Header 1

Our future, our universe, and other weighty topics


Wednesday, January 3, 2018

Origin of Language Theorists Are Like a Circular Firing Squad

One of the great unsolved problems is the problem sometimes called the problem of the origin of language. But to use that term is actually to oversimplify the problem. It is better to refer to the problem as the problem of the origin of language and linguistic biology. What needs to be explained are two different things:
  1. How what is it that any language was ever able to originate?
  2. How was it that humans were ever able to acquire particular biological features used in speaking and understanding language?
There is a kind of “which came first, the chicken or the egg” problem involved here. Let us imagine a certain time thousands of years ago when human ancestors had a primitive larynx and pharynx. At such a time it would not have been possible to speak words with any clarity. Trying to speak words would have been like trying to say “the rain in Spain stays mainly on the plain” while keeping your tongue on the bottom of your mouth (or like trying to speak with your mouth filled with rocks). It seems that language could not have got started under such conditions, because spoken language would have been too ineffective as a communication tool. But if language never could get started under such conditions, how could there ever have been any reward-propelled evolution of speaking ability that resulted in humans able to speak like we can speak?

Then there's the problem of how some particular language ever could have been invented and spread about to achieve any following greater than that of just being the custom of a single little tribe. If there was some bright and domineering tribe member determined to teach language, we can imagine him teaching a few words to his fellow cave men (such as words for cave, bear, hunt, day, night, cook, and eat). But how could the first language teacher ever have thought up a grammar? And how could such a teacher ever have taught a grammar or abstract non-noun words to those who had no language?

Imagine you arrive by boat on some remote island in the Pacific consisting of people who don't speak English. You'd be able to teach them quite a few noun words, and a few verb words. But how could you possibly teach rules of grammar to such people? And how could you possibly teach abstract words such as “soon,” “possibly,” “tomorrow,” “went,” “hope,” “try,” “almost,” “towards,” “beyond,” “until,” and a thousand other words that you couldn't teach by using gestures?

And even if some teacher might get a language to be adapted by some little tribe, how could such a language possibly spread beyond that little tribe? It would seem that the only way a language could get a decent following would be if it were mandated or encouraged by some little form of government or some organization greater than a tribe. But such a form of government or organization could never get started, it would seem, until there was already some type of language existing. So we have another “which came first, the chicken or the egg” problem.

These very great problems are ignored by two recent books on the origin of language. The first is a book called The Truth About Language: What It Is and Where It Came From by Michael C. Corballis. In Chapter 2 of this book, Corballis attacks the thinking of language theorist Noam Chomsky. Chomsky has suggested that language arose rather suddenly, perhaps about 60,000 years ago when somehow humans acquired an ability to create a “universal grammar” that he thinks is the key to the origin of language.

Chomsky's theory deserves to be attacked, as it offers no real explanation for the problems discussed above. But Corballis doesn't offer a better theory. He merely advances the idea that language arose from bodily gesture. His theory is that before there was a spoken language there was a gesture language. This is an idea that has never made sense because of the impossibility of giving a plausible explanation as to how humans might have made a transition from a hand gesture language to a spoken language. A gesture language would not be a stepping stone to a spoken language, so the idea of a gesture language that preceded spoken language has no explanatory value.

The “first a gesture language, then a spoken language” theory of Corballis is attacked on page 241 of the book How Language Began by Daniel L. Everett. Everett states, “Gestures could not have been the initial form of language.” The language theorist Everett attacks both the theory of the language theorist Corballis and the language theorist Chomsky. So we see that the modern circle of origin-of-language theorists is kind of like a circular firing squad, with no consensus and lots of volleys being fired back and forth. This may suggest that we should have no confidence that anyone has any credible explanation for the origin of language. 


Everett's book offers no real explanation for the origin of language. His book should have been entitled When Did Language Begin? rather than How Language Began, for rather than offering any substantive attempt to explain the origin of language, his book is mainly devoted to defending the claim that language first appeared in the species Homo erectus. We can make a good guess as to why Everett has advanced this strange idea: probably in order to make the origin of language seem like nothing very impressive and miraculous. If we can believe that dumb Homo erectus used language, then we might think that the origin of language was not very astonishing.

But the idea that Homo erectus invented language is an extremely unbelievable claim. On page 129 Everett says that Homo erectus had a brain size “roughly two-thirds the size of a modern human's.” Explaining the appearance of language in a species with the brain size of a human's is hard enough, and you are compounding the difficulties if you try to suggest language originated in some species with a brain much smaller than ours.

Everett admits on page 5 that “few linguists claim that Homo erectus had language.” Everett tries to support his claim by making the strange claim on page 16 that grammar “really is at best only a small part of any language,” an erroneous claim he repeats on page 105 with the weird claim that “there are several reasons to reject the idea that grammar is central to language.” He apparently wants us to think that Homo erectus organisms too dumb for grammar might have still had a language.

On page 88 Everett makes the claim that “there is nothing in the body dedicated to language,” but on page 174 he completely contradicts this claim by giving us the real truth:

The creation of speech requires precise control of more than one hundred muscles of the larynx, the respiratory muscles, the diaphragm, and the muscles between our ribs – our “intercostal muscles”-- and muscles of our mouth and face – our orofacial muscles. The muscle movement required of all these parts during speech is mind-bogglingly complex.

Exactly, and the specialized biology required for speech also includes substantial brain biology (as we know from some stroke victims who may lose the ability to speak but not to understand speech, and other stroke victims who lose part of their ability to understand speech).  Because Homo erectus didn't have a great deal of this specialized biology, that is a good reason for thinking that Homo erectus should not have been able to speak.

Everett tries to get us to believe that Homo erectus had language by arguing that Homo erectus used symbols, and “symbols are just a short hop away from language” (as he states on page 106). The idea that we can assume some creatures had language because they had symbols is not sound, and Everett is only able to produce very weak evidence that Homo erectus used symbols. He refers us to a sea-shell with some scratches on it, and includes a picture of a 300,000-year-old bone that looks rather like a penis. Hardly very convincing evidence of symbolism. The wikipedia.org article on the bone in question (called the Erfourd manuport) says no evidence of carving or shaping has been detected in it. Everett also wants us to think that spears are evidence of symbol use. On page 94 he says, “For their original owners these would have elicited thoughts of, thus symbolized, hunting, of bravery, of caring for their families, and of death.” This is strained reasoning indeed – spears are not evidence of symbolic thinking. There is, in fact, no substantive evidence that symbolic thinking appeared in Homo erectus.

One huge difficulty in any theory of Homo erectus using language is that such an organism would have lacked the complex vocal biology to produce speaking. Everett has an answer to this objection – an extremely lame answer. He states the following on page 188:

In fact, computers show that a language can work just fine with only two symbols, 0 and 1. All computers communicate by means of these two symbols....All the novels, treatises, PhD dissertations, love letters and so on in the history of the world can ...be translated into sequences of 0 and 1. So if erectus could have made just a few sounds, more or less consistently, they could be in the language game, right there with [Homo] sapiens.

This is completely fallacious reasoning. Let us consider the dependencies that are involved when something like a novel or a marketing plan is stored on a computer system. The first dependency is the existence of a particular language – if you imagine a dictionary, you can visualize that dependency. Then there is the dependency of a particular alphabet. Then there is the dependency of the ASCII system, a complex table used to map alphabetic letters to decimal numbers. Then there is the dependency of there existing an algorithm to translate decimal numbers into binary numbers. This ends up being a huge set of symbolic dependencies vastly more complicated than the simplicity of 0 and 1. So you cannot at all argue from the simplicity of binary numbers in binary computers that some organism would “be in the language game” if it merely “could have made just a few sounds.” No organism would ever be able to understand a complex meaning when it heard another organism use some “two sound” language with sentences such as “oh-ah-ah-oh-oh-ah-ah-ah-oh-ah-oh.” There has never been a natural spoken language that used only a few sounds.

And, of course, trying to argue that sub-human organisms would have been capable of something because computers are capable of such a thing makes no sense, and is like arguing that a human or sub-human could sort 20,000 words in 30 seconds because a computer can do that.

On page 106 Everett summarizes his evidence that Homo erectus used language:

The evidence thus strongly supports the claim that Homo erectus possessed language: evidence of culture – values, knowledge structures and social organization; tool use and improvement (however slowly, compared to Homo sapiens); exploration of the land and sea, going beyond what could be seen to what could be imagined; and symbols – in the forms of decorations and tools.

Some of these claims are dubious or debatable, and even if all of these claims were true, none of them would be evidence of language use. Everett completely fails to substantiate his very implausible claim that Homo erectus invented language.

As for the claim of Corballis that a gesture language preceded spoken language, such an idea does nothing to explain the origin of spoken language. To the contrary, there is a very strong reason why the possibility of a gesture language makes the origin of spoken language very much harder to explain. I will discuss this reason in my next post. 

Postscript: In this article Chomsky states the following:

There is little evidence of anything like human language, or symbolic behavior altogether, before the emergence of modern humans. That leads us to expect that the faculty of language emerged along with modern humans or not long after, a very brief moment in evolutionary time.


This is very much at odds with the claims of Everett.