From Protolanguage to True Language

Pidgin speakers use the closest contemporary approximation to protolanguage, but it is hard to find records of their language because they are in such a low status in their societies. The picture shows migrant laborers in Abu Dhabi and surely they use some pidgin to communicate with one another, and just as surely both their employers and the migrants despise the pidgin they speak.

Derek Bickerton’s most important contribution to the study of language origins was the idea of a protolanguage that was spoken before full language appeared, and it remains the most important idea in his new book, Adam’s Tongue: How Humans Made Language, How Language Made Humans. Put simply, protolanguage is a simple language, but Bickerton is more specific than that and puts the matter more formally. He also reveals that he had a series of e-mail exchanges with Noam Chomsky on the subject of protolanguage and, as he tells it, he came out on top.

Human language as studied by generative grammarians (and Bickerton belongs to that group) is defined as the set of sentences that can be produced by the hierarchical rules that govern the language’s syntax. There are also a set of rules that govern the construction of words (e.g., cat+s to pluralize English nouns, dis+approve to change a word into its opposite). But these rules assume the existence of a conceptual ability that animals lack. [Is this true? To follow Bickerton’s and Chomsky’s argument we must assume it is.] As Bickerton put it, “Language must have evolved from a prior system, and yet there doesn’t seem to be any such system out of which it could have evolved” [p. 35].

Bickerton’s solution was to suggest that first came a protolanguage, that’s a language without syntactical or word-construction rules. When protolanguage began, it used words to express thoughts that apes were capable of having. The use of these words themselves made the speakers aware of their limitations leading to an understanding of words that enables them to be used in compound or syntactical form. (Hence the book’s subtitle. Early humans are making language, while language is turning speakers into more modern humans.)

The formal difference between protolanguage and true language is that protolanguages puts words together like beads on a string, without rules (at least beads on protonecklaces are so strung), but they share a unique property as well. In both proto- and true language, words are combinable. As Bickerton puts in, “Languages combine lawfully and protolanguages combine lawlessly” [41],

Bickerton discusses Chomsky’s rival theory at some length (see: Chomsky’s Theory of Language Origins). Chomsky not only believed that protolanguage did not happen; he doesn’t believe it could have happened. In this area, however, Chomsky is at an unusual (for him) disadvantage. Bickerton is an expert on pidgin languages and the earliest forms of pidgin combine as lawlessly as a protolanguage would. Thus, as Chomsky raises theoretical and logical arguments against protolanguage, Bickerton can point to real-world counterexamples. The result of this dispute is that Bickerton is able to produce a much more dynamic theory than Chomsky imagined.

Both Chomsky and Bickerton begin with the proposition that our pre-linguistic ancestors, like other animals, had concepts that “won’t Merge.” Merge is a theoretical process by which words are strung together according to rules. As I say, I’m not going to trouble myself here with the truth of the proposition. Generative grammarians as a group accept it.

In Bickerton’s theory the human lineage began talking, using protolanguage. Protolanguage, lacking rules, does not require Merge. The talking began, says Bickerton, as a means of recruiting others to help scavenge carcasses. It sounds possible, although my time on the Serengeti persuaded me that carcasses are a very short-lived phenomenon. Bickerton seems to think he is absolutely right, but he has no better reason for thinking so than that he has a good story.

Over time (and this time could be two million years), talking itself turns lawless speech into lawful speech. The experience of speaking turns concepts that won’t Merge into concepts that will.

In response to this change in concepts, the brain is rewired to better handle linguistic rules. Bickerton does not insist on this point, seeing it as only plausible.

With the new powers of linguistic organization complex thoughts and planning become possible. Modern humanity has arrived.

Chomsky’s theory has many of these same steps but they are not dynamically arranged. A mutation simply transforms animal concepts into mergable concepts. Another mutation rewires the brain so that complex thoughts and planning become possible. Then still another mutation allows people to start expressing their thoughts  aloud (i.e., they start talking).

You can see from this that both Bickerton and Chomsky are trying to explain the same thing—how to get from animal thinking to linguistic thinking—but only Bickerton approaches it as an evolutionary process. So that gives him a leg up, but at the same time, he does not surpass the limits of his field.

It should be obvious that the origin of language is a multi-disciplinary subject. If you want to put forth an account of how language began, no matter what your background you have to broaden your scope and think about questions that don’t come naturally to a person of your inclinations, interests, and training. Everybody, of course, has to look into the story of evolution, but that should be just the beginning. There is by now a very long list of things besides language that distinguish humans from apes. Here is a partial one:

Anatomy: upright biped, hairless, eyes white instead of black, different hyoid bone, lacks chest air sac, brain enlarged and moved above brow…

Genetic: FOXP2 mutated …

Life cycle: breast-feeding period reduced, childhood period between infancy and juvenile state, adolescence between juvenile and adult state, elongated post-menopausal state in women…

Social structure: identity goes beyond self to membership in larger communit(y/ies); relationships between community members include duties, trust, support; mother lets (and expects) others to care for infant; adults and infants play together; joint attention common; emotional bonding activities common…

Some of these differences may be relevant to the origins of language, but Bicketon ignores them as a group. Instead he argues that people started talking using simple (lawless) combinations and evolved (culturally and/or physically) the ability to make syntactical (lawful) combinations. He had to work this point through in more detail than it merits on its face, because of the skepticism of professional colleagues who deny that language could ever have been simpler. (Even there he stays with his own work. David Gil’s paper presented in Barcelona last year reporting on modern, syntactically simple, languages did not make it into Bickerton’s bibliography. (see: Complex Grammar Has a Simple Solution.)) Many books and papers on language origins leave me going wow. In this case, I was back to an old ninth-grade reaction… well, duhh.

Note: It’s a paradox, I suppose, but in relation to this book I keep thinking, and another thing… so next week I will discuss one other aspect of Bickerton’s book: his list of  requirements that any theory of language origins must satisfy.

Comments

[Uzza]

Should language be on any list of things that distinguish humans from apes? There are plenty of adult humans who lack language, for instance the 400+ J. Kegl has documented. If Bickerton is right and “language makes humans” does he classify these people as some non-human species?

They possess modern human brains and the difference between them and humans with language cannot be explained by species-level differences in brain structure, only by sociocultural environmental factors. The set of hominids who lack language includes these particular humans as well as apes.
--------------------------
BLOGGER: A living person today without language is still the product of a long history of language-species interaction. It's that interaction that Bickerton is talking about.

[John Feliks]

The Idea of Protolanguage Considered in the Context of Archaeological Evidence

Quoting from Bolles' comment on Bickerton:

"It should be obvious that the origin of language is a multi-disciplinary subject. If you want to put forth an account of how language began, no matter what your background you have to broaden your scope and think about questions that don’t come naturally to a person of your inclinations, interests, and training."

There can be no doubt that Bolles is correct. Any specialist field on its own, be it physical anthropology, psychology, genetics, neuroscience, or even linguistics is obviously far too narrowly-constricted to claim any comprehensive understanding of modern language, let alone, ancient language or protolanguage. In my opinion, Bolles' observation or something similar should be the motto for beginning a new era of linguistics if we are interested in understanding language origins on a meaningful level.

I would like to point out that most discussions of language "origins," including Bickerton’s idea of protolanguage, tend to be philosophical or analogical because they make very little if any reference to actual physical evidence of symbolism in the archaeological record. Analogies based on things observed in modern languages are meaningful, of course, because that is the only way we can relate to early languages, but to discuss language origins in a context without reference to physical evidence of symbolism from the time periods being considered perpetuates a notion in the public's mind that there actually is no physical evidence of language from early time periods. But this is simply not the case as I know very well after 15 years of research in this area.

The first unequivocal proof of early language was presented at the XVth UISPP Congress in Lisbon, 2006, in a paper called The Graphics of Bilzingsleben. Unfortunately for Bickerton and others writing on language origins, the paper and all of its geometric data has been held back from publication for nearly three years. This is an advantage only to those scholars who have held the paper back because they are aware of the data while other scholars continue to write from a perspective which is less informed.

Bolles quotes Bickerton, "Language must have evolved from a prior system, and yet there doesn't seem to be any such system out of which it could have evolved."

The idea that language evolved from a prior system is the standard notion and often presented as an axiom. However, when one is thinking in scientific terms, it may be better to focus on the second half of Bickerton's statement. Instead, everyone is so convinced that the first half must be true that they continue trying to fit a square peg into a round hole even though that square peg never did fit and still doesn't, but to add a touch of sarcasm here, we know that it must fit so we must keep looking for ways to make it fit.

More in alignment with the second half of Bickerton's statement, the Lisbon paper deals with Homo erectus symbolic artifacts securely dated to between 320,000 and 412,000 years old and offers fully-testable and falsifiable geometric proofs that by this time, the system of human language was already fully-developed with no sign even then that it was in any kind of a transitional half-way-there or protolanguage stage. (Granted, Bickerton may have been thinking in terms of other species but I am trying to keep this comment in the realm of physical evidence of symbolism.) The paper's primary evidence is that of the earliest motif duplicated on two separate artifacts. The two motifs are the same size, they feature the same outer angles and inner angles and they are situated in entirely different contexts within their respective artifacts (which rules out the possibility that they were engraved as mere reactions to a medium or represent mere scribbles); and finally, the two motifs were apparently engraved with a straight edge.

Language or protolanguage is all about duplicated or agreed-upon motifs (e.g., words), so if duplicated motifs are discovered in the archaeological record from a certain time period or for a particular human species, that automatically brings at least one human language time zone out of the realm of mere speculation or analogy and into the realm of science. To further demonstrate that these two motifs represent the same linguistic competence known in today's world, apart from the proven design association between the two motifs, there is the added dimension of another association indicated by the use of a straight edge. As explained in the paper, any line engraved with the aid of a straight edge is directly symbolic of the straight edge itself, being a representation of the edge. Therefore, each engraved artifact, while apparently referring to the other, additionally refers to the exact same "external" object or concept, the straight edge. This double-layered conceptualization further demonstrates that by the time of Homo erectus (regarded as the standard half-way point in modern anthropology) thought and language were as capable of complex exploration as they are in modern times and adds force again to the second half not the first half of Bickerton's statement, i.e. that there seems to be no precedent from which modern human language evolved.

While Bickerton was likely not aware of this data when he wrote Adam's Tongue, the physical evidence, at least at this point in time, does not indicate any gradual development of language capability. So, I would not regard Bickerton's evolutionary perspective as any kind of a "leg up" on Chomsky as Bolles suggests that it is; and please remember, I am specifically talking about instances of direct physical evidence directly associated with fossil hominins, and am not discounting the possibility that Bickerton's or some other philosophical ideas might be logical or even likely explanations for language origins.

In conclusion, my belief is that discussions specifically related to the origins of language, whether discussing full-fledged language or protolanguage need to reset the detent 0-point of fully-developed language to Bilzingsleben rather than the modern era however vaguely defined. If the data were allowed to be seen by the public, it would be clear that there is no longer any need to consider the possibility that fully modern human language began any later than 400,000 years ago.

[JoseAngel]

The very notion that any human symbolic activity was "fully modern" 400,000 years ago seems to negate any relevant role for evolution or cultural interaction in the development of language.

I find Bickerton's book thoroughly enjoyable and thought-provoking, indeed a must for anyone interested in the interface of human evolution and the origin of language.

[John Feliks]

Jose has perhaps misunderstood my point. It is not regarding any accumulations or variations whatsoever within languages or cultures; this is the part that is easy for any culture once a certain capability has been attained, and the larger the population the easier and faster the accumulations or variations. Rather, it is that the capacity for fully modern language as we know it today was already present 400,000 years ago. As I had mentioned, those who have had privileged access to the data have already begun tweaking their stance regarding the intellectual capabilities of Homo erectus, both in print and online, while many who have not been permitted to see the data are still firmly locked into the idea of Homo erectus as a "half-way-there" hominid. My disappointment in academia involves the withholding of empirical geometric data from the public. This withholding of data encourages those who see Homo erectus as less capable than us to continue defending a paradigm which is assuredly no longer valid.