it'll lead to probably lethal developments it'll come back and strike us are they giving them the vaccines no we are so consumed by greed and commitment to maintain profits for pharmaceutical corporations that leaders all around the world are saying okay let's commit suicide i mean that's what we're watching is that a thinking creature of a global warming it's much more serious um coven get over somehow global the melting of the arctic uh the polar ice caps we if it's there what's done you know and we're watching it and not solving the problem so we know how to solve it so is that a thinking creature if you extrapolate future doesn't look too bright so uh there may actually be no at the moment thinking creatures in the universe to some extent but let's hope let's hope that future generations will look back and not draw that conclusion it's been an honor to speak with you professor chomsky thank you so much for joining us thank you the origins and function of language are the subject of intense debate among linguists but what can the brain itself tell us about what language is and where it comes from evalina federenco is an associate professor in the department of brain and cognitive sciences and the mcgovern institute for brain research at mit where she runs a research lab devoted to studying the cognitive and neural mechanisms of language she joins us from her home outside of boston welcome ev also joining us now is daniel dorr professor at the dan department of communication tel aviv university he's a linguist and the author of the instruction of imagination language as a social communication technology in which he argues that language is a collectively inventive tool that allows us to harness the power of imagination he joins us now from his home outside of tel aviv welcome daniel so thank you for joining us so noam chomsky has laid out his perspective which you both are already intimately familiar with that language is fundamentally a tool of thought and only in a peripheral sense is it a mechanism a means a medium of communication now as i understand it both of you have a very different perspective on the fundamental role of language just to get us going daniel what do you consider to be the fundamental purpose the fundamental role of language well first of all it's a tool of communication pure and simple it's not a tool of thought but what makes is it's different from other tools of communication that we use and the other animals use is the fact that language opens a totally new channel for communication for us that no other system opens and let me very briefly describe what this is when you look at all the other systems of communication that animals use and that we use what you see is that those systems allow us to present our interlocutors with materials to perceive with their senses so i show you a picture you see the picture i play music for you you hear the music i hug you you feel the hug language doesn't work that way if i uh tell you about an event that i participated in um the other day you don't see the event or feel it or hear it what actually happens is that you imagine the event inside your mind and the way you do that is by taking my words and the structural connections between them and using them in order to get from your long-term memory clusters of experiences that are related to the words and then rearrange them according to the structures of my sentences to create an imagined com experience inside your mind and what that has allowed humans to do for hundreds of thousands of years now uh is to communicate about stuff that you cannot show stuff that you cannot present to the sentence to the senses and that includes stuff from the past stuff from other places talking about the future that doesn't yet exist at all talking about stuff that just simply doesn't exist and it also very very importantly it allows us for the first time to communicate about mental dynamics questions that we have ideas that we have arguments that we want to get into and to create a sort of collective consciousness a sort of larger brain made out of all our different brains that allow us to think in totally different ways but the way language contributes to thinking is not by helping us individually to think in ways that are different the way language helps us is by allowing us to do parallel computing by allowing us to connect our brains on a network of thinking and i think that every big thing that humanity has done and including all the bad things that are big are the result of the fact that we no longer only think by ourselves and that is the thing that language has helped us do it has created an entire domain of collective imagination where we all live which characterizes our life our lives much more deeply than this property or the other of our individual brains so ev do you do you have a similar view that that unlike chomsky's perspective language is not a tool of thought primarily and in particular if you do hold that view i'm wondering if you can address one of the arguments that chomsky has put forward which is every time it comes to a kind of competition between efficiency and computation and efficiency and communication he argues that the computation seems to win out which suggests that primarily that is where language has its fundamental home in the nature of thought uh yes so um i find the question of the relationship between language and thought obviously deep and important and it's a question that drew me to science but i am a very strong empiricist and i like testable ideas and to the extent that one wants to argue that language shares machinery or mediates something like complex thought and reasoning abilities we should be able to evaluate it and we should be able to see if that's on the right track and so for the last decade or so i've been trying to evaluate that idea and it just seems like there is um no support for it from all sorts of um approaches so uh in the human brain the system that works right now in your brain as you're understanding what i'm saying and in my brain as i'm generating this is absolutely silent when we're thinking logically when we're doing math when we're doing kind of common sense reasoning about object properties event properties in the world and you so on silent just so i understand you mean it's like not lighting up in exactly some kind of mri yeah exactly whatever measures we have to um uh measure brain activity which is it's an indirect measure but it has been linked to cell firing at some very um basic level and so the language system just doesn't do anything whereas other parts of the brain are working really hard so it's not like something in the brain is doing that work obviously it's just not the same system that allows us to produce and understand utterances now you can also take individuals with severe language damage so these are individuals who have you know grown up and their brains are well formed and working well and then suddenly they have a something like a major stroke to the one of the big arteries that supports the parts of the brain that underlie language ability and these people um so that's a syndrome known as global aphasia so it's a very very severe linguistic deficit these individuals cannot produce language they cannot understand language and yet they can play chess and they can do sudoku puzzles and they can think very very complex thoughts about the world suggesting that language is no longer is not needed at least once the brain is fully formed and so empirically just doesn't seem like there's any evidence for this link between linguistic processing and thought it could have been that way it's just not and and and in those examples are you measuring the individual's lack of language capacity through the inability to speak or you're also doing it through examining what's happening inside of the brain while they're say playing chess can you have someone playing chess in an fmri machine you can make it so but um uh a lot of these individuals have conditions which actually prevent them from being like a lot of them have pacemakers and other things that make them unable to undergo fmri but so you assess them behaviorally but you're not just tapping the kind of surface output structure we have tests we as a field of language research has developed tests over the years that allow us to tap uh what kind of knowledge representations are retained and the best we can tell um the the language system was effectively gone and if you just look at the extent of damage it's basically the whole cortical chunk uh that supports language is gone in those people it's not like a subtle little lesion there it's a massive massive damage and so and and you do not think in those cases that there's some kind of individual rewiring that allows the individual to undertake certain cognitive tasks that ordinarily might involve language but in their case the brain has found a way to avoid that damaged region um well combined with fmri evidence i don't think that's likely because otherwise in your typical individuals in you and me for example um the language brave each should be working hard when we're engaging in complex thought right so the fact that they don't seems unlikely but even so um the argument would still stand that complex thought is possible without language even if it somehow moved around in the brain and it's not quite the same as a neurotypical yeah absolutely so so daniel what about the argument that chomsky gives in terms of the the intrinsic structure of language somehow favoring the computational qualities as opposed to the communicative qualities when you look at the design features of language itself let me let me uh take a step back and and say something about the structures of language from my point of view i think that as long as you think like professor chomsky you're thinking in this way or the other in terms of the computer metaphor there's this computer on our on our shoulders and this computer is a computational machine it does all kinds of things and the idea is language is a sort of software inside that computer i think that everything we've learned in the last 50 years actually tells us tells us that we should finally replace our metaphors and move from the computer metaphor to what we may think of as the internet metaphor language resides on the net it is a property of the community the structural properties that you find in language when you actually get your hands dirty and do the linguistic analysis what they reflect is an entire history of social negotiations within a culture within a community about what we need from language as a tool of communication how to structure it and once you think about language um as something that resides on the internet like facebook or tiktok then you can see that we as individuals are end users we are not our brain is not the site of language we are end users of language and there's a lot of fascinating stuff that we need to that we already know and a lot of stuff that we need to discover about what happens in our brains as end users but i think that the most important thing for me as a linguist is that once you look at the structures of language and you start thinking about them in terms of interaction in terms of the relationships between speakers and listeners in terms of the different interests of speakers and listeners you can actually explain the structures much much better and you don't need to stipulate stuff of the of the stuff that we've heard of the type that we've heard that for example sentences that are ambiguous in language actually emerge inside the mind as a non-ambiguous sentence i think this is something that just doesn't make any sense if you think in terms of the relationships between speakers and listeners you can see for example that speakers and listeners have different interests in the story as far as you're concerned you would best prefer to say a word or two and let me figure out everything else as a listener do all the work by myself that's what sometimes happens in intimate situations right a good friend asks you what happened in the meeting and you say oh it was okay because you understand that the other person can put in all the other information for me as a listener when i listen to you my preference is for you to put everything correctly and right and with all the details so i don't have to work hard to guess what you meant and between those different interests we come up with a certain compromise not just you and me the whole history of the community with the language a certain compromise that actually determines the logic of the structures and we as individuals who are end users do not have to represent all this logic in our minds we just have to be able to use language the way we use facebook which means that just like you cannot understand facebook on the basis of what's happening in your own computer you have to look at the network the same thing is also true for language so if if language is as you say fundamentally a tool of communication then ev for instance would you argue that it has a relationship to other tools of communication that we do see in the animal kingdom from look i can't talk without my hands flying into motion right and and there are various other facial expressions and things that we do is language a kind of natural next step in that progression of communication strategies i think so i mean i think we um have lots of prior reasons to think that there is biological continuity in general but in terms of what we know from studying um modern day modern day humans that does seem to be a deeper link between the systems that support social perception uh and cognition and language compared to the systems that support kind of abstract complex reasoning and language um and the evidence comes uh in a few different forms from the fact that um the general topography so how the landscape of activation in our brains elicited by linguistic processing versus social processing are similar or parallel they are distinct so as i mentioned the language system is actually quite specialized for processing language so it doesn't it's not active when we engage in complex thought or when we think about others mental states or imagine their faces or gestures and so on but the regions that support those other social functions are right next door to the regions that support language processing and that gives rise to this interesting possibility that evolutionarily maybe there were kind of broader regions attuned to social stimuli because we are social species very deeply in you know very fundamental ways that this last year has made perhaps uh abundantly clear um and it's made it more clear to me than i than i would have thought before yes same and um eventually as we got more and more complex and the relevant parts of the brain expanded because this is the kind of what's known as the association cortex so it's cortex that's not primarily involved in perception or motor control so it's basically the cortex that does all the interesting cognitively complex stuff as it expanded and became more complex in humans we now have this fractionation into some parts that say perceived faces some parts that perceive gestures right or body language and some parts that actually store mappings between some symbolic representations like words and meanings in the world right because that's effectively what language is is a system that allows us to map some aspect of the world some object some type of event onto a string of sounds or a string of signs if we're talking about sign language but um i think there is um a lot of suggestive evidence that language and social cognition are very deeply linked so so daniel can you give us a a a brief history if you will of the steps by which you imagine language was brought into the world through whatever pre-linguistic modes of communication our forebears may have been struggling with and then language resolved this design issue of how to bring the level of communication to a higher place i i i think that you got a tail there of three parts the the first part basically starts in the cambrian explosion a half a billion years ago when brains actually began to evolve and what you see through a very very long process of mental cognitive emotional evolution our brains that are specialized in all kinds of ways for different animals for what you may call private experiencing the brain is supposed to be the commander of behavior of the individual carrying it all right and and this is something that is true as far as we can tell of all animals maybe apart from social insects even when animals live together socially even when they cooperate you don't get mental cognitive cooperation in the sense that you don't get parallel computing individual animals do all the computations they need inside their own head sometime around two million years ago in homo erectus or maybe in abilis you start to see in the archaeological evidence signs that human communities broke a barrier and they broke the barrier of private experiencing and moved to a mode of collective experiencing using memetic tools of communication that include pointing and eye contact and demonstrations with the hands and all kinds of stuff like that where individuals within the here and now of the shared experience could start to compute together to think together and this worked wonderfully for a million and a half years but it turned out probably that the needs for more and more communication for parallel computing became more and more severe and at some point as far as i can tell human communities broke a second barrier first they broke the barrier of private experiencing and moved to collective experience and and now they broke the barrier of experiencing and went beyond that into collective imagination and once collective imagination proved to be such a useful thing for human communities because it did very very quickly it opens your possibilities in unbelievable ways uh you got into this spiral of co-evolutionary dynamic in which communities languages and individuals co-evolved with each other and within this process individual brains came to be selected for their ability to participate in the social linguistic game so that eventually we're today in a situation where you have communities very dependent on language you have full-fledged languages and you have language-ready brains language-ready brains in the sense that the brain the whole thing is capable of um recruiting enough brain energy in order to meet the challenge of language acquisition uh without there being any universal grammar or set of formal rules that are somehow given to us in our genes because evolutionarily speaking this idea doesn't make sense then ev would you then come to the conclusion that if we at some point make contact with extraterrestrial life that we shouldn't be surprised that they have some form of communication because presumably in order for them to make some sort of contact with us in some form they had to work as a group and therefore would have the very same social pressures that daniel was just talking about and then that would naturally yield to communication of the sort that vaguely we're familiar with as opposed to say chomsky's perspective where there's a good chance perhaps that we're the sole thinking linguistic species in the universe yeah yeah i think that's unlikely i mean um i you know i have no stake in this i i can't assess the likelihood i mean i can guess based on the numbers of planets that are out there but um uh yeah i think if there is a species um some kind of biological species that have enough of a social bias to interact with each other and a sophisticated enough internal world then of course they're gonna find a way to share information because you are empowered when you can do things together right you can protect your group better jointly you can share knowledge with offspring i mean you can do all sorts of things that you couldn't do and of course you can um you know invent calculus and figure out how to fly to other planets and things like that so language is certainly empowering but i think thought development uh fundamentally precedes language so daniel ev made an interesting remark there in terms of say inventing calculus that makes my ears perk up because i'm one who uses calculus in in my day-to-day life is calculus and just view calculus of course as a representative of what the human mind human creativity is able to accomplish is that an example of instructing the imagination going overboard have we sort of just swung past the utility of imagination instruction into a place which is fascinating but is really just an over swinging of that capacity i don't think it's an overswinging of the capacity but i think that something slightly different is happening there one of the most important things for me is that once you look at language at the right ontological level as a system of communication you start looking and you start seeing the fact that just like any other tool it fails on us from time to time it has weaknesses it doesn't always work very well and the foundational reason why it sometimes fails us is the following you have your own experiential world that you live in and i have my own and they're not the same we're not the same person and when we use the same language for example english we get this sense which is sometimes illusory that your words mean the same thing that they mean to me and because of that very often you'll say something i'll think i understood you but it turned it will turn out that i didn't and we'll get all kinds of misunderstandings so i think that once language is there you start a process which is again collective innovative of trying to brush the thing to get it to be more and more precise it's one of the huge forces that plays there and i think that all the stuff that you guys physicists and math and mathematicians work with are tools that uh uh you guys the tradition managed to make to make so precise managed to make so unambiguous that you eventually were able to discover all kinds of fascinating things about the world as a result of that process of honing the meanings so i don't think it's a different story i think it's one of the different dynamics that emerged from the fact that we started to communicate and in day-to-day communication uh people who are doing um a work in conversational analysis for example nick enfield is a very important person there they can show you in a regular conversation that sometimes we spend more than 50 percent of our time of conversation repairing misunderstandings exactly because language is not that ideally designed tool that chomsky is talking about it is something that our ancestors invented hundreds of thousands of years ago when they still use stone tools and it is a very tentative system it's ingenious but it's also tentative so mathematics is one of those places where the entire developmental process that we humans went through got us to a place where yes our words one two three plus minus have come to be so precise that now you can use them as wonderful tools for the understanding of the world so just to be clear you you are saying and i agree with this perspective you are saying that math is a language i am saying that math is a language in the sense that in in the way you guys use it the people who use language uh to work with i think that one of the failures of educational systems is that children are taught mathematics in a way that doesn't go all the way to turning it into a system of communication for these kids right um but but yes so ev one of the other ideas that flows from chomsky's perspective that there was this singular mental rewiring as he described it that allowed us to have this capacity to acquire language that gives a nice explanation for why there is such a deep similarity between the world's languages after all any kid can learn any language and the reason for that would be that this mental rewiring is latching on to this universal quality that's common to all languages without that what do you consider to be the the strongest explanation for why there is this deep similarity between the world's languages well first of all i think the similarity should not be overestimated um and a lot of um work in the trump skin tradition has been done on english and all the examples they'll tell you come from english so there's actually drastic difference in how the world sound systems lexicons structures are that said there is a set of common universals if you wish but these universals are all to do with cognitive and communicative efficiency so they make our production system efficient they make it easy for you to select words and put them together in quickly and robustly so that any comprehender can extract meaning from those utterances and on the comprehender's side on the side of understanding we have mechanisms that allow to fill in the gaps in the signal just like um dan was saying where um you know we can uh extract very rich meanings from very minimal linguistic input now what i think makes the human brain language ready is in part just its size i think that's something that's been very undervalued uh in language research where somehow the idea of some new brain region emerging that can support some fundamentally new computation has found a lot of appeal but from everything we know about biology and neuroscience um it just doesn't look like this and the evidence that has now found all sorts of homologies between human brains and brains of other primates and so um i think this ability to store a vast number of these mappings between forms and meanings is something that is grossly under a pre has been grossly underappreciated i think it's now getting more attention and the standard argument is you know my dog can learn some words yes but it cannot learn 10 000 words it cannot learn 100 000 words and humans capacity to be able to encode these very vast numbers of symbolic representations that they can use to talk to each other about a vast number of ideas i think that's a uniquely powerful capacity that has enabled a lot of what we see in the world today i'm always i'm always fond and i give the example of my dogs being smart but i i i don't think that they speak physics or understand general relativity but every time i say that i do wonder if they're off in the corner barking among themselves and saying how silly those humans are we've got the unified theory and they haven't even gotten there gotten there yet but yes daniel you uh had a remark on that i i think that th this whole idea that languages are essentially more similar to each other than they are different from each other this is one of the greatest mistakes and i think that now we're eventually in a situation where researchers actually do comparative studies using a lot of computer power for analysis over hundreds and hundreds of languages you have 6 000 languages around the world and on every level that you want to look at syntax and morphology and phonetics and what have you you see more and more remarkable differences and if you put into it not just the spoken languages but signed languages which are languages for you know all purposes and you even put into the story prototype languages that are now being investigated where people who are deaf and blind communicate by tapping you see that on all the structural levels there are huge differences between languages this is actually one of the most important things in the whole story if you go all over the world and you meet people and you communicate with them with facial expression and pointing you can go very far you can say pointing is indeed a universal thing but there's no language in the world that you can speak that you can use to speak to more than 20 30 percent of the population of the world and if you really want to get a window into human nature you're not going to find it in what is shared by all languages which is very very little you're going to find it exactly in the sense that the human species has managed to find 6 000 different solutions to the same basic engineering problem of how are we going to be able to communicate with each other and instruct each other's imagination the window to our nature is in the diversity between languages not in whatever putative universal there is there because there aren't many i mean does it disturb you then that the world is losing so many languages at such a rapid rate definitely i think this is one of the most worrying things that uh we have to think about uh according to some estimates of the 6 000 languages that we have today around the world approximately 90 will be lost by the end of this century and i think it's a huge worry and of course there's a lot of sociological political reasons for that and very often you see this beautiful beautiful language with such strange engineering solutions that you never even thought about a product of ingenuity uh that right now you have you know 10 speakers 20 speakers 100 speakers and we're definitely losing together with those languages that are dying we're losing traditions of thought about the world that are very very important beautiful inspiring and so on and so forth so yes it's a very very good very serious reason for worry yep ev why don't you take the last word as we wind it up here sure i i mean i just wanted to um disagree a little bit in saying that i don't think it's uninteresting to try to understand which properties of languages are similar because there are certainly parallels in how humans have solved these problems of communicating you know infinitely well don't have to go to infinity a large number of meanings okay um but i think what these universals to the extent that they have been discovered have told us is that all of these constraints are basic cognitive constraints there are constraints from memory limitations from how our attention works from how far we can predict into the future for how we hold on to information and package it and um abstract away from the input to more generalized things and i think that tells you again um that it's not very likely that language kind of emerged as this little snowflake in our brains and suddenly everything was different it was built on the very basic fundamental components of neural computation that characterize many many species um that are you know lower on the evolutionary dream and so the one final question i said it was going to be the last one we asked one more as you look to the to the future then maybe each of you can just give a brief answer daniel what do you see as the next major thing on the horizon and studying human language i think that once we manage to go through the the conviction that uh language is this way or the other not a tool of communication and we manage to settle on what we find empirically as part of our scientific work and we managed to start looking at the way communities have built languages and we start looking at the histories and the way they interact with cognitive capacities we get to know more and more about the differences and similarities between languages and we get to understand more and more of the evolutionary dynamics that have characterized this because we haven't talked about evolutionary developmental psychology for example the whole biology story of evo devo evolutionary developmental biology we understand the co-evolutionary dynamics now in a much much better way and i think that once we really understand the diversity of languages and we really understand the dynamics of the brain together i think will have a sort of first step towards a unified theory of language that is realistic and i think that's the goal ev how about you look into the future yeah um so to circle back a little bit to um deep learning and artificial neural networks i've been incredibly excited by the developments and in computation artificial intelligence um i think the models that exist these days um that are language models are exquisitely good at language they're often criticized but what they're criticized for is their inability to think now they're not trained to think they're trained to predict the next word in a sequence and uh now that we've shown in our work and some related work that the system that supports language comprehension and production is actually quite separate from the system that supports thinking abstract kind of hierarchical uh relationships among all sorts of elements be it in math or in common sense reasoning we can utilize these language models along with trying to build models of thought which quite a few people are trying to do and then if we you know understand how those fit together then we have a more complete um computational explicit model of how the human brain may be solving these problems and then we can probe these systems to really understand what are the representations that we store what are the algorithms that apply when we generate a sequence or we generate a thought right and um and that can have all sorts of profound clinical implications educational applications because if we can build a system and make it do um something that humans do well i think that's quite powerful and now some people have argued following uh daniel's comments that maybe you want to build systems that from early on are built to interact with each other uh because that may provide some features that just wouldn't emerge ordinarily from if you just construe it as a kind of singular uh organism um and i think that's a that's also a very exciting direction yeah it absolutely is wonderfully exciting so daniel dora evelina federenko thank you both so much for joining us a fascinating conversation that was great thank you very much for having me stephen pinker is the johnstone professor of psychology at harvard university he conducts research on language cognition and social relations and is the author of 12 wonderful books including the language instinct the stuff of thought and the better angels of our nature books from which i and many others of course have been deeply influenced he joins us now from his home in boston welcome steve how are you doing great thanks how are you brian good great to see you so we've had some discussion earlier in the program about the origins of language and the purpose and the role of language and there's been a variety of perspectives and i think it's well caught by a phrase that guy dutcher a colleague of yours he wrote some time ago where he described the impressive range of theories circulating for how the first words emerge from shouts and calls from hand gestures sign language from the ability to imitate from the ability to deceive from grooming from singing dancing rhythm from chewing sucking licking and from almost any other activity under the sun so i gather part of the issue here of course is that words don't fossilize they don't leave an imprint in the physical world that we can examine the way we can other structures and the evolutionary records so from your view where does that ultimately leave us will we ever have an answer to the question of the origin of language i don't think you'd ever say we won't have an answer to the question right now we don't have an answer to the question i think as with many questions about the evolutionary function and origin of a trait the function is often more obvious than the origin so for example there's not a whole lot of mystery as to why we have kidneys they filter out the blood while we have lungs they oxygenate the blood where the kidney came from or where the lung came from was it a swim bladder or was it something else it's not exactly the same question the digressions are related but even if you don't know where the kidney originated that is what was the precursor structure you really do know that it filters the blood likewise with language i think it's pretty clear that it's there to exchange ideas and to negotiate social coordination what the first word was is a much harder question to answer and and so but right there there's a bit of a controversy i guess because noam chomsky famously says that language is really a tool of thought and quite peripheral it's a tool of communication what's what's your stance on that i i think it's a a a radical an extreme and a rather implausible position uh for one thing if language was primarily a medium of thought why why do we have words that can be pronounced out loud why is it the entire component of language called phonology namely the mapping from uh symbols to motor commands in speech or sign and in the other direction the acoustic phonology the ability to parse sound waves or in the case of sign language manual and facial gestures back into the words so that would be a complete mystery if language was a medium of thought where it's in your head that all you would need is to be able to register the symbols why do you have to pronounce them also why would words have to be conventional why couldn't i have an idiosyncratic representation of of dog and plate and couch and sun and moon why is my representation the same as your representation if all we were doing with those words was thinking thoughts because the thoughts could stay inside each of our skulls so the fact that that is two of the most salient and universal features of language namely there are conventions uh if i have the subject before the verb before the object so do you if we belong to the same community that would be a coincidence it was just a way to think and the fact that we can pronounce words uh would also be a coincidence if language is just a way to think well presumably chomsky would say that language has become a tool of communication as well as a tool of thought but its primary purpose where it began was in thought maybe the qualities that you're describing emerged as language was used in that way but that was a secondary use as opposed to a primary use is there any thing in that that strikes you as as plausible no because if you're simply talking about a system of mental representation of having thoughts and a system of inferring one thought from another and there are no shared symbols and there's no way of getting it in or out of the skull then it's no then i wouldn't call it language that's thought uh and indeed of course in chomsky's own theory there is a level of representation of meaning he called typically calls it logical form and that is almost continuous with thought to the and presumably a lot more universal than the particulars of a of a given language or indeed of logical form belonging to a particular language but if you're talking about it may just be a matter of as we say semantics in the sense of meanings of words but a system of simply of having thoughts is just not what we mean when we use the word language and and so when you think about language as a tool of communication is it general purpose communication or for instance like daniel dorr emphasizes language as really a tool to say instruct the imagination of the other very specific role that he believes language played in the ancestral past and the reason why we have it it is not our a uh our only channel of communication and there are some things that it's good at communicating others not so good one of the reasons we have facial expressions is that there are emotional colorations that are difficult to put into words one of the reasons that we have gesture and pantomime is that there are spatial configurations motion configurations that are very hard to convey in words precisely because language is is digital we've got words they we don't stretch out words more for bigger things and shrink them for smaller things they're they're just you you pick one word or you pick another word and we we order them from from uh early to late in the sentence that's excellent for conveying who did what to whom uh what where and why not so good for conveying high dimensional analog relationships it's virtually impossible to say to define something like a spiral without going well you know it's like this if you can't move your hands you're gonna have a hard time using words to tell someone where spiral is and so on for an infinite number of other concepts on the other hand language is it is very general purpose in the sense that for example a subject needn't be the agent of an action we can say you know john sustained an injury we can say the syst the situation justifies drastic measures the verb doesn't necessarily refer to an action the object is not necessarily something acted upon even though prototypically they are so there is a kind of generic set of semantic relations that language is optimized to convey uh highly useful not specific to any activity like eating or having sex or running on the other hand it is not identical with the thoughts we have and the thoughts we want to share now you've also noted that you see language as being part of the way that we humans fill the cognitive niche can you explain the role that language played in that and and what what do you mean by that particular place that our species has has filled yeah it's a a concept that i borrowed i.e stole from john tube irv devore and lita cosmetics and it pertains to the fact that humans are zoologically really really weird they're things that you see in homo sapiens that are rudimentary at best perhaps non-existent in other species language being the most obvious but not the only one we also are very unusual among animals in the degree to which we cooperate among non-relatives easy to explain cooperation among kinship among kittens they share genes not as easy to explain cooperation among non-kin who can much more easily exploit each other we also are have a lot of technological know-how we're not the only species that that makes tools but we depend on them a whole lot more than any other species tools that are not part of an innate repertoire but that we uh we invent we pass down culturally now uh devore and tube and cosmetics suggest that these three uh zoologically unusual traits and they listed a number of others are uh it's no coincidence that you find them in the same species because each one multiplies the value of the other two so it's not so easy to negotiate social contracts agreements if you do this for me i'll do that for you without language not impossible but but very inefficient it's not impossible to convey technological know-how how do you extract a poison from a plant and prepare it so you could put it on the tip of an arrowhead how do you fashion vines into a baby sling language is is pretty helpful for that as well and in terms of getting social cooperation going the evolutionary theory tells us that the cooperation is uh easiest to evolve when it's possible to confer a big benefit to someone else at a small cost to oneself because that sets up a kind of a market a kind of economy of favor trading uh you you don't get that started if it's just one uh cow say pulling up a lump of grass and dropping it in front of another cow the second cow could say well thanks i can get my own grass but when it comes to know-how just in a few seconds of breath it costs me virtually nothing and i can give you a an invaluable skill and so uh the uh language uh ideas can be duplicated at uh trivial cost we're seeing that of course in the digital revolution where you can make perfect copies of things at minimal cost it goes way back and the ability to share ideas means there is much more reason for people to be as we say on speaking terms namely uh you uh hang out with other people be if you have language and if they have something to say to you that is language evolved because we have something to say and we're on speaking terms and and vice versa we're on speaking terms because we have language and we have something to say because we have language language also allows you in the technical sphere to pool innovations so not everyone has to invent the wheel that people can improve on other inventions you can pass it down to your kids you can share it with others and so it's uh they probably co-evolve we have reason to believe that some degree of cooperation and know-how probably got the cycle going simply because we see them in in chimpanzees our closest relatives and we don't see language in them but it may have kind of primed a pump and created a kind of virtuous circle where each one made the other two more valuable right now among the other things that certainly seem quite special to the species we can investigate the universe we can write down laws of physics we can write novels i mean there are all sorts of qualities of the world that you can trace back in one way or another to the emergence of language so a natural next question is how do we leverage language into these abstract realms of thought and that naturally takes us to the subject of metaphor and i'd like to turn to that now and look i mean we both write books for general audiences i find all the time the exercise of writing a book is finding the right metaphor the right analogy for communicating abstract ideas to general audiences i also find i'll come back to this in a moment that actually in my research work metaphors are actually valuable even right there and trying to push the boundaries of understanding but one thing that was really surprising to me is the ubiquity of metaphors in the language many of which you know are hidden i mean if you if you focus your attention you realize that our language is just stuffed full of metaphors and of course right there there were two bad metaphors right of focusing attention we focus telescopes we don't focus our brain you know we stuff turkeys we stuff draws in general we don't do it with metaphors so so it's all it's all over the place i mean is there is there a sense i mean what fraction of our language is is metaphorical has been quantified uh it hasn't been quantified and partly because we don't know of those metaphors how many of them are dead metaphors that may have originated as metaphors in the minds of the first coiners the first poets the first colorful speakers but have been uh dumbly memorized ever since but the answer is surely most language at least historically and you can just prove this by looking up almost any word in the dictionary for an abstract word and its root in in in germanic and indo-european will uh refer to something pretty concrete i did this in my book the stuff of thought with the opening of the declaration of independence where almost every word like it went in the course of human affairs well course as in a water course or a race course is is a pathway and word after word you dig deep enough and you do find some concrete referent and i think it is a profound observation and indeed if i was to answer your your most general question namely how did i a and this this is a uh puzzle that goes back to alfred russell wallace the co-discoverer of the theory of natural selection namely presumably if the selection contingencies that shaped the human brain certainly did not involve being able to to do physics as we know it or or write novels how why are we capable of doing it how did our brain evol how did our brain evolve in the pleistocene that uh uh tens of thousands of years later allows us to discover you know black holes or to to to write novels and i think chomsky identified one of the key engines and that is the open-ended recursive combinatorial power of language and by extension thought that is language does not consist of a list of sentences that we memorize it is a set of grammatical rules that allow us to combine words and to embed combinations in bigger combinations and still bigger combinations without limit so there's no finite limit on the number of sentences we can utter but more importantly since each one of those sentences conveys a thought there is no limit on the number of thoughts we can think i think the other key to answering that question comes from in a sense chomsky's bet noir his uh long-time opponent within the field of linguistics and former student george lakoff uh from the book metaphors we live by who i identify the phenomenon that impresses you and me namely that language is just uh saturated to take another metaphor with metaphor uh we don't even realize it we say you know he attacked my argument and i tried to defend it well that's kind of appealing to an underlying metaphor of argument is war and uh as soon as you start to notice this you just see that that that almost all of abstract language is metaphorical now so i i think that the the the metaphorical abstraction where we can take ways of thinking that originally evolved to for you know chucking rocks and for for for pushing logs and extend it to these abstract realms even something is as abstract as uh say the uh analogizing the value of a variable to a location in space something that we do every time we plot a graph or to analog and analogize uh application of force to change to causation uh immediately our concrete cognition can be put to work for abstract scientific uh deliberation and then combine it with chomsky's insight that there's just no limit to the thoughts that we can think given these combinatorial rules of embedding one thought into another and you can you see the provision for an explosion of thought but there is one other piece to the puzzle though because just being able to think an infinite number of thoughts and to have all of these poetic metaphors can't explain why science should ever work why should given that you know a metaphor is just our reaching for something familiar why should it even apply to something as abstruse as the universe the fundamental forces the elementary constituents of matter i mean it just seems like an amazing coincidence that a mind that there are a cognition for rocks and sticks and and uh and and shoving things should work in these abstract domains and so there i think to uh to to solve that puzzle i i um appealed to an idea from a philosopher named richard boyd which is that they're the successful metaphors in science are ones that are really more analogies than metaphors and they appeal to some abstract uh relational system that embraces both the source of the metaphor and the thing that you're explaining maybe you know complexity theory or complex systems theory would be the source of these models but just to be give you a concrete example why would um you know a a a tetherball give us a good metaphor for the solar system and then you know and then even to some extent the adam yeah the bore adam uh i mean why did god make tetherballs which we play with and know about a good model for adams i mean that seems did he want us to understand it well presumably not presumably it's because there is a an entity abstract indie called say centripetal systems which can be characterized by energy and force and geometry that embrace uh tether balls and solar systems and atoms so it's not so much that they by an amazing coincidence we can understand an atom via a tetherball but just that their the universe contains systems that have many examples some of which are accessible to our everyday experience and our luck is that our everyday experience can exemplify these systems that also apply to much bigger much smaller much more abstract entities but the amazing thing there is while there are echoes of the phenomenon that take place on one scale happening on other scales and therefore allowing say the tetherball to be a reasonable bottle for the earth going around the sun when you do apply it to atoms it works just so far right it worked for the initial version and then somehow we're able to use the power of observation experiment and mathematics to break free of the initial metaphor and come to a new understanding of the world which then we invent yet new metaphors to describe so there's a great deal of flexibility that somehow is built into this whole schema where we're not locked in rigidly to only recapitulating our understanding of one scale on another scale is that telling us something deep about the human mind is that some quality of mathematics that we've developed and now we're allowing mathematics to take us forward are we learning more about ourselves or more about the universe yes no ab absolutely that's a vital point and again you and i as would be science explainers are constantly faced with the problem of reaching for an analogy and the analogy only goes so far and in fact if you take it too seriously then you'll misunderstand the phenomenon and i know i mean when when the the higgs boson was uh confirmed the metaphors some of them were just seemed so crummy that they you know running past a crowd and slapping people's hands and you know molasses and i mean i i've used molasses so don't don't don't diss molasses here okay glasses is fine no no it actually has some problems i agree no but you're absolutely right that the uh in addition to metaphorical or analogical thinking the mind has to be nimble enough to compartmentalize the analogy and to pick out what aspects are uh ought to be taken seriously and which ones ought to be ignored that requires a pretty sophisticated kind of representational computational uh system a kind of a logical form with variables that can be bound but also unbound i mean it speaks i think to the sophistication of thought that we can both enjoy and appreciate an analogy but uh also know how or ideally know what aspects to attend to and which ones to ignore absolutely vital so so we spoke a little bit before about the origin of language generally where did the first metaphors come from i mean is that a separate question i mean presumably language began in a very to use a metaphor concrete way we were actually speaking about specific things in the physical world around us at what point did a metaphorical version of the language start to inject itself to use another metaphor and and and and thereby give us this distinct way of describing things in the world well it may that step may not have been that hard if we mentally represent things not as an unanalyzed glob a gestalt but if we represent it in terms of its constituents its components its shape its weight its it's motion then two things that actually share components will automatically call one another to mind might literally be represented in overlapping neural real estate that is if we think of and you mean literally literally here literally literally yeah exactly literally literally i mean it could be that that in fact there is the challenge of how you don't confuse them but if for example you see a um see a tetherball and it isn't just the the tetherball scene but you represent it as oh there's an object and the object spins and the trajectory is a circle and there's something uh pulling it and then if you see say an animation of an atom and the way that it's represented in the mind again is not just adam you know full stop but oh it's got as an object in the big object in the center and a little one in the periphery and uh an invisible line of force connecting it and it goes around in a circle then you've got circle and circle and line and line an object and object it might be quite natural for one of them to trigger the other it may be even automatic if say your representation of object is just say even in an extreme case let's say it's just one place in the brain and then you think adam you think uh ball and parts of those thoughts might literally overlap so that part that part might be easier in fact probably the computational challenge in terms of neural computation is how do you not confuse them right now you mentioned and we've been playing around with it that metaphors do die at some point in the sense that we no longer sort of hear them as metaphors you just hear them as words describing things in the world and of course one way to establish that is just to look at the number of times people mix metaphors in ways that just don't make any sense and i was just looking at a few examples that i had written down you know people have written burning the bridge at both ends biting off the wrong end of the stick here's one that i loved it was the bottom of the ninth so they threw a hail mary i mean that's got so many interconnected parts i don't even know where to make sense of that another one that's a pet peeve of mine they invested heavily in the company resulting in earnings taking a quantum leap that's a funny one because i get it quantum leap is a leap that couldn't happen in the classical world but a quantum leap is the smallest possible leap at all and one that i'd never heard before i read in your book that maybe you can remind us of one that rabbi baruch korff wrote in referring to to nixon do you do you recall that one by any chance how how could i forget that the um the american press has literally emasculated the president yes right right um making him eligible for the the club called uh awful americans who figuratively use literally there you go and and so is there a sense of how long it takes for a metaphor to die it's not like a a constant and it depends i think on how many contexts it's used in whether people hear it uh divorced from its original context in which case since we can we have the ability this is a another elementary principle of linguistics one of the foundations is just arbitrary signs that is one of the things that makes language so powerful is that we can associate a sound with a meaning now often those sounds have a transparent relationship to the meaning as in metaphors as in onomatopoeia as in to some extent in what's sometimes called uh phonetic symbolism the fact that humongous is refers to big things and teensy refers to little things kind of echoing the or the size of the cavity in your mouth so there can be some connections on the other hand we'd be quite crippled in communication if every one of our words had to be like a pantomime where you could just sort of see what it meant just by its shape we can do these arbitrary memorizations tens of thousands of them and so what originated as a transparent metaphor if people just don't see the connection they just memorize it the same way you memorize that the word for english for dog in english is dog i mean the word dog doesn't you know look like a dog or bark like a dog we just all memorized it and we can do that and sometimes we do it too much as in the case of mixed metaphors is there some sense i mean we talk about words themselves in fact that we memorize them there's some connection to the external world is in there the answer to the question which i i've puzzled around and i think i asked you this once when we were together a couple years ago if you look up a given word in the dictionary it's defined in terms of other words in the dictionary so this book kind of seems somewhat circular right is the way out of that that there are certain words that somehow are so basic and so uh within our need to know and just to live in the world that we understand them without a dictionary and that's what breaks the loop is is that where the answer comes from uh abs yes absolutely part of the answer and this is connected to your answer is that of course some words in the dictionary have little pictures next to them uh so that that breaks the circle uh also because dictionaries are intended for human users who bring all of their intelligence common sense expectations to bear when reading the dictionary you could they uh we draw on a lot of our tacit knowledge in figuring out um what a word means from its definition in those cases where we have to look it up in the first place the vast majority of words of course we learn from experience but yes we hear a word in context we uh now there's an old theory that it's just raw association that like almost like pavlovian conditioning you hear the word you see the uh the action and so word action word action or word object word object chances are that's that's not the way we do it uh for example when a child is told to stop doing something then you know the word stop can't refer to what the child is doing by definition the the extra level of sophistication is that we use our intuitive psychology our so-called theory of mind referring now to not a scientific theory but an intuitive theory to make a reasonable guess as to what what the speaker is probably trying to get across and then we uh solve the puzzle of what sound it coming out of the speaker's mouth or our hands in the case of of signing probably is what the speaker was trying to convey so there has to be enough redundancy in the situations especially in the case of of uh children that they can that language is partly redundant redundant enough that they can crack the code and figure out what those words mean but indeed not only sensory experiences just you can't define red at the end of the day red is just you know what what what triggers your long wavelength cones and likewise there may be a a kind of almost an alphabet of elementary concepts uh what what computer scientists sometimes call an ontology uh in the case of representations of the basic ideas and relationships that are form the basis of all of this metaphorical stuff that we can then elaborate so maybe elementary things like an object a causation a place a path and there are linguists like ray jackendoff i've tried to to do it myself to lay down what is this inventory what is at the end of the day the language of thought which in a sense is not itself language but it is the system for understanding and reasoning about the world and each other that language then externalizes so so we spoke a little bit about dead metaphors we spoke a bit about metaphors and analogies as they play a role in science where it's a very specific tool to allow us to say go from one level of understanding to a different level of understanding but what makes a live metaphor a metaphor that just we we we look at in awe at the author having written this phrase about the world what is it about those kinds of metaphors that just move us as emotional beings yeah they it um excellent question at the intersection of um cognitive science including linguistics and you know poetics and and literature what makes for a a brilliant metaphor it's probably a combination of its aptness in terms of the relational structure and this is partly what makes for a good scientific analogy and here i'm drawing on the work of the cognitive psychologist dendri gentner that the what makes a the flow of liquid a good analogy for thermodynamics is not that you know water can be hot in fact that you got to forget that water is hot uh you've got to have abstract concepts like a a potential difference and flow and uh to the extent that the structure of what causes what and what enables what and what prevents what if all of those abstract relationships carry over from the thing you're describing and the thing you're reaching for that makes an analogy or a metaphor feel deep in addition that's that's what makes a good scientific analogy what makes for a good literary metaphor is if in addition the the affect the emotion conveyed by the uh the vehicle or the uh uh sheds light on the emotion that we feel toward the the target or as they call it the the tenor uh so the the emotional connection is crucial and there are uh i i illustrate this in the stuff of thought uh by its opposite the bad metaphor contest from the washington post that they had a number of years ago where people would submit humorously inappropriate metaphors like he uh he took his life and jumped out the window and hit the ground like a hefty bag filled with tomato soup now it's kind of in conspicuously bad taste and it is a bad metaphor because that's precisely not the kind of connotation that you want um in conveying something as precious as a human life uh so did he say you can understand something by seeing what it's not those bad metaphors uh underscore what is wrong with a metaphor like that whereas something like i am the you know nabokov i am the shadow of the waxwing slain by the false azure of the window pane where he's talking about the loss of uh his own uh emotional uh crushing evisceration of the death of a daughter analogizing it to a bird that flies into a pane of glass where they're the patheticness of the the pathos of the the bird something that we feel as beautiful as fragile as delicate being um it was misled by a a transparent window and seeing blue on the other side uh there's something so poignant about a blue sky a beautiful bird leading to its death and so not only is the content uh transparent but every one of the emotions elicited by the metaphor the azure the the the falseness uh the shadow uh each one of them triggers an emotion that's appropriate to what's being uh conveyed in this case the the death of a child well i was i myself was moved by another choice that you made in illustrating powerful metaphors and the stuff of thought i recently reread madame bovary i'd read it of course as a student a long time ago and there was one passage in in my copy that i'd heavily underscored i don't underscore a whole lot and then i saw that passage in your book so i thought it was fitting to end with this passage which goes no one can ever give the exact measure of his needs nor of his conceptions nor of his sorrows as human language is like a cracked tin kettle on which we hammer out tunes to make bears dance when we long to move the stars and so with that stephen pinker thank you so much for joining us i enjoyed this conversation on language and uh thanks a lot always a pleasure thanks for having me brian thank you four distinct views on language from four of the foremost thinkers on the subject language as a tool of inner thought language as a socially constructed means for instructing the imagination of others and language as the natural evolutionary outcome of humankind filling the cognitive niche now there are profound differences between these perspectives but there is also common ground the fact that we humans are able to use language to understand so much about the cosmos and about ourselves that to me is just wondrous i'm brian greene thank you for joining us and join us again for future episodes in this world science festival series focusing on the big ideas of science [Music] do [Music] you

Channel: World Science Festival

Views: 849,617

Keywords: Brian Greene, Noam Chomsky, Steven Pinker, Daniel Dor, Evelina Fedorenko, science of Language, Is language a human invention?, Mind Your Language, Metaphor, Thought, Imagination, Linguistic science, Cognitive science, language and the brain, analytic philosophy, language philosophy, How language shapes the way we think, origins of language, Linguistics as a Science, How Does Language Change Your Brain?, World, Science, Festival, Big Ideas Series, New York City, NYC

Id: 6LXHtDUXkS0

Length: 111min 57sec (6717 seconds)

Published: Thu Sep 30 2021