Babel's Dawn

Primates and Genetics

2011-12-20T21:23:27-05:00

One of the big quarrels in the study of speech origins is over whether it is relatively new (about 100,000 years) or decently old (millions of years). This blog has consistently leaned toward the old origins story.

Basically, every time there is evidence of a complex genetic history, it's a point for old origins and every time there is evidence of a straight cultural story or one mutation does all, newbies look better.

A story in today's New York Times is titled "Genes Play Major Role in Primate Social Behavior, Study Finds" adds to the old origins score.

Human societies are quite distinct from chimps and gorillas. We are cooperative and modular, building up relationships across groups.

The study reports: Social structure does not change readily with the environment (presumably because the structure is genetically programmed); complex behavior does not evolve step by step (allowing for a genetically based, complex difference between chimps and us); and brains do not get bigger to handle larger social groups (challenging Robin Dunbar's theory that the human big brain and language origins reflects our larger groupings).

In the old days I would have raised a few questions about this study, but today I'm just going to say read the story. I read it while taking the train to work this morning and want to bring it to your attention this evening, before I fall into my bed.

Marchetti Comments Online

2011-12-05T19:35:25-05:00

Italian Philosopher and hero on this blog, Giorgio Marchetti, has posted a comment and prescis of Babel's Dawn on his website here.

Janet Kwasniak's Review

2011-11-22T19:20:41-05:00

A regular commenter on this blog, Janet Kwasniak, has done me the kindness of reviewing the book on her own fine blog ("A blog on consciouness"):

If you are at all interested in language, human nature, evolution, culture (and I expect many of my readers are interested in that type of subject matter) get the book and have a good read over Christmas... it is many head-and-shoulders above other attempts to trace language’s origins.

Read the whole review here: http://charbonniers.org/2011/11/22

How Language Began

2011-10-30T23:45:00-04:00

Dawn comes up in Kenya's Masai Mara National Reserve.

I've held many consulting positions while working on this blog but now I have one that includes 4 hours of commuting each day, meaning that most of my non-working life is put on hold for the weekends, especially during World Series week. Typically I work on this blog on weekends, but I cannot do that and have the rest of my life. So I think I'll take a break. After 5 years and a book I've covered a lot of ground. In this post I'm putting a bit of what I learned into scenario form.

As I see it there were several stages in speech origins. The first was a long period of vocalizing in which emotions were shared but not named. Then came a word/phrase period, followed by true sentences (subject + predicate). These first languages were likely very concrete and only later developed metaphorical and abstract capabilities.

Vocalizing probably began around 3.3 million years ago when, according to louse DNA information, our bodies had become hairless enough to support two distinct forms of lice. (see: Lousy Timelines) Hairless primates have a problem because they cannot form social bonds via grooming and their young cannot hang onto their mothers' hair. Robin Dunbar and Dean Falk have done good work in this area. Falk especially has analyzed the development of vocalizations that are not words.

As to why we became hairless, I don't know. I don't know why we are bipedal either, but balding mammals are not unknown in African woodlands. Elephants, rhino, and wild pigs also have little hair. Presumably the process of losing hair was accompanied by the counterprocess of making up for hair's disappearing social benefits. Mothers and infants maintained contact vocally while the mother put the infant on the ground. Adults vocalized together while grooming faded.

A vocalization period persists to this day in infants and is generally seen as a movement toward speech, but that goal may reflect the prejudice of speakers. A great deal of emotional development goes on during this vocalizing and the process needn't look to future language for justification. Myself, I suspect that there was a period of more than a million years during which the human lineage had its own distinctive birdlike nature—no flying, but plenty of chirping.

The vocalization-only period began when the lineage was likely some species of Australopithecus and ended after the rise of the Homo genus. This critical period saw the emergence of the African savanna and our lineage's adaption to the new environment. One important adaptation was the formation of cooperative communities, something very rare amongst mammals who are more given to herding and forming hierarchical societies whose members are unwilling to share what they know. On the open plains, however, slow, weak humans have no hope of survival in herds or even the more complex ape societies. They stood together or went extinct.

The oldest Homo known are about 2.7/2.8 million years old and sometime after that members of the lineage began using their vocalizations to name things. The first words were probably formed by random associations. I've seen infants in their first year do the same thing, making what I call "toy words" by themselves. Most such words fade, but a few are picked up by the family. I seem to recall a baseball player in the 1960s and 70s named Boog Powell and, supposedly, the nickname Boog came from a word he had used as a baby.

The critical step in the word origin was not the association of sound with thing—that's pretty much inevitable with a brainy species doing a lot of free form vocalizing. The key lay in others being willing to repeat somebody's association, and that willingness comes only when individuals are already motivated to share knowledge. Once word sharing had begun, Baldwinian evolutionary processes would support the development of an instinct to learn and use each other's words.

The critical evolutionary step was the rise of a cooperative genus and not some new semantic or syntactical intelligence. Apes are already able to associate a number of hand signs with specific things and focus their attention on a named object. What they needed was a reason to make use of that capacity.

How much more did Homo biology have to change once the lineage got to that point where it could name things?

One step was discussed in last week's post, the ability to combine two phrases into a whole. Apes cannot do this, and the reason is not hard to grasp. Two distinct phrases—e.g., that red ball and a blond lumberjack—direct attention in two separate directions. But we can combine the two by finding a common verb that unites them: a blond lumberjack kicked that red ball. So that seems to be an intellectual power that we have and apes do not. Our thinking in this regard is not just a scaled up ape ability; it is something new. Language, as we know it today, reveals a discontinuity between humans and apes.

From time to time I read a paper or get a comment saying that discontinuities in biology are impossible because evolution works in small steps, but today's world is full of such breaks. An obvious example is in the world of birds. They have a complex evolutionary history related to the dinosaurs, but the dinosaurs are long dead so when you look at the world today there is a huge discontinuity between birds and other animals we see. Similarly, the intense group selection that was part of the evolution of the human lineage appears to have left us without any race that can speak phrases but not sentences.

There is, however, an equally important continuum from words to phrases to sentences. The evolutionary process has not led us away from the early association between words and attention toward some more abstract syntactical and semantic processes; it moves in the opposite direction—toward more elaborate powers of attention.

When the transfer to full sentences occurred is uncertain. Last week's post noted a possibly relevant genetic doubling 2.4 million years ago, so we can put that as the earliest possible date (and frankly I don't believe the lineage was even using words back that far). Archaeological evidence puts the possible date as early as 1.8 million years ago when Acheulian axes first appeared. These are the first bifaced tools in the record, meaning both sides of a chopper have been worked to form a cutting edge. To make the axe, an artisan must be able to hold two points of attention in mind at the same time. If you are smart enough to join two sides of a stone by making an edge, you may be smart enough to join two phrases with a verb. Let's say that by 1 million years ago Homo erectus was speaking full, concrete sentences. I don't think it could be much later than that.

There is, of course, a big difference between the ability to speak about concrete things and what modern language can achieve. Is that all the result of cultural evolution? I doubt it. I think in particular some biological change was necessary to support metaphors and perhaps abstractions. Whatever that took, I figure that the biological determinants of modern language were in place by 150 thousand years ago. Although some people still date language as recently as 60 thousand years ago, my 150 K date is reasonably orthodox. There are quarrels over whether anything came earlier or if language sprung de novo out of a 'great leap forward' [i.e., a genetic mutation] of some kind. I'm firmly in the long-history camp.

I defend my position by noting the way we put abstractions and metaphors into concrete form. If we began right away using abstract symbols, I doubt that we would have organized abstractions so insistently in perceptual space, as every language on earth does. Instead, I think language would have a much more mathematical flavor. With something like Dirac's equation we can calculate quantum physics outcomes without having any perceptual understanding of what is going on. That means we cannot discuss the results meaningfully in ordinary language. Physicists talk about superposition, collapse, and entanglement but nobody can explain those terms. The reason for the failure is clear enough. Language organizes things according to a perceptible space; Dirac's equation does not.

A good discipline for anyone interested in writing is to try to write a purely concrete paragraph that hangs together. The tone tends toward that of a hard-boiled crime story for there are no abstractions or interior goals. Emotions are external. Someone can shake but not feel fear. There are no motives, only actions. Without metaphors we can never get into somebody's head or heart.

Metaphors are often defined as a "figure of speech" that says one thing is another thing. When it came to the honor of France, General de Gaulle was a lion in defense of his pride. There is a metaphor (and a pun too) used figuratively. You could just as easily say he was "like a lion in defense of his pride," turning the figure into a simile. But there are some metaphors that cannot be deflated. Her flirtatious behavior shattered his trust. Here the metaphor is shattered and no non-metaphor can be substituted.

Why not? Because trust is a psychological state, not a concrete thing. We can substitute a concrete thing for trust. We might belong to a culture that locates trust in the heart, so we could say shattered his heart but now we've replaced a metaphor with a metaphorical phrase since a shattered heart cannot be taken literally. Instead of calling this sort of phrasing "speaking poetically" we might say the user is speaking psychologically, for there is no way to talk about subjective psychological states without using metaphors (or, if you happen to be a professional psychologist, empty jargon).

How do we understand metaphors of subjective experience? In my own case it seems to me that the word directs my attention to my own knowledge of the experience and if I have no acquaintance with some subjective experience I don't know what the metaphor is about. As I have gotten older, much literature has become more clear to me.

Autistic people are reported to be baffled by metaphors, suggesting that there is some biological contribution to understanding them. Perhaps they cannot direct their attention inward. Naturally there is more to autism than understanding metaphors and I'm not suggesting Homo erectus was autistic.

Wherever metaphors come from, they allow us to speak of invisible things as though they were part of the concrete world. In fact, many people don't even realize they are using language in a way that demands interpretation rather than literal acceptance. I have spoken to many people who get indignant when I point out that nobody takes the Biblical reference to God hardening pharaoh's heart literally. Pharaoh's heart did not literally become like a rock or something. Yet language treats it as though and to understand that statement you have to know the feeling of a "hardening heart." It is not the feeling of a heart attack, but something else.

A second feature of modern language not present in concrete speech is the use of abstractions. We live in a world of abstract references. To take an example almost at random, look at Paul Krugman's column in the New York Times last Friday. Here is a purely abstract sentence (for emphasis, I'm putting all the abstractions in blue).

So a crisis brought on by deregulation becomes a reason to move even further to the right…

You have to be pretty comfortable with abstract words to follow that clause, and yet it feels almost normal because of all the structural components that give it a concrete context: brought on by, becomes, move even further. I can imagine some verbose Home erectus saying something similar: So a rainstorm brought on by dark clouds made us move even further into the cave. Thus, we have the paradox of organizing very modern ideas using verbal structures that may have been recognized a million years ago.

Whether this rise in abstractions required a biological change is uncertain. I have the impression that the change has happened so recently that it is essentially cultural and depends heavily on intensive education. Until very recently many abstractions were dealt with mythologically, presenting ideas in the form of living spirits. But there may be a biological component to the process as well. If there is, it was likely added a while back since everybody in the world today can speak of abstractions like justice, freedom, and rights.

The relation between perception and language is not news. The oldest advice given writers is Show; don't tell. In other words, make it as concrete as possible. Now that I have spent five years focusing on the origins of speech, it makes perfect sense to me, and I will end this period of maintaining my blog with a passage from my book that sums up what I have come to understand about language. Because of language we are presented with

a true biological novelty: a cooperative species able to share perception of a topic by directing attention with words. It is almost as though our lineage developed a new sense, a cooperative sense that for the first time let individuals know what another sees, hears, or even feels.

Duplicate Points of Attention

2011-10-23T23:45:00-04:00

Some news that I've been waiting for seems to have come along. If so, a number of puzzles may have solutions and some suspicions I've held seem likely to be true. The news appears to be a confirmation of an idea put forth a couple of years ago by Terrence Deacon and which, when I heard it, set me thinking.

Deacon is best known on this blog for his book The Symbolic Species but he is also an expert on evolution and the brain. An interesting idea he has been promoting for a few years is that of "relaxed selection," a means of producing a random change and sheltering that change while it finds a role. Most changes to the genome are point mutations in genes and the results are immediately tested. If a mutation is harmful, it does not spread, while helpful genes do increase in a population. Neutral mutations are neither selected nor filtered out. They may increase over time through chance but probably will never amount to much of importance.

Deacon focuses on a more rare type of change in which a section of a chromosome duplicates. Although relatively rare, doubling has happened many times in evolution's long history. An example Deacon likes is the introduction of color vision into the primate line. It began with an accidental doubling of the genes that govern the retina. Color vision did not result immediately. The original genes provided normal vision. The doubling was basically useless but at least not ruinous. In essence, the doubling worked like a neutral mutation, neither selected nor discarded.

Once the doubling occurs, the doubled genes are of course subject to mutation and can change and sometimes produce something valuable. This process led to the eventual appearance of color vision as a neutral duplication evolved into something well adapted to the forest environment. Deacon calls this process "relaxed selection" because the doubled strand of genes is not immediately rejected and allows time for something new to appear.

Deacon has been arguing that the Homo brain's development is the result of a doubling and relaxed selection. Now comes news that a doubled section of the human genome controlling brain development has been identified. The doubling took place an estimated 2.4 million years ago, early in the history of the Homo line.

As far as I know, James Hurford never linked any of his ideas to Deacon's doubling but I did when I covered one of his presentations on subject and predicate. He talked about it as subject and news (see: The Word-Sentence Continuum). Hurford's approach enabled me to think of sentence and predicate as a kind of doubling.

Standard analysis of sentences defines the subject and predicate as a [noun phrase] + [verb phrase]. That rule allows for simple sentences like [Jesus] [wept] and complex ones like [Alexander and all his army] [defeated the Persians and their allies]. But Hurford got me thinking in a more functional way. News can come in many forms: verbs (e.g., died, as in Our neighbor died), adjectives (red, The leaves turned red), and nouns (book, Jack wrote a book).

Looked at this way, sentences can contain two elements and a link. The elements are the subject and the news. The startling thing is that a sentence can ask the listener to attend to both subject and news at the same time. This double attention is made possible by a link that fits with both ends of the elements: The leaves turned … turned red; Jack wrote … wrote a book. If the link does not work with both elements, the sentence is confusing and the doubled attention breaks: The leaves fell … fell red; Jack splintered … splintered a book.

I'm not proposing that sentences always come from doubled attention. Sometimes the link alone is sufficient for the news—Our neighbor died. Perhaps this kind of sentence is older than a doubled one, but ever since I combined Deacon and Hurford in my thinking I've been keeping an eye out for news that might support the idea that speech reflects a doubling that enables people to hold two points of attention at the same time.

Naturally I sat up straight when I came across a news item headlined, "Doubled Gene Means Extra Smarts," by Tina Hesman Saey. The gene in question is called SRGAP2 and has been previously identified as one of 23 genes that has been duplicated in humans but not in other primates. The gene was partially duplicated 3.4 million years ago and then 2.4 million years ago the partial duplicate was itself duplicated. The first duplicate is not found in all humans, but the second duplicate is fixed (i.e., universal) in the human population. 2.4 million years is considered a short time for a duplicate gene to become fixed, and suggests the change is important. Exactly what the changed gene accomplishes is unclear, but the research team—headed by Megan Dennis—proposes that it thickens the brain's cortex.

From my perspective, the date is good (before speech probably began, but not hopelessly before). It's something to keep an eye on.

Beyond Dictionaries and Rules

2011-10-16T23:45:00-04:00

Last Tuesday (Oct. 11, the feast of St. Jerome) was International Translation Day. By chance on that day somebody gave me a link to an item in the Paris newspaper Le Figraro and, when clicked, the link jumped me to a Google translation of the article. As usual, every sentence of the translation seemed ok, if a little off, but the whole effect was disastrously bad. I've been trying to grasp what it is that the translation misses and, in keeping with the theme of this blog, where such a thing might have come from.

It is tempting to say the translation missed the article's spirit, but I'm trying to get a more concrete understanding of what went wrong. What is the spirit that turns a collection of sentences into a unit?

The article translated was a review of the new Tintin movie. (Opening in France this month, and in December in the USA.) Part of the problem was the translation machine's insensitivity to idioms. The article's title is given as "We saw Tintin and it does not disappoint," which is correct (although perhaps have seen is more accurate than saw) but carries with it the taste of French idioms. If I were translating, I would make it, "We've seen Tintin and were not disappointed." Another version might be, "We saw Tintin and it did not disappoint us." The trick is that the French has a change in tense and a change in subject. A French writer can get away with these switches because the sentence uses a cliché (does not disappoint) of Frenchcritical opinion. A translation that includes this cliché-based freedom might go, "We've seen Tintin and it holds up." There we've got the cliché, the tense shift, and the change in subject, but it sounds like English.

Every language has its clichés and perhaps like idioms they need to be translated whole rather than by parsing it, but a phrase's context becomes even more important when translating. The cliché in I had that meeting with my grouchy aunt and it did not disappoint cannot be translated as it held up.

There is an amusing and incomprehensible example of what you get when you translate an idiom's parts, piece by piece. What do you make of this passage: Verdict hot most anticipated film of the year? Idioms are, by definition, phrases that cannot be understood literally, but Google has translated an idiom (Verdict a chaud) literally when it means something like immediate reaction or first impression. It's a rare phrase (Google offers only 14 instances) and is typically used without a verb, making translation even harder. To make matters still more obscure the translator also skipped a preposition, du. The French translates as Here's a first impression of the year's most awaited film. (Since Tintin is a hero of French comics, the coming movie about him has caused more excitement in France than elsewhere.) The passage is a bit of pawing the earth. The reviewer should just get on with it, but bad writing still deserves s a clear translation.

Translating that simple little bit requires the translator has to look beyond the French-English dictionary if he is to make a decent go of rendering a French sentence into English. But what might that beyond be? If a language can be defined by the rules that produce its sentences, then why should we ever need to check beyond the sentences themselves to get the best translation?

Regulars on this blog may feel a couple of familiar themes lurking just off stage. One is the speech triangle, specifically the corner with the topic. The translated essay is a review of a movie, moreover a movie about a beloved comic character familiar to every person who grew up in France, moreover a movie produced by a much admired filmmaker, Stephen Spielberg. The more knowledge the translator can bring about Tintin and Tintin's place in French culture to the project, the better the translation is apt to be. This issue is specific to this bit of writing, but every translation task includes a topic and the translator. So every sentence generator needs syntactic rules, a dictionary, and a reference work on the topic. Rules plus dictionary won't suffice. And topics have, by definition, been part of language from the beginning. Even if we imagine a group of Homo erectus exchanging no more than phrases we would have to know what they know about their topic to understand what they said. If they are looking for a good stone to work, with they might well be assuming a trove of knowledge about stone tools that most of us do not have. Thus, even with a protolanguage-English dictionary we might have a hard time getting exactly what they are saying.

A second recurring theme on this blog is how language directs attention. That is to say, language constantly turns the reader/listener's attention outside the sentences. A translation like "verdict hot" may be literally correct, but we know immediately that something is wrong because it provides no point, abstract or real, to direct our attention. Computers, of course, have no attention to direct and therefore need some work-around solution to discover when a translation is simply hopeless.

Attention directing too has been part of language from the beginning. Hopping into our time machine and watching that group of erectus discussing tools, we would do well to direct our attention to the stones. Closing your eyes and focusing just on the words spoken will not do.

Another element in the text but not in either the syntax or the dictionary is the author's excitement about a Tintin movie by Spielberg. If the author was speaking, we could hear that excitement in his voice. That is what we would have gone by when eavesdropping on our erectus. Do you think they were excited or matter of fact when they examined stones? In written language tone of voice has to be replaced by something else.

In the Tintin review the tone is established in the title and persists through the whole review. The author wants the movie to be fine but feared it would be bad; he is relieved to discover that the movie is pretty good. The Google translator misses the whole of this tension, naturally. This deafness to tone leads to some funny translations. One section is bafflingly headed, "Snowy to life also," when a glance at the French makes it clear that a better translation is, "Snowy too comes to life." Snowy is Tintin's dog, called Milou in French. It is typical that the machine was able to translate the dog's name correctly, but unable to give us a comprehensible sentence expressing the reviewer's enthusiasm for the movie's accomplishment.

Particularly striking is the machine's problem with "come to life." In the review's body, the machine gives us, "As for Snowy, he takes life too …" Takes life? Doesn't that mean kill? So there is another difficulty, metaphor. Metaphors are never literal. The French might say a fictitious character "takes life" but in English fictitious characters "come to life." Sadly, a dictionary cannot just say translate take as come when followed by life. People can take to life as well: e.g., Luke took to life in the two-street town surprisingly well. To translate the metaphor accurately the Google translator has to know what it is talking about, but of course it never does.

Meaning is the great mystery of language. This blog has a running hypothesis that it works by directing attention toward some detail of a topic. The problem is, we don't understand how attention works, so we cannot make a machine that can duplicate the function. The Google translator, like all translation machines, has been built to work around the fact that it doesn't know what the text means. So it uses a dictionary, a set of syntactical rules, and a statistical analysis of a large corpus of translated documents.

When language began the human lineage could already pay attention, focus on simple topics, and express an attitude in tone of voice. So from the beginning Google translator likely would have been missing the subtleties of the original speech.

Getting Less and Less Perceptible

2011-10-09T23:45:00-04:00

What shall we talk about?

One of the greatest satisfactions of writing this blog for five years (and I hope also of reading it) has been the much sharper understanding I gained about how language works. As a professional writer my whole adult life (and an apprentice writer for most of my childhood) I liked to think I had a decent understanding of the way language worked, but this blog has brought old instincts and habits into consciousness. I've made a few corrections, as a result.

Central to understanding language is the speech triangle: a speaker and listener paying joint attention to a topic. Each part of this triangle has its own evolutionary history. In my lecture the other day I went briefly into the evolution of topics. Animals pay attention to things, but a topic implies shared contemplation which animals don't engage in. Jean Louis Dessalles has noted that no animal has ever been observed drawing another animal's attention to something because it is interesting.

I'm not sure that is entirely true. Adult animals sometimes teach the young by demonstrating. Mama lions bring small, captured prey to their growing cubs to learn killing. Since there is no talking, the prey isn't a topic, but perhaps it is a prototopic, a shared point of interest. I mention this because it indicates that once things had names no further evolution was required to turn those things into topics.

We also know from experiments that apes can make words in sign language. They don't engage in the kind of chatter fantasized in the movie Rise of the Planet of the Apes, but they are able to provide the correct sign when asked to identify many things, such as a bird or a cow.

Examples like these suggest that the ability to recognize potential topics is much older than speech itself. What was needed was not a topic-spotting skill but a motive to consider a topic. Really, dual motives are needed. The speaker needs a reason to draw attention to the topic, and the listener needs some reason to care about and trust in what the speaker says. This kind of dual motivation is cooperative. Thus, it didn't take the evolution of a new curiosity about the world to add topics to the speech triangle.

Work with apes and sign language have made it pretty clear that apes cannot put together a subject and a predicate to form a true sentence even though sign language is perfectly capable of expressing a complete idea. James Hurford has described this advance as the introduction of news, although one can argue that the speaking of a topic alone presents news.

The critical advance is the transformation of two points of attention into a unified whole. I could say, "Mother. Apple pie," and produce a kind of emotional or associative collage. A sentence goes beyond collage by fusing the two parts. A sentence like, "Mother burnt the apple pie," combines the two foci, thanks to the verb which functions with both noun phrases (mother burnt … burnt the apple pie). If the verb doesn't work with both foci (e.g., mother swam … swam the apple pie), the sentence is meaningless.

Since apes cannot construct such sentences, our ability to do so must result from some evolution. We have to be able to link two things at once to produce a new whole. There is archaeological evidence of this ability in the Acheulean axes. About 1.8 million years ago, Oldowan tools—stone flakes that are sharp on one side—began to be replaced by axe-like stones that had been sharpened on both sides to produce a chopping edge. It's an open question, whether this ability to make an axe head is related to the ability to produce sentences. It would be nice if a strong link could be established because then we could give a date as to when sentences were possible.

Modern sentences can be much more complicated than subject + predicate. William Faulkner was a master of elaborate sentences, one of which reads: "The store in which the Justice of the Peace's court was sitting smelled of cheese." The subject and predicate are united in the classic manner ( the store smelled … smelled of cheese) but there is a long aside that inserts a second, more challengingly organized, sub-sentence (in which was sitting … was sitting the Justice of the Peace's court). Did we have to evolve something new to add a subordinate clause to a sentence?

I don't know, but I doubt it. Languages have so many twists and turns, and they vary so much. I'm inclined to say that normal changes in language overtime is sufficient to account for the syntax that follows the union of subject and predicate. But others have other opinions and I'm open to experimental evidence. In particular, I wouldn't be surprised to learn that the scaling up of the brain was required to handle more complex sentences.

When you focus on topics, however, the issues are different from syntactical matters. Instead of worrying about recursion and move rules, there is the steady straying from physical reality by topics. Apes and toddlers name concrete things, but there is more to talk about than things that can be sensed. In particular, things that happen internally cannot be named directly.

A common internal experience is feeling hatred. How do you talk about that in concrete words? The Bible, when talking about Pharaoh's returning hatred for the Jews of Egypt says his heart hardened. Even people who call themselves literalists don't take that one literally. It's a metaphor for the changing feeling going on inside the pharaoh.

The ability to handle metaphorical topics probably evolved. I'd like to see an ape understand one. Also, not every human can understand metaphors. It is particularly common for autistic spectrum people to have trouble with them for they find them illogical. How could pharaoh's heart harden without killing him? This kind of problem suggests that there is something in a normal human brain that understands metaphors and that doesn't work properly in autistic brains. Other explanations, however, may be possible.

Finally, the most abstract topics are concepts. Whether we needed to evolve a special brain function to handle concepts is not clear.

One thing that is very apparent about metaphorical and conceptual language is that it uses the same structure as concrete language. A sentence like ,"There is no justice in that courtroom," makes justice sound like a physical thing. Even if we speak more abstractly, "The system is unjust," we are talking as though the system exists and unjust is a real property, like, "The apple is red." Languages don't seem to distinguish, at least syntactically, between abstract topics and concrete ones. It suggests to me that while topics have been evolving, syntax has not.

Of course, getting at the bottom of these details calls for much more study, but it seems plain that the topic part of the speech triangle has a more complicated history than appears at first glance.

Going From Blog to Book

2011-10-08T15:37:07-04:00

I spoke last Tuesday at an event sponsored by the Science Writers in New York (SWINY).and now I've managed to put the lecture together in a series of videos. I got some compliments on the slides I used and include them in the video. If you cannot start them by clicking the image above, you can go to the Playlist here and click the Play All button. (Sorry, there is no transcript for the hard of hearing. My notes were the slides.)

What Shapes Language Besides Genes?

2011-10-03T02:04:52-04:00

Genes and how they work may be the hottest dispute in biology right now, and since most other questions in biology cannot be settled until the gene question is resolved, everybody is marking time just now. I'm reminded of this point by a paper just published in the online journal¸ Biolinguistics. The paper ("[Eric] Lenneberg's Views on Language Development and Evolution and Their Relevance for Modern Biolinguistics") is by Cedric Boeckx and Victor M. Longa, both are linguists making a biological argument.

Ostensibly the paper is a look back at Eric Lenneberg, author of the 1967 classic The Biological Foundations of Language, but it strikes me as an attack on classic nativist position of generative grammarians: the belief that a universal grammar born into us and that the development process has nothing to contribute to our grammar. Language acquisition could be instantaneous and it wouldn't make a difference to how language turned out.

Eric Lenneberg was a lasting influence on this blog. I read his book in the early 1970s and ever since then have taken for granted that language has a biological and evolutionary history. This blog has staked its reputation on an evolutionary approach and I owe that confidence to Lenneberg. However, I confess that while I still have my copy of his book close to hand, I haven't read it in decades. It's nice to return to it. The book is, as the authors say, "a body of work of the highest interdisciplinary quality" [p. 254]

A basic question of biology has long been how a new generation acquires its traits. There are two basic answers:

Preformation – the next generation already exists in potential form. In the eighteenth century naturalists spoke of a homunculus, or tiny being that lived in the seed of the present generation. That idea proved wrong, but it exists in a more modern form, information controlling the formation of an individual is passed through the generations. Differences in individuals arise from changes to the information.

Development – a series of developmental interactions between genes, environment, and the rest of the organism's body. Differences in individuals arise from differences in interactions.

For the past half-century and more the preformation explanation has dominated. Genes are viewed as causative agents, the major element selected in evolutionary processes. Thus, if modern humans and Neanderthals share a gene, they share the trait as well. This point is especially important in the FOXP2 gene, associated with language. If the form of the gene supports language in us, it supported it in Neanderthals as well—or so the preformative theory insists.

The authors add, "that Generative Grammar has traditionally embraced the preformationist perspective by means of its defense of the need for a genetic program for language" [258]. The Universal Grammar is assumed to be built into the brain whose modules are determined genetically.

More recently the emphasis has shifted. There is ample talk about Evo-Devo and epigenetics, but, suggest the authors, language scholars have not paused to absorb the full implications of the change. Genes, of course, still matter, but they are only one part of an interactive process. An abnormal gene, for example, may result in an abnormal organism, but it does not follow that the normal version of the gene causes the normal form of the organism. Similarly, breaking a wire in an automobile engine may mean the car won't move, but that doesn't mean the wire is what causes the car to move. This analogy has obvious implications for the FOXP2 gene, which was discovered in an abnormal form in a family with abnormal language difficulties. We cannot argue from that connection that the normal form of the gene causes normal language.

So if genes are not the causative agents shaping the organisms, what are they? They are part of the story, contributing to the probability of the organism attaining a particular state. We've already seen that the genetics of a population can determine the probability that the population speaks a tonal language, so the idea of genes shaping probabilities rather than serving as causes is not new to this blog. (See: Was the First Language Tonal?)

What does this tell us about the evolution of language? The authors don't say, but they seem to imply that the process is statistical and the result is statistical. Developments depend on interactions and it is quite arbitrary to pick one part of the interaction (such as the gene) and to say it is the key.

It is well known, for example, that children who are raised without human contact (there are a few wretched cases) do not develop language. Traditionally we say that this is an environmental effect that overcomes the genetic process. But if you think more broadly, it might be that human interaction doesn't just permit the genes to do their work. They might contribute to the development of language. In that case, there should be many examples of children acquiring language at different paces, depending on their interactions with speakers. It sounds doubtful, although every nursery school teacher knows that their children show a great range in language abilities. Identifying the sources of all this variety is plainly an impossibility and is unlikely to be attempted.

The paper's authors write that "innate traits are due to … any factor influencing development. Genetic factors are just one of them, but in no sense can they be regarded as the main or unique factors" [263]. Yet the difficulty of identifying all the other factors leads me to suspect that for a long time to come the focus is going to remain on genes.

This Child is Under the Weather

2011-09-25T23:45:00-04:00

See you next Sunday with the post I had planned to put up today.

5th Anniversary

2011-09-18T23:45:00-04:00

This post marks five years of blogging. Usually on my anniversary post I summarize what I've learned so far. But since I've just published a book setting forth everything I've learned, I'm celebrating this year's anniversary by taking the day off.

The Minimum Requirements for an Acceptable Theory

2011-09-11T23:45:00-04:00

Giants of poetry stand as an illustration of how intertwined language and being human are. I'm tired of theories that ignore the mix.

I read a technical article the other day that discussed a way of evolving language without needing to cooperate. (The paper is available online, here.) When I first approached it my expectation was that I would present the author's argument and leave it to the readers of the blog to accept the finding or not. I changed my mind, however, when I got around to reading the paper. Working on this blog, and my book too, has broadened what I require of an account of origins.

Work on this blog has shown me that the simpler the account of language origins the more marginalized the role of language itself in one's account of human existence. The simplest of all explanations is that language is a spandrel, a side effect of some other selection process. Stephen Jay Gould was the biggest proponent of this idea. I've never seen a spandrel theory that specifies what produced the linguistic side effect, but the effect of the "explanation" is to trivialize the origins and to make language itself a secondary part of human experience. It is an ironic result as Gould himself was a great writer who used language to sharp effect.

Slightly more complex is a "big bang" theory—a single mutation rewiring some portion of one individual's brain which was then selected because of the improved thinking capacity that resulted from the mutation. This account rests on the old Greek idea that humans are distinguished by their reasoning abilities, and plainly we are smarter than apes, although apes are not idiots. They are capable of rational problem solving. I have read a variety of variations on this level theory—the most prominent version comes from Noam Chomsky—and I am often struck by the narrowness of their view..

Most people throughout most of the world and history have not based much of their activity on deep reasoning or scientific inquiry. Yet they have mastered their environments and become the dominant species in their locality. Scientists and philosophers are unusual in stressing human rationality and overlooking the many other forms of creative and adaptive behavior that characterizes human society. Big-bang type theories ignore all that.

Still more elaborate are gene-level selection accounts that involve multiple stages rather than a big bang. The richest such theories are able to place language at the center of human behavior and are able to view society in terms of altruism. The best example of this approach comes from Jean-Louis Dessalles in his book Why We Talk (see: Why We Talk: Summary). He takes a really detailed look at language, perception, and evolution, but in the end there is no room for love, for moral behavior, for art, for religion, for humor, for singing, or even for two people working together to carry a log. Those are all presented as spandrels that cannot have evolved directly.

Dessalles has defended himself on this blog by saying it is not his job (or even his place) to question orthodox evolutionary theory. It's a fair point, although gene-only selection has never been as orthodox as its defenders say. All the founders of the modern synthesis of evolution and genetics, giants like John Maynard Smith, accepted group selection as important to human evolution. A variation on group-selection, known as multi-level selection—has also gained strength since Dessalles published his book. But, as the paper that sparked this meditation was written by Dessalles himself,

The richest theories manage to at least imply room for the breadth of human essentials. One of them comes from Michael Tomassello, another from Terrence Deacon. The first focuses on cooperation, the second stresses the need for symbols that can organize a cooperative society. Their accounts differ, but in each case they point to a view of humanity that embraces its breadth. Without that as a minimum, I'm no longer interested in the theory.

Well, I can imagine one exception. If the theory of origins can explain that language is not central to the human experience, I'll have to look at it seriously. Until then …

A New Account of Human Origins?

2011-09-08T10:00:00-04:00

The latest issue of Science, just published, devotes several papers to a consideration of Australopithecus sediba, a species identified in southern Africa and dated to between 1.95 and 1.76 million years ago. When the species was identified in 2008 I wasn't much interested because the genus Homo was already perhaps a million years old by then, so this finding seemed to outside the human lineage. There were many bipedal apes three million years ago and even after the emergence of Homo bipedal apes persisted. The new Science issue considers the possibility that A. sediba led into Homo as we now know it.

How is such a thing possible? We've just had an example of what can happen as DNA has shown there was a crossbreeding between H sapiens and Neanderthals. There could have been a hybridization between A sediba and some Homo species, maybe even leading to Homo erectus.

A study made of the interior of a very well preserved juvenile cranium shows that the brain was much smaller than those of Homo species but suggests that the brain was being rewired in direction of the modern frontal lobe. Now there is a surprise. If true, it suggests a double story—brain growth which had begun with Homo continued with a reorganized brain from last Australopithecus.

Another surprise, the A sediba pelvis included modern features that are usually explained as adaptations to allow for the birth of a big-brained infant. But sediba did not have big-brained infants. So why had its pelvis shifted?

The hand fossil is the most complete such fossil from the period we have yet. Details suggest both tree climbing and a precision grip. Such hands could have been used for making tools.

A well preserved ankle suggests a human like arched foot with a heel more typical of apes.

All of these suggestions are quite important, if true.

A Lifetime of Wondering

2011-09-04T23:45:00-04:00

The porch on the Acropol Hotel in Morogoro, Tanzania. A friend recently told me I used to pontificate there on the subject of language origins. I remember the pontificiating, but not the topic. (This photo of New Acropol Hotel is courtesy of TripAdvisor )

This Tuesday, the 6^th of September, marks the official publication of the Babel's Dawn book, presenting the story of the 6 million year long process that led to the appearance of true speech. I learned ninety plus percent of the material in the book by working on this blog, but the passion I bring to the issue is very old. More for my own curiosity than anything else, I thought I would celebrate the publication by figuring out the various key moments in my lifelong search.

Age 11: While walking along a street in Paris I realized that language had to have begun at sometime, but, since it couldn't have happened in a meeting wher people worked it out jointly, there was no obvious way for people to start talking to one another. Back in those days paradoxes sometimes terrified me—e.g., what happens when I go to the end of the universe and stick my hand out?—but the language question held my interest without inspiring dread. [This paradox gets a simple solution in Babel's Dawn: before we had locally generated words, we had a long history of biologically produced vocalizations that could turn into words.]

Age 20 During my college days I learned that the academic world took a dim view of my interest in language origins and there was no way to pursue the matter at the university level. The question of origins would be a sideline for me.

Age 24: A friend who was in the Peace Corps with me recently said that he recalls me sitting in Morogoro, Tanzania's Acropol Hotel and discussing language origins with him. I remember the Acropol and those conversations very well, but had completely forgotten discussing the language question. Apparently, despite my discouragement, the origins issue stayed in my mind.[Later, when I returned to the USA, I sold my first article to a newspaper, an essay about Swahili which ran in the Washigton Post.]

Age 25: While still teaching as a Peace Corps volunteer I had a conversation in Swahili that was so different from English—tense unknown to English speakers; single word served as full sentence—that I wondered if reality itself was just an invention that varied from culture to culture. At once, however, I saw that there was something in common--both English and Swahili expressed subject-object relationships. Since that time I have believed language has universals, although I have bounced between thinking them learned or inate.. [A few years later I learned about Chomsky's work, and of course was immediately attracted to its emphasis on universals; however, I never could become a true Chomskyan because I could never take seriously Chomsky's level of abstraction, removing language from all social and worldly context. I would read him thinking, great, great, great and then end up disappointed because the points were not connected to a larger whole.]

Age 28: While falling into bed it occurred to me that I might be wrong about how language is learned. Perhaps, instead of being taught language by their parents, perhaps each generation invents language anew. That led me to thinking about language as an evolutionary product, but it was not an interest in evolution that led me to that thought. Ever since the Peace Corps I have believed in the fundamental creativity of ordinary people. [This bedtime thought was what led me to learn about Chomsky, psycholinguistics, and modern linguistics. It suggested that there might be an answer to where language came from.]

Age 38: Near the end of writing a book about children's language—So Much to Say!—I noticed something peculiar about speech: we say words and then correct them. Computers do not produce output and then issue a correction. I decided that we must have two separate systems for producing language, and I never again took seriously the idea that sentences are the product of a single computation. [Work on further books suggested that those separate systems are two forms of memory—recall and recognition, which produce novel interpretive powers when combined.]

Age 60: As the 50^th anniversary of my original question approached I thought it was about time that I made an effort to see what I could learn about language origins. I read some books and took some notes. Eventually I decided the best way to proceed was with a blog that would let me do research, provide some contacts, and enhance my credentials. Shortly before launching the blog I had one further thought: language depends as much on listening as it does on speaking. The idea prepared me to embrace the speech triangle when I came across it early in my work on the blog.

So there you have a lifetime of milestones working toward today's publication. Of course, the milestones are really just points that stand out in memory. The real story is a lifetime of reading, writing, and wondering about language.