table of contents


Martin Lewis and Asya Pereltsvaig on the Origins of Language

Martin Lewis is a Senior Lecturer in International History in the Department of History at Stanford University. He studied at UC-Santa Cruz and UC-Berkeley, receiving his PhD in Geography in 1987.  His dissertation, and first book, examined the interplay among economic development, environmental degradation, and cultural change in the highlands of northern Luzon in the […]

download transcript [vtt]
This is KZSU Stanford. Welcome to entitled opinions. My name is Robert Harrison and we're coming to you from the Stanford campus.
You're not a bother.
Preparing for today's show, I check 20 different translations of the opening verse of the gospel of John and every one of them translates the Greek word logos with the English word word.
In the beginning was the word and the word was with God and the word was God.
Now what does that mean? The barn's notes on the Bible says I quote, "The meaning is that the word had an existence before the world was created."
Great, but what does the word for word have to do with the world that came into being?
Even if we take it on faith that in the beginning was the logos, no one really knows what logos meant in the beginning because by the time our words begin to mean something, they already have a past, they already reach us by way of our four mothers and four fathers.
Regardless of our dialect, we speak with the words of the dead. That in a special sense is how and why we speak at all. Just as humanity begins where there is already an ancestor, so language begins where it has already begun.
A language is natural to the degree that its native speakers are born into the generative source from which the saying power of its words springs.
That means in turn that the origin of our words lies not so much behind them as in them.
And that in turn means they contain within them humic underworlds where the dead holds sway over the converse we hold with one another.
To descend into the underworld of words, philology needs the golden bow that allowed a nius to go down into Hades and consult with his ancestors.
If words are genetically related to world, then Heidegger has a point when he declares in the context of his lectures on Nietzsche in 1936, I quote, "to relegate the animated vigorous word to the immobility of a
particle, mechanically programmed sequence of signs would mean the death of language and the petrification and devastation of dasa."
Our basic words he goes on to say possess a plurality of meanings that derives from the plurality of prior worlds through which they have passed on their way down to us.
The word "bildung" meant something different for good they and Hegel than it did for the educated person of the 1890s, says Heidegger.
Not because I'm quoting him here, the formal content of the utterance is different, but the kind of world encapsulated in the saying is different, though not unrelated.
He goes on to affirm the following about the historical nature of our words, quote, "That does not mean simply that they have various meanings for various ages which because they are past we can survey historically.
It means that they ground history now and in the time to come in accordance with the interpretation of them that comes to prevail.
The historicity of the basic words understood in this fashion is one of the things that must be heated in thinking through these basic words."
I have with me in the studio two scholars who have spent a lot of time thinking about the historicity of our basic words.
They hold in their hands the golden bow that opens up the underworlds of human language.
Martin Lewis is a geographer who teaches in the Department of History here at Stanford.
He has been my guest on entitled opinions before. I did a show with him about geography a year or so ago.
Joining Martin today in the studio is Asya Pearltsvig, a highly accomplished linguist who teaches in Stanford's Department of linguistics and who specializes in the study of syntax, particularly in Russian and other Slavic languages.
He is currently working on Tatar, a Turkic language from the Volga region in Russia.
Martin and Asya co-taught a course this year on the major human language groups and they are presently co-authoring a book on the origins and expansion of the Indo-European languages which will be published by Cambridge University Press sometime soon.
They have kindly accepted my invitation to talk with us today about some of the problems entailed in the search for the origins of language.
Thank you. Thank you.
I have a copy of the proposal of the outline that you have for the book that you are co-authoring at the moment that will be published as I mentioned sometime soon.
We hope in any case.
You are right there that the initial impulse for it came from an article that was published in the New York Times back in August of 2012 where a group of biologists presumed to have solved the fundamental puzzle of the Indo-European language family.
They had written a paper in the journal Science where they claimed using computational techniques borrowed from the physical sciences to have located the point of origin from which the largest family group, language family group in the world, Indo-European, 46% more or less of the world's population speaks an Indo-European language that they had located somewhere in Anatol's language.
And the assumption there was one singular point of origin from which there was all this dissemination.
And both of you have a lot of problems with the methods that they use and some of the conclusions that they arrive at.
Can you just briefly reconstruct a little bit what your issues are with this new brand of research?
There are a number of problems. One is that it goes against the research that's been done by most historical linguists and archeologists who have found many lines of evidence that suggests that the Indo-European language family originated in the so-called Pontic Steps, the Grasslands North of the Black Sea, among pastoral peoples or nomadic horse riding peoples.
This would be about some three to five thousand years ago, whereas the article in Science suggested it was farmers in Anatolia much much earlier in the Neolithic period.
So they go against the bulk of evidence that has been patiently accumulated for decades, really centuries.
They really ignore that evidence. They don't address it at all.
Instead they take wordless of cognates, words that are related, and treat them like a biology retreat, proteins or genes, plug them into extraordinarily sort of complex mathematical program, and then return what they think are likely lines of descent and areas of descent.
But the problems are manifold from the words they choose to the notion that language evolves the way species do.
Really we think that the processes are entirely different, partly because language is shared so much more than species do.
Species can share a few genes can be moved around by a virus, but that's very rare. Languages are borrowing massively all the time.
So that's just part of it.
So maybe I can ask you the new methods that they're using the biologists.
You would think that as hardcore scientists they would be the most empirically oriented.
But is it your view that they are actually ignoring a great deal of empirical evidence that has been amassed by archaeologists and linguists and geographers like Martin and are actually operating on the basis of a theory rather than on the basis of the real evidence?
Absolutely. As a linguist I think the biggest problem I find with this line of research is that it focuses exclusively on the lexicon on a small selection of words as opposed to looking at the language as a whole, and in particular looking at sound changes and grammatical patterns of language, which are much more reliable as evidence of language descent and language expansion.
So however we're dealing here with dead languages that have not been spoken for millennia, and yet, so how does one deal with a whole language when you have such only, I'm presuming do we only have fragments of that proto language at our disposal now?
Absolutely. But as you said in your introduction the really cool thing about historical linguistics is that it can tell us something about human past,
based solely on the languages that are spoken today. So we can examine in the European languages that are still spoken from Irish to Hindi, and we can draw conclusions about proto in the European, what it must have been based on what it developed into over these, well we think about maybe six millennia, but we can make those conclusions without it.
Without even necessarily a making reference to extinct languages, hit tight or latin or Sanskrit, and we can do this, and we can look at grammatical patterns in languages that are still spoken to do that.
So your specialised syntax, the study of syntax, is proto-European, are we able to reconstruct something like a syntax or a grammar for proto-Indo-European, or do we only have conjectures about a common vocabulary of words?
There are some, there's some work on reconstructing the grammar of proto-Indo-European, in fact some scholars seem to be able to converse in it in the reconstructed form, or compose stories in the reconstructed form, but it is much harder to do because for centuries the work indeed focused on sound changes and vocabulary changes looking less at the grammatical side of things, especially at this syntactic change.
So Martin, you're a geographer primarily with disciplinary formation, and these scientists that you're kind of a nerve you in their methods, you write there that they employ these novel paleo-geographical methods developed to trace the spread of viruses along with the previously used phylogenetic methods designed originally to trace lines of biological descent.
So when I see the word "filog geographical methods" I would assume that a geographer would actually sympathise with the use of this method, but I take it that you don't.
Well I think it's a very simplistic model, it's based on the diffusion of viruses, it's based on the notion of diffusion, and diffusion is a physical process that can be modeled through Brownian motion or a random walk model, basically molecules bumping around into each other.
So they hypothesize the languages, language, family, spread by "demic diffusion" in other words, groups of individuals who speak that language have a population, that population is expanding, and individuals are randomly moving and in that process the language grows, the area covered by that language.
I think that's wrong for a number of reasons, first of all people don't move in a random fashion. Now our opponents would agree with that but say it nonetheless can be modeled as such.
But actually people often move over long distances in my grational patterns, but more importantly languages often spread not by a "demic" process at all, but rather by a process of language contact people are, people are all the time losing their languages and adopting those of other people.
So there's no biological movement of the people, it's simply a movement of the language, although usually it's some complex combination of the two processes.
Sure, so I gather then that one of the, and this is something that has afflicted the study of Indo-Europeanism from the inception of the early philology, which is an over-easey,
an over-identification of linguistic family with a actual race of people, no? And you're saying that the diffusion of language is not reducible or is not identical to the diffusion and dissemination of peoples across the...
So that geography, where you find linguistic diffusion does not necessarily mean that you find the racial components going along with it.
Right, that's a problem that has the long be-deviled historical linguistics, it doesn't anymore but the possibility is still there.
So the Nazis identified the Proto-Indo-European with areas and thought it was a racial group, and this has been done with actually many other language families, and the class we like to show atlas for the
As late as the 1960s, the depicted fins and Hungarians and Turkish people as racially Mongolian.
And the reason they did that is because all those languages were at that time classified in the Ural-Alteic language family, which included the Mongols, hence they were all thought to have been descended from one group.
Well, we know now that the Ural-Alteic family doesn't really hold true, but even of language family it does, like Uralic includes people who look very East Asian and people who look very European.
So obviously the languages have moved independently of the genetic flow and the population flow, if you will.
Great, so if I can ask you about, you said, Martin said the Ural-Alteic has lost a lot of its credibility as a thesis, and yet there does seem to be indeed some kind of syntactical or linguistic commonalities between these weird Hungarian Turkish and Finnish.
Geographically, it's a completely unlikely place to find commonalities, but linguistically is it just by accident that there are these commonalities or are the commonalities not there at all?
The commonalities are certainly there, so for example the languages you mentioned all have this phenomenon called Val-Harmony.
So in these languages each word or each root has to have vowels that are either all pronounced in the front of the mouth or all vowels that are pronounced in the back of the mouth, but you can go back and forth and have sort of mixed and match as we're perfectly happy to do in English and other languages.
So commonalities are certainly there, and Finnish and Hungarian are related even if distantly so the members of Uralic language family.
But Turkish is now thought to be a member of a separate Turkic family, the biggest member actually.
And the commonality is probably shared through not common descent but rather borrowings and contact between the languages in a sort of great Eurasian zone.
One thing if I can add in that zone in sort of central Eurasia you had a lot of people who were migratory pastoralists who formed large multilingual confederations.
I mean the Mongol Empire would actually be a great example which probably had more people who spoke Turkic languages than Mongolian in that.
So there was just tremendous interchange of all sorts of linguistic features over a large area and a long period of time.
Right. So, and I say you were saying about the vowel formations but there's also I was always told that there was the fact that they these three languages are a glute native in their grammar and we don't need to go into that because I kind of might get too technical for our listeners but nevertheless there is more than just the vowel formation that's.
Absolutely. There's a lot of commonalities and these are the commonalities that first made linguists think that they belong to the same language family.
But there are many agglutinative languages outside of Eurasia in the among Native American languages, our regional Australian languages, Papuan languages, African languages.
For example Swahili spoken over a large spoken as a lingua franca in East Africa is an agglutinative language but nobody that I know of proposes that it's a member of the same language family as Finnish or Turkish or Hungarian.
So they are certain large design features like a glute native morphology or like core word daughters. So many know all of these languages are subject, object, verb in their word order.
But these sort of design features are found all around the world and they usually do not correlate the language families. So they're not the right type of phenomenon to look for evidence of a common descent.
Okay, here I'm going to begin a series of naive questions from the person who doesn't know about the deep history of languages. So I read on these typical low-life sources like Wikipedia, I'm a big fan of Wikipedia.
I love Wikipedia too. So I want to retract that. However, sometimes you don't know whether it's a misprint or something like that.
When I read or read and preparing for today that the Indo-European languages are originated.
Some are started getting to fuse somewhere between four and six thousand years ago according to one theory which is the one about the Pontic steps.
Whereas the old Renffrews, one of the great scholars of Indo-European, he thinks it has an older origin because it takes place in Anatolia through Neolithic farmers and not through the pastoral.
So I kind of wore your patriarchal society. So when I read four to six thousand years old and I think, okay, here we are, I go back to Europe.
And I know that there are caves with cave paintings that are 37,000 years old and that these were Homo sapiens and were obviously language speaking.
Neanderthals, there's a lot of debate about whether Neanderthals had the actual apparatus vocal and apparatus for speeches.
We know it, but nevertheless they also predate in Europe.
So Europe has been populated for many, many, many millennia before the spread of Indo-Europeanism.
What on earth were they speaking in Europe if they weren't speaking Indo-European?
Well, we don't know, but we have some interesting evidence because Indo-European didn't take over all parts of Europe.
One important language, Basque, which is considered an isolate, not related to any other language, spoken today.
For example, Persists, for a long time, persisted in a much larger area.
There are a lot of ancient languages that persisted up until the classical period.
So, I think that was a little bit more of a bit of a little bit of a little bit of a little bit of an idea.
So, there were many different languages that were most of the languages that were in the world.
So, there were many different languages that were in the world of the European Union.
But, there were many different languages that were in the world of the European Union.
So, there were probably many different language families.
People speculate there's something they call the "tearsennian family" that might have included a number of these.
But, we don't have enough information to go on to put them together in larger families, but we certainly know they were there for a long time.
So, people were speaking a number of languages.
So, Europeans came in and those pre-existing languages may have really influenced the Indo-European languages that did come in.
This is quite controversial, but Greek, also the Germanic languages, may have a large substrate of words and even grammatical forms that may have reflected the languages that were spoken before in the European spread.
I was always amazed, I'm always amazed at Thalasad, which is the Greek word for "c" is not actually an Indo-European word at all.
Almost no maritime terms are Indo-European.
Building terms, a lot of political terms, we just cannot reliably reconstruct Proto-Indo-European terms.
So, here's another naive question maybe asked you to answer that, which is, is it really plausible that you have a small group of people because, you know,
the Pontic steps, you tell me where it's located, it's north of the Black Sea, it could be, it's not the cradle of civilization, it's not where the Neolithic revolution began, it's not where you have the first earliest artifacts or anything of that sort.
Is it credible and plausible that a tribe of people, you know, speaking this proto-Indo-European, could have conquered the entire world, you know, all the way to Western Europe, you know, the Eastern Europe, the India, and it's a vast range.
So, either these people were a super race with their chariots and they conquered, you know, ruthlessly, which is usually, you know, the kind of stereotypical image that you get of the Dorian invasion, you know, these people on horseback terrorizing indigenous populations, or there has to be some other mechanism by which one particular language in a provincial area of the earth became so dominant.
Absolutely. And I think this sort of mechanism is intermarriage. So when you have man and women from different linguistic groups getting married and having children, it's you, well, usually it's the language of the fathers that's passed on, but the mothers also have a say and sort of modify this language as it goes on to the next generation.
But it's in these kinds of inter-linguistic communities you get transmission of language. And to sort of illustrate how this could have worked.
Just imagine we did this show 2000 years later, 6000 years later, and we were looking into the past and sort of imagining how Russians could have spread from a very small area around Moskovy.
And it wasn't, you know, the cradle of civilization, it was a small group of people sort of embattled by people who spoke a different language from a different family, from a Turkic family, and they managed to spread over a vast territory.
If you look at the map of where Russian is spoken today, well, that's pretty huge. It's not quite as large as in the European, but it's a vast territory that extends all the way to the Pacific coast of Eurasia.
And if I might add a little bit to it, it used to be thought that the proto-Indo-European spread in this organized invasion, very bloody conquest, the contemporary thinking it was probably a much more gradual process of incursions of small scale, because they didn't have a state, they didn't have an organized army.
But there are actually many cross-cultural examples of languages spreading in that manner. For example, Semitic languages were associated with non-horsewriting pastoralists from the Arabian Peninsula that moved into the fertile crescent, supplanted Sumerian and other languages. Even in North America, we know that the Comanche language was spreading rapidly, because it became a trade language. Other people picked it up, because the Comanche didn't necessarily conquer in rule them, but they were kind of domineering.
And also, we're very important economically, and so something may have been happening like that.
So to ask a few more technical questions, here about linguistic diversity.
So how many major language families are there? Do we call them language families, or is that politically incorrect? Do we have to call them language groups, or can we call them families?
Oh, we can call them language families. The problem is linguists use the term language family to refer to groupings at different levels.
So one can talk about English being a member of West Germanic family, or of a larger Germanic family that would also include Scandinavian languages, or a member of Indeuropian family that would also include Russian and Hindi.
So it's sort of hard to answer that question, because it depends. If you want to look at the highest order families that we're talking about, maybe around 100 families, but they also have these groupings within them. That link was also referred to as language families.
So the biggest ones, after Indeuropian would be like Sino Tibet, what are the biggest ones after Indeuropian?
In the number of speakers, Sino Tibetan is the second largest, and Indeuropian is the largest. But if we estimate the size by the number of individual languages within a family, then Indeuropian is only fourth largest, I believe.
And Nijir Kongo is the largest, and Austronesian is the second largest.
The third largest is Trans New Guinea, found only in New Guinea, and one of a number of families only in New Guinea.
And is that because of the isolation of the different population groups?
Well, it's kind of a mystery why New Guinea has so much linguistic diversity. One thing is humans got there at a quite early period.
There's a large plateau in the central part of the island that is almost separated from the rest of it, and was largely not entirely, but largely isolated for millennia.
And these very small scale, but relatively densely populated agricultural populations.
And it ended up in with this just extraordinary diversity.
Okay, this brings a question about diversity, then something I need. I know that I asked Asya this before, you know, when we talked about this.
But the, I don't know if it's in the biologists that you were engaging with, but the idea that at the origin, one way of tracing on original languages, phonemic diversity.
And when you start losing phonemic diversity, that means that a language has started to disseminate itself around.
So why would there be more phonemic diversity at the, at an earlier stage in the development of a language rather than less? Is that a mix sense as a question?
It sure does. The problem is that this is actually a claim that comes from the work of one of the authors of the science paper, who in a different paper also in science, claimed exactly that that language originated in West Africa.
And that's where you get the most phonemic diversity, the most different sounds in any language spoken in that area. But as you get further and further away from that location, you get languages with smaller and smaller inventories of sounds.
The problem with that claim is that actually is just simply not true factually. There are languages that are outside of West Africa, even outside of Africa that have a lot of sounds.
So for example, languages spoken in the Caucasus regions are known for their very region inventories of consonants, a language that is now sadly extinct called Uber.
It's last speaker died in the 1990s. The language, Uber had something like 84 different consonants about twice as many as English has.
So you can find a lot of phonemic diversity outside of that proposed origin area in West Africa.
And if I may add it, it is true that Africa, though actually Southern Africa, not Western Africa has a huge inventory of sounds because they have all of these clicking sounds.
And we used to think that that was one language family, the hoist on family, but now we know there are four maybe five separate families there.
They all have these click sounds that can be articulated in many different parts of the mouth. They all sound the same to me, but they're not the same at all.
Well, here among the speculators on the origin of languages, it's been speculation for thousands of years, Plato has it and Dante has the whole idea of a dynamic language and where there's this one origin, it's one speaker.
And it sounds crazy, but the more I started thinking about it in relation to today's show, the more it struck me that with all the diversity of languages and diversity within particular languages, phonemic and so forth, that the study of the origins of language is really when you push it to an extreme, it becomes belongs to the domain of natural history in so far.
As we're talking about the evolution of a species, the Homo sapiens, however much diversity there is linguistically among the human race, we're all part of a species and that that species has its own evolution and we know that it seems highly, highly likely that Africa was a point of origin for the spread of Homo sapiens around the globe.
And that therefore, it doesn't seem so absurd after all not to take the academic language in the literal sense that there was an original couple that spoke, but that there was a family of people who were speaking something like a common language and from that point of diffusion it went around the world.
Now, I know from your outline that this whole assumption that there's some of these same scientists have tried to pinpoint the one point from which all of human languages originated.
I don't know where they claim that is, but this seems given what I said in my introduction, a pretty hopeless sort of project because I believe that the origin, especially when you're dealing with language, is that the same thing.
And I'm not even speaking as a deconstructionist where you have the Doridian dogma about the decimines cian de d'Ochizin and all that stuff.
But it is from the point of view of natural history, it's not absurd to imagine that there could be a point of origin for language.
Yeah, I think most historical linguists believe that there was one single origin. Now, that's been long debated.
Did language originate gradually in which case you may have had many languages sort of evolving or was there a major mutation in certain genes that happened once this gnome tombsky's view, by the way, as it was.
A huge change that reoriented the human brain made language possible and then you would have had one language that started almost certainly in Africa and then disseminated outward from there.
So that is sort of the basic view. The problem is though, is that this would have happened 50,000,000, maybe 150,000 years ago, maybe even 200,000 years ago.
And after about, oh, 10,000, 15,000 years of language change, everything has been shifted around so much you can't really reconstruct past that.
The sounds of change, the lexicon has changed, the grammatical patterns have changed, and it's like it's shifting a deck of cards and after you shuffle it some many times, you can't trace back.
Unlike, say, species where the farther back you get to the root, the more clear those relationships become with language after a certain amount of time.
It's just like, "Sans" or your hands, you really can't grab onto anything.
Now, some people think you can't, some people think that you can take a list of words and put them in your computational machinery and come up with some sort of proto-human language, we're skeptical.
If I can add to that, indeed, the current thinking is that language originated only once, and as Martin was saying, historical linguistic tools are not really adequate to figuring that out or answering this question.
But there's a different way that linguists are trying to answer this question now, and that is by looking at the shared commonalities that are these big design features of language that seem to be in disshared by all the diverse languages spoken by humans.
And by looking at these patterns, we can see that languages are not just any conceivable way, but there are certain restrictions on what a human language can be like.
And even though this is not a conclusive proof, it's certainly an argument for there being one original language from which all other languages developed and every human language is being sort of a variation on the theme, if you will.
So, I guess I gather that you're a chomp skin of sorts, and that means, rather than me saying what it means, why don't you tell me what it means to be a chomp skin in this particular point of view, namely a commitment to kind of universal human capacity or gene for a certain grammatical understanding.
Well, yes, that's, you can say, I'm chomp skin of sorts, and the views that I share with that Chomp skin platform is exactly that human language has this sort of common blueprint.
So all the different languages of the world have a common blueprint, the all variations on a common theme, and we're very far from understanding exactly what that blueprint is in all of its details.
But for example, all languages seem to have consonants and vowels, all languages seem, although different ones of course, all languages seem to have verbs and nouns and maybe adjectives even though that's more controversial.
All languages seem to be recursive, again, that's somewhat controversial, but there seems to be some big design features that are shared by all human languages.
And it's understanding what exactly the common features are that's the research program.
So if one were not a chomp skin, what would be the most important counter view to the chomp skin view that there's a universal substrate to all human languages?
Well, the alternative would be that human languages are not restricted in what they can be like.
So in theory, they can be a language that was just vowels or a language with verbs but no nouns or anything of the sort.
So that would be sort of a belief that there's no universal shared common core, which also according to the chomp skin view is transmitted genetically in some shape or form.
And of course, it's very difficult to flesh out the details because even though at this point we think that a gene called FoxP2 is a gene that's responsible for language, but we don't actually know how it does its job.
So we know what the sequence is, we know what proteins are supposed to be produced by this gene, but we don't know how these proteins translate into anything that linguists can understand really, things like word, order, sounds and things like that.
We actually think about that gene as other animals have it and it's linked to their vocalizations as well, but if a human has a defective copy of the gene, they're not going to have fully articulate language.
Well, here, Mark, you're a geographer and I know that you'd like to stick to the hard facts, but let me give you the speculations of the French philosopher Jean-Jacques Rousseau who wrote a little treatise called the origin of human language.
I didn't realize that. He did and it's a rather sophisticated piece of argumentation that is also a theory of music because he believes that the human language has musical origins in the sense that, and if I get him right that
In the southern climbs and the southern regions, the first languages would have arisen and they would have arisen out of a sense of a love of life or certain passions.
No, Rousseau is always looking for a passion as the origin of things. And he's not alone and there's also, my philosopher was very dear to me, Jean-Bati-Stavico, the 18th century Italian theorist who said that language, human language must have arisen under a very strong violent emotions, either joy, grief or something of that sort.
In fact, I think there's a lot of evidence that grief has played a certain kind of role in the origins of language where you have to find a way to control a otherwise chaotic passion by certain vocalizations and techniques.
But I'll leave that aside, Rousseau's theory is more simple that human language was originally only vowel sounds and that through the intonation and the musicality of the vowelization, you would communicate what we would call semantic content.
He imagines, you know, boy meets girl at the well and says something that translates into French as "emmois love me" but it wouldn't have had any consonant, say it wouldn't have had a grammar, it would have just be a vowel sound like an operatic sort of bird song.
And this is quite interesting, where would consonants have been introduced, well, they go to someone like Hegel and it's the awareness of death and what do consonants do?
And the consonants arrest the melodic line of a thing and it introduces the negative, this is the power of, you know, it introduces the death principle into the life of the melodic line and therefore, especially the northern climbs, as Rousseau says, they're very harsh with their consonant, all the things.
What are we to do, someone like me who teaches these texts as a cultural historian, are these just fanciful or is there some way in which linguists in a more professional mode can retrieve and recuperate something actually that has some insights there that are worth revisiting?
Well, they certainly are imaginative texts and they're important for intellectual history, as far as the origin of language.
I mean, it's so highly speculative at one time the French linguistic academy declared henceforth we're not going to talk about the origin of language because of some of these issues.
Now, we know that a lot of non-human species, not just apes but even ground squirrels, have fairly sophisticated vocalizations and you can analyze these little chirps of these ground squirrels and one will mean hawk and another might mean ferret, so there definitely is meaning there and they're connected with emotion fear, I suppose, in this case.
So I think it's quite plausible that human language would have had an evolutionary process like that. Now, whether or not there was one big breakthrough where there was some sort of massive genetic change that allowed fully articulate language, perhaps.
But there's a big gap between sophisticated vocal sounds that indicate emotional states vis-a-vis actual language which as Asya said has these qualities of recursion you can embed clauses and clauses and clauses and basically go on forever.
I mean, that's a new thing under the sun, I suppose you could say.
If I may add to that, indeed the different species have their patterns of vocalization or their communication strategies, but they're different from what humans have in a very important way and that's that they're very limited there, their finite.
So ground squirrel cannot produce the kind of imaginative attacks that Rousseau did. The reason being that they can only, they have a limited set of vocalizations and they cannot operate on them and reshuffle them in infinite ways the way humans can.
And also, if I may add to that question about Rousseau and what's fanciful and what's factual there. I think these texts from European philosophers, they are very informed by European languages, the way that a very small subset of human languages happens to be.
So I think if Rousseau was familiar with any of the so-called "Hoisan languages" that have upwards of 100 different sounds including many collects consonant sounds that is, or even Ubech that I mentioned earlier with its 84 consonants, it would not necessarily occur to them to say that consonants tend to cluster in northern latitudes because of course these are fairly close to the equator.
So, well, close to them, French, I would say. So it's kind of interesting that people, scholars of European origin, use to produce these theories of human language, they're very Eurocentric in a way, but not very applicable outside of that small group of languages.
>> Well, without getting into the fanciful speculation, because I think the French Academy
did well to abolish, you know, debates about the origins of language but here's a more philosophical,
a matter of – I don't want to call it metaphysical, but, is it – is language just a tool,
a pragmatic tool of communication among humans, or is – would – the human language
as we know it have been possible in a species that did not have an intense and terrifying
awareness of death and mortality.
Is death or the sense of an awareness, some self-awareness of being in time and being in the
mortal thing, where we know that the anthropologically and archaeologically, where you define
the human, not as a species, but as a cultural life form, is where you have burial of the
Because a burial of the dead indicates your awareness of mortality and your ritualizing
the passage.
Is that awareness that is responsible for things like burial of the dead also somewhere embedded
in the very possibility of language as such?
Yeah, I wouldn't be surprised at all.
Partly, it's just a matter of cognitive capacity.
Once cognitive capacity reaches a certain level, then I think awareness of death is going
to loom large.
Now there's some speculation that some of the more intelligent non-human species have
some death awareness, but we can't really say for sure and it certainly aren't going
about funeral rituals and the like.
There's some evidence that Neanderthals did some of that, but very little compared to
homo sapiens sapiens.
And again, we don't know if Neanderthals had full-articulate language or not.
It used to be thought they didn't.
Now the evidence is a little more uncertain.
So that's certainly an interesting way of putting it.
Yeah, I think that there might be some connection and that connection is symbolic thinking.
So when humans bury their dead, there are some elements of symbolism or symbolic thinking
in that.
And symbolic thinking is really what is thought to be associated with language of being
at the core of human language absent in the communication systems of apes and bees and dolphins
and other species, although dolphins are kind of, if you eat dolphins are very human-like.
But I say, I'll tell you why I have problems with the word symbol or symbolization is
because a burial of the dead, they're not engaging in symbolism the way we tend to understand
They had literal beliefs in the continued afterlife of the dead in some other, sometimes
literally under the ground, otherwise they wouldn't be bringing them food and these
funeral repasts and so forth.
And the fact that the dead can carry on and after life in the world of the living, it has
to be related to the fact that all are words.
Almost all the words that we speak were spoken by our predecessors and by ancestors and
by the dead and that we inhabit the world of the dead insofar as we're using their words,
maybe borrowing their words.
Our words are on loan to us from the dead and therefore symbolization, I think, tends to categorize
it in a way that doesn't do full justice to the pre-inhabitation of our human consciousness
by the predecessor, the ancestor, and perhaps the pre-inhabitation of language by the previous
speakers of that language.
And there are theories, poets I could invoke them that when a native speaker hears a certain
word, I'm thinking of Leo Padd of the theory where he pointed out certain words which
have Latinate roots and when a native speaker, a Italian someone who was born into the
language, one means of a native, that it resonates with deep time and there are associations
that go back millennia in the psyche that are evoked by the word no to nor the word
said or something like that and in English it would be a different set of words and the
poets are the ones who know what these words are because they come with a deep history that
is related somehow to that world of the dead.
Yeah I think in the sort of the Bardic tradition in many parts of the world brings that
out very clearly where poets are sort of the guardians of that kind of collective consciousness
if you will, that communion with the dead.
I certainly find one of the most interesting things about studying historical linguistics
is precisely this taken into the time and to some extent even into the minds of our predecessors
long ago and to be able to see how linguists have reconstructed words in Proto-Indo-European
and see how they've descended in many different languages in different forms but have
these common roots.
I find it sort of intellectually thrilling but it very much speaks to what you've been
talking about.
Yeah in fact this Vico who mentioned earlier in my introductory remarks I'm playing on
the word and world he said words mean something in the particular worlds that were inhabited
and he thought that there was a development or evolution in worlds he begins with the premise
that this was the order of human institutions.
First the forests then the huts, the villages and then the cities and finally the academies.
Then he takes one word which is the word "lex" and he gives a kind of conjectural reconstruction
of what the word "lex" would have meant at every stage of these different worlds and so
in the forest it would have meant a collection of acorns.
In the huts it would have been the collection of vegetables.
People laugh when they read this in the villages it would have been the other thing.
Finally you get letters or you know all because "lex" means to bind together and what Vico
is trying to do is what would to bind together have meant in a particular environmental context
or in a particular world.
If you're living in the forest the acorns gather the acorns rather than gather letters
together and create words.
So the genetic research in words and world is where I agree with what you were saying
earlier that this is a fascination of historical linguistics which is what kind of world
the otherness of the world that our predecessors inhabited are being kept a memory of it is
recuperable through the study of language.
Yeah it's interesting one of the tools that historical linguistics use this is something
that's called linguistic paleontology which is to try to reconstruct the environments
that a group came out of based on the words that they had even the social order and this
is a very fruitful field I would add though that the biologists who we've been criticizing
don't buy it one bit they totally reject that notion because they don't think it's
rigorous enough although I think historical linguistics have shown that it's not mathematically
rigorous but it's pretty solid in many cases.
So let me ask both of you a question with the publication the future publication of the
book that you're working on.
What will change in our understanding of the origins of Indo-Europeanism if anything in terms
of are you proposing a theory are you because when Renfuse book came out and was in the 80s or
early 90s?
Yes that was like now we have a kind of standard gold standard for the study of Indo-Europeanism
and now you're talking about these biologists and so forth but you say in your outline
that your project is setting forth an empirically credible account of the ways in which
languages actually originate, diverge and spread across landscapes with special attention to
the Indo-European family.
So in the best case scenario at the end of your project what will you be proposing a new
theory about how these languages actually originate diverge and spread across the earth?
Well no I'd say we're not doing that for Indo-European it's been done by others there's
an archaeologist named David Anthony who wrote a award-winning book called the Horse the
Wheel and Language just a few years back and he surveys the evidence of the origin of
Indo-European languages exhaustively and so we really don't need to add to what he's
saying but we do want to put a lot of these issues sort of on the table in a very clear
form about how languages spread, how they interact with each other to show that the kind
of what seems to us simplistic biological models just simply can't account for that.
We think that language really has to be taken on its own terms and not put in sort of
a reductive set of procedures drawn from another discipline and a lot of this is going
to be very specific to the Indo-European languages but a lot of it will also be much more
In the way what we're doing is in this in this book project it's saying there's a lot
of theories always already proposed, already on the table let's just stop and weigh the
evidence and also examine our own procedures in how we do that.
So part of our concern is epistemological we want to see how do we know certain things
what are the ways that we use to examine the evidence and what sort of evidence really matters
for reconstructing human past.
And interactions between different language groups is really key.
Well I want to applaud you for this project because what you're doing in the realm of
linguistics and geography is what phenomenology is really all about and when phenomenology
began the famous cry of the phenomenologist or the slogan was back to the things themselves
namely get out of the discourse all the presuppositions and assumptions and theories about
why something must be the case which always leads to reductionism because you're always
looking you already know what the result of what you're looking for.
The key, the thing itself look at it exhaustively from different points of view and let
it actually reveal itself on its own terms and it sounds to me like you have a kind of
exemplary phenomenological approach to your topic so I want to thank you both for coming
on again we've been speaking with Martin Lewis from the Department of History and Astia
Perolsvig and pronouncing it correctly and you can get their full bios on our website and
we'll hopefully have you back to have a follow up on this conversation.
Thanks for coming on.
Thank you very much it's been a pleasure.
Thank you for having us.
Bye bye.
(gentle music)
(gentle music)
(gentle music)
(gentle music)
(gentle music)
(gentle music)
(gentle music)
(gentle music)
(gentle music)
(gentle music)