On the Origin of Language

Introduction and Initial Assumptions

The question of the origin of language is one that has fascinated mankind for thousands of years; some of the earliest writings that have survived to our time have to do with that question.1 The answers that their authors have given us have ranged very widely, from "it was a gift from God" through "it was an inevitable consequence of natural selection" to "it was just a lucky accident." And as Professor Roy Harris has pointed out in the volume cited in note 1, the question of language's origin has often been conflated with related but logically independent questions such as "what was the first language men ever spoke?" and "are all present-day languages descended from a single ancestor?" and "how do children learn languages so easily?"

Because the question has seldom been put with great precision, and because direct evidence bearing on it, however it is understood, is so meager, attempts to answer it have fallen into disrepute among linguists, almost as much as have investigations into the supernatural or perpetual motion among scientists. But however unsatisfactory most past inquiries into it have been, the question itself does not deserve scornful dismissal: language did originate at some time and in some manner, and any progress made in investigating that event or process will have much to tell us about who we are. So while we may never arrive at a fully satisfactory answer, there is nothing inherently foolish about attempting to reach it.

To avoid the kind of confusion that some past inquiries have gotten bogged down in, the first step is to form a perfectly clear view of where our inquiry begins, including all our presuppositions. What assumptions are we making about the origins of the human race itself? What conditions do we imagine earliest man found himself in? What do we count as language? Answering such questions is no guarantee of success, but failing to answer them practically guarantees failure. For the present inquiry, the answers are these:

• I assume that human beings, the species homo sapiens, are related to, if not descended from, various earlier hominids and, more indirectly, anthropoid apes. In other words, man is, for purposes of this inquiry, produced by whatever processes produced the earth's fauna in general, rather than by special creation.

• I assume that the earliest human beings — more briefly, man — lived by hunting and gathering, activities that entailed moving from one location to another as prey animals migrated or were locally depleted, and edible plants were consumed or went out of season; that he cooperated closely with others of his own immediate group — his tribe or extended family — but had little to do with other groups that might to modern eyes be indistinguishable from his own, even turning to hostility and warfare against them under stressful conditions.

• I assume that language means a system of recording, elaborating on, and communicating thoughts that enables its users to deal with matters remote from their immediate circumstances, and is not simply a body of involuntary or semi-voluntary sounds made in reaction to current stimuli.

I have no proof nor even private certainty that these assumptions are correct; I list them not in order to persuade anyone to accept them, but just in order to make clear where I begin my inquiry.

From Animal Sounds to Human Speech?

My next step is to ask, what would induce human beings to undertake the work of building a language? I assume that the earliest specimens of homo sapiens (and even earlier hominids) uttered cries just as nonhuman animals did, but these cries did not constitute language any more than animal cries do. How did those cries evolve into real language, then, as is widely supposed? What exigencies of their lives would be pressing enough to motivate early men to develop an elaborate symbolic system for discussing certain matters in absentia, and of referring to them in complicated ways: with respect to their place in time, their proximity in space, their relations to other things, and so on? The answer usually offered in response to this question is, of course, those that deal with the most urgent necessities of life: instructions to fellow hunters during the chase, warnings of imminent danger, and the like.

But the very urgency of these issues makes it unlikely that man, in trying to respond to them, went from making the kind of vocal utterances he shares with the beasts to something deserving of the name of language — he needed some simple and immediately available way of doing such things as warning others of imminent danger, and telling them that he'd located some game animals or other fugitive source of food. Communicating such urgent matters required speed above all, not some relatively complex and subtle tools that were difficult and time-consuming to use well, and that some members of his group — children, for example — might not yet have mastered. So I reject the idea that language developed in response to the need to convey urgent messages.2 Surely the need for such urgent communication was met by vocal utterances — but these would be little more than the cries emitted by apes in similar circumstances, although perhaps more consciously uttered than when uttered by animals — and not by anything amounting to language as I have just defined it.

Whatever motivated man to develop language must, on this hypothesis, have been of great importance to him, although not of the greatest urgency: how would he choose to deal with those needs? I suggest that it is unlikely that man's most natural course, when it came to finding some way to meet those needs, was to use his vocal apparatus for that purpose. Why should he have been attracted to the idea of issuing sounds through his mouth to fix and communicate ideas of a type quite different from the few, simple, urgent matters for which making such sounds was the natural, indeed largely involuntary, course? After all, man's principal sense is vision, not hearing — would he not tend to rely on his primary sense, his visual sense, for his more complex and longer-term needs? We are sight hounds: we say "I see" when we mean "I understand." When we want to praise someone's deep understanding, we say he has insight, or is a man of vision; when we want to deride him for being obtuse, we say he's blind. Until computers and photo-editing software arrived to disillusion us by facilitating the creation of illusions, we thought that seeing was believing. When we want to go undetected, we strive for invisibility, not inaudibility or inolfactibility. The Invisible Man is, for us, the invulnerable and nearly omnipotent man.3

What's Important Enough to Justify Creating Language?

I suggest that what called for and justified the labor involved in the creation of language were some features of man's life that, unlike the vital but momentary and ephemeral matters of urgent necessity, were instead stable, long-lasting, and relatively complex elements of his experience — that is, his own cultural artifacts and social structures. I conjecture that long before he had developed true spoken language, early man conveyed important information to others, and recorded it for himself, in some visual medium — marks on stone, on animal hides and bones, on suitable plant leaves.4 And in fact we know that early man produced sketches and tallies and primitive diagrams — we have the stick charts, or sailing diagrams, produced by the islanders of Oceania on which they recorded their knowledge of currents and winds, and the petroglyphs produced by very early peoples of what is now the southwestern United States, and the hieroglyphs of ancient Egypt, and the cave paintings of the Pyrenean region; such things, I suggest, both preceded true spoken language and motivated its development. Spoken language resulted, under this hypothesis, when man first adapted vocal sounds to stand for elements of what was originally conveyed by these conventional marks on stone, clay tablets, papyrus, or taro leaves.

These are the matters that, while not urgent, are yet important enough to be recorded as marks on some durable material, and then to warrant the development of something as complex and unnatural as a system of vocal sounds to represent, in turn, these primary representations. Such things as maps — of winds and currents, of the way to important fishing and hunting grounds, of the routes taken by migrating animals, of religious rituals that had to be performed with exact accuracy — were surely among them; and, as Professor Harris has suggested, counts of objects of value were another such subject: how many of the goats in this herd are mine, and how many yours? How many bags of barley do I owe you for this cask of beer? These things would have been recorded first, and most conveniently, graphically, but the trouble with such representations is that they are not always available when needed, and can be awkward to handle even when available. It would often have been convenient, even necessary, to refer to such pictorial records without actually having them at hand, especially when those representations were carved into slabs of stone, or incised into heavy baked clay tablets, or written only on friable plant or animal material, or stored under the jealous guardianship of tribal chiefs or medicine men.

A Picture Is Worth — and Blossoms into — a Thousand Words

If man cannot always have quick access to such graphic records, one thing he always has at his disposal is his voice, his capacity for making vocal sounds. I suggest that it was the keen desirability of being able to avail himself easily and quickly of the information preserved by such graphic records that induced man to start associating vocally produced sounds with the chief elements of those records; the desirability of being able to enjoy their benefits without actually producing and handling them motivated him to elaborate speech, that derivative and secondary way of representing information, until its capabilities were so advanced as to let it largely displace the physical handling of the graphic forms it was created to refer to. I conjecture further that the development of speech had among other effects that of liberating graphic representation; with speech taking on more and more of the immediate practical needs of man, the graphic alternative was freed to develop in ways that greatly enriched it — freeing it to become "art" — although possibly diminishing its value for its original purposes, now being dealt with by speech. And it is certainly plausible that there was back-influence from speech, the newcomer, on the course taken by graphics, the old traditional tool — it may well be that part of the development of writing was emulation of the new, elaborate speech elements, just as speech began by referring to graphic elements.

What Do Babies Learn First, and When Do They Learn It?

One fact that seems to lend support to the conventional belief that speech came first is that babies in present-day societies learn to speak before they learn — if they ever do — to read and write. And in a kind of homage to the ontogeny recapitulates phylogeny dictum, this may be taken by some as evidence that man as a species spoke before developing written language. But what we ask of our children when we teach them to read and write is to master a highly elaborate system that has been under development for millennia, not just to learn the symbols for a few animals of interest, and for showing the way to running water from our camp, as children would by my hypothesis be expected to do in prehistoric times. What we do in teaching them to speak — if it can be said that we "teach" them — is quite different. From the moment of their birth, and long before they can even hold a book or a pen, we expose them to spoken language (some mothers-to-be talk to their fetuses, or play educational recordings for them, unwilling to wait until they're born to start communicating with them). So if children today learn to speak long before they can read and write their language, that is no evidence that primitive man spoke before he wrote. There is also, perhaps, a trace of political correctness in giving speech priority over writing: everyone except the brain-damaged speaks, but relatively few write, so it would seem more democratic (in the modern sense, where that means egalitarian) to promote the claims of speech — the possession of the common man everywhere — over the elitist, snobbish practice of writing.

But in any case it is clear that the notion that speech preceded, and formed the basis for, writing is unfounded. Indeed, the second of these ideas — that writing is just a way of encoding speech as marks on some suitable medium — has been closely examined by Professor Harris in two books,5 and shown to be quite improbable. Speech and writing show many signs of having developed at least semi-independently, each largely following its own course (though doubtless having frequent interaction with the other); what I hope to have made plausible is that graphic expression, far from being merely a way of recording speech in a visible and relatively permanent way, is the precursor that made the development of speech both necessary and possible.

Pure Conjecture — Like Everything Else

It will be said of these conjectures that they are just that — conjectures — and are offered without a whit of hard evidence or "proof." That comment will be true, but hardly of much value. The conventional view — that man first talked, and only afterwards had the idea of developing some way of recording visually what he had long been saying — is likewise just conjecture, and conjecture that has the disadvantage of failing to come to grips with the points about early man discussed earlier. This conventional view has for so long been one of the things that "everybody knows" that questioning it almost always evokes shocked gasps.6 But when the shocked have recovered, and are asked to cite the evidence for what "everybody knows," shock becomes embarrassed silence; it turns out that no one has the slightest idea of why it should be believed, except that it's what all decent, normal people believe. Perhaps it is a child of, or cognate with, the linguists' dogma that speech is real language, writing at best a derivative and second-hand shadow of speech — practically an epiphenomenon.

Of scientific evidence the conventional view is totally bereft; indeed, if one wanted to be hard-nosed and legalistic, one might point out that all the direct evidence shows that writing is far older than speech, since we have inscribed tablets and fragments of vellum and papyrus that are thousands of years old, while the oldest record of speech is a nineteenth-century recording of Thomas A. Edison reciting "Mary had a little lamb."7 Direct evidence that anyone uttered an audible word before that date is nonexistent. Of course there are many written records that assert or imply that man spoke before then, but as linguists are always insisting, writing is a late development, barely language at all, and has little value as evidence for the history of language development.

An End Without a Conclusion

This study is clearly a work in progress, and ends at this point not because a conclusion has been reached, but because its author has for the moment run out of answers to the questions he has opened. In recognition of its inconclusive status, I have retained here the notes that I've written to myself about lines of inquiry to follow up and thoughts to develop. I invite any reader who has a contribution to make to this ongoing investigation to communicate it, and I will respond as appropriate — and of course give credit to such a contributor in any further development of this hypothesis. The great thing to bear in mind is that conventional "sound" thinking has made no progress whatever in the search for the origin of language, and that bold, imaginative conjectures are to be welcomed — not adopted blindly, but entertained, explored, used as heuristic tools. Truths are, in their infancy, metaphors; in youth, conjectures; in maturity, theories; in old age, clich�s. In this investigation, what is needed is more conjecture.

Article Footnotes

1. A good collection of such speculations is Roy Harris (ed.), The Origin of Language. Bristol, England: Thoemmes Press, 1996. A brief survey of "Theories of the Origin of Language" is offered in A. S. Diamond, The History and Origin of Language (Citadel Press, 1959), pages 258–275.

2 Nor do we today use language, spoken or written, to convey urgent, life-and-death messages if we can help it; we use red lights, red flags, sirens, alarm bells, skull and crossbones graphics, diagonally slashed icons, and so on.

3 Further indications of the priority of vision: (1) the differing attitudes toward eyeglasses and hearing aids — the former are no longer considered a mark of age or other shameful condition — men do make passes at girls who wear glasses, so much so that some girls wear glasses they don't need because they think glasses make them look more interesting, while hearing aids are regarded as signs of decrepitude; (2) our fear of darkness and of creatures that go bump in the night, the period when our vision is almost totally dependent on artificial light, and of little use without it; (3) "eyeing" something means considering it; (4) new discoveries are said to "shed light on" or "illuminate" problems; the saying, "Out of sight, out of mind"; (5) the terms "lucid" and "elucidate," (6) the military reliance on the "Mark 1 eyeball." And consider the difference between the lives of the deaf and the lives of the blind; the deaf, at least those who live in the modern world, communicate well with each other and with the hearing people who have troubled to learn sign language. In some cases, they form a closed society, refusing to see themselves as suffering a disability at all, and sharply rejecting attempts to restore some degree of hearing. The blind are quite different; they may get along fairly well, but they do not benefit much from the presence of others lacking sight, and for any but the most routine activities in well-known surroundings, they are dependent on the sighted. In short, if you go deaf, you can compensate to a great extent; if you go blind, you are severely disabled for all but a very limited range of activities, and lead a very circumscribed life. Note also the use of "seeing" someone as a euphemism for "having sexual relations with" that someone. Also consider that many are "afraid of the dark" — is anyone (other than a teenager, perhaps) afraid of silence? And note that when we wanted a pejorative name for some centuries, we called them the Dark Ages.

4 Richard Dawkins, The Ancestor's Tale (Houghton Mifflin, 2004), page 36: "Or perhaps representational art � was the bridge to referential language. Perhaps people learned to draw bison, before they learned to talk about bison that were not immediately visible."

5 Roy Harris, The Origin of Writing. London: Duckworth, 1986; and Rethinking Writing. Indiana U.P., 2000.

6 When an academic expert has to mention this dogma explicitly, he is usually embarrassed at having to utter such a commonplace, and covers that embarrassment with a little verbal moue. For example, from Dominique Charpin, Reading and Writing in Babylon (Harvard U.P., 2010), page 176: "The oral — who can be surprised? — came first: Mesopotamia was fundamentally a civilization based on giving one's spoken word. Writing was used only to document the agreements that had been concluded." That moue stands in place of any evidence for the generalization it accompanies; untypical for Professor Charpin, who is throughout the book quite hard on other Assyriologists who fail to provide evidence for their assertions. Perhaps part of the reason for his uncritical acceptance of the dogma of the priority of speech is to be found in a few lines from his Preface: "Some people are fascinated by the search for origins, but I must confess that is not so in my case. I am more interested in how a civilization that has reached maturity functions. That is undoubtedly a matter of taste, but it also reflects a preoccupation with being able to find fairly speedy confirmation — or invalidation — for the hypotheses formulated, which the many unknowns associated with the most ancient periods scarcely allow."

7 The Edison recording is described in, inter alia, John McWhorter, Doing Our Own Thing (New York: Gotham Books, 2003), page 18. But in March 2008 there came news of an earlier recording: a group of American audio researchers discovered in an archive in Paris a "phonautograph" recording of a French soprano singing "Au Clair de la Lune" on April 9, 1860, thus setting the terminus post quem for evidence of human speech — or at least song — back some seventeen years. See Jody Rosen, "Researchers Play Tune Recorded Before Edison," New York Times (March 27, 2008), pages A1, A18, and David Perlman, "1st known recording astounds scientists," San Francisco Chronicle (March 29, 2008), pages A1, A7. Further information at informationweek.com/1180/recording.htm.

Notes to follow up on

• Denise Schmandt-Besserat (Before Writing, volume 1: From Counting to Cuneiform. U. of Texas Press, 1992) thinks that the Near Eastern clay tablets she calls "tokens" were "a primitive accounting system and one which led to the creation of writing. Words, in a sense, began with numbers." And numbers — more precisely, numerals in the form of tally-marks — are just such graphics as I suggest language started with.

• From paleographic remains, it appears that hominids have had a modern vocal tract — that is, a pharynx and larynx — only since the Neanderthals, who emerged about 150,000 years ago. Is there evidence of graphics — artwork, maps, etc. — before their emergence? (Apparently not; Mithen, Prehistory of the Mind (Thames and Hudson, 1996), says the oldest surviving specimens of human art are 60,000–30,000 years old. But he also points out (page 156) that there is nothing primitive about those specimens, suggesting that there was a long artistic tradition and history behind them.)

• Although the present conjecture is, as I have freely admitted, without a shred of hard evidence to support it, it is superior to its competitors in one way: it tries at least to account for the tremendous effort mankind must have made to develop full-blown language. Many studies that purport to be about the origin of language are really about our species' acquisition of the physical basis of language, and particularly of speech: the growth of the brain, the development of the necessary vocal apparatus. Of course these studies are not really about the origin of human language, but about the evolutionary steps that may have given us the physical basis for speech. The great question, I think, is not when or how man acquired the physical capacity for speech, but how and why he made the enormous effort involved in realizing the potential given him by that physical development — another question altogether. Certainly having the physical capacity, both in our brains and in our vocal organs, is a first requirement — but once we had them, what made us use them so intensively? We all have legs, but we don't all become ballet dancers.

• From the Preface to Carl Darling Buck, A Dictionary of Selected Synonyms in the Principal Indo-European Languages (U. of Chicago Press, 1949 [paperback. ed. 1988], page v:

Yet such cries, which comprise all that primitive man inherited from his ancestry, must perforce be the starting-point of human speech. Primitive man, with his more highly developed brain, gained an awareness of a connection between the sound and something expressed by it.

Here the assumption that human speech must be an outgrowth of animal cries is made explicit, and in being made explicit is shown to be quite untenable. We human beings still today utter nonlinguistic sounds — animal cries — when astonished, terrified, or otherwise momentarily shocked out of our usual articulateness. But there is no evidence that this trait that we share with the higher animals ever leads, today, to the invention of speech; as the sad stories of feral children invariably tell us, children deprived of socialization during their first six years or so never acquire language, but remain limited to animal cries throughout their lives. Why is it to be supposed that primitive man was able somehow to build language on the foundation of animal cries when modern children, with brains presumably at least as advanced as his, cannot?