The Jiukiukwe Indians live in some swamps near the headwaters of the Orinoco. They are the most primitive people on the face of the earth, and their homeland is so isolated and dismal and totally lacking in anything of value that they will never be discovered by civilized men.

They do not know the use of fire. They wear no clothing, although on special occasions the married men do drape pieces of vines around their necks. They don’t tie them, however, for they have not discovered the knot. They build no shelters, and they use no tools, except for unworked rocks with which they bang on trees in order to dislodge the bark and expose the grubs and worms on which they live. Once in a while they find a dead fish floating in the water close enough to the shore to be hauled in without having to wade out too far. The waters teem with nasty creatures. They celebrate the occasion with a communal meal and a religious festival, at the climax of which the married men drape the pieces of vine around their necks. When the party is over, they squat happily in the muck and jabber away at each other in a language that contains thirty-four separate noun declensions, each providing forty-five separate forms, fifteen singulars, fifteen duals, and fifteen plurals.

Their verbs are even more complicated. The conjugations, eleven in all, indicate not only the usual things like person, number, tense, voice, and mood, but also the relationship of the speaker to his listener. A younger brother who is nevertheless not the youngest of his mother’s sons and who has never found a fish must speak, in the dry season, to the eldest of his paternal uncles in verb forms that would scandalize the mother-in-law of his youngest female cousin, especially in the rainy season.

The Jiukiukwe language has an enormous vocabulary in matters of interest to the Jiukiukwe. Although their technology is limited to the banging of trees with stones, they have scores of words to describe it. They have separate and unrelated words for flat stones, round stones, big stones, little stones, sharp stones, and so on, but they have no word for “stone.” In the case of trees, the vocabulary is even larger, since trees in which grubs or worms have been found are distinguished individually from one another by words that amount almost to “names” and are devised as needed, while unproductive trees or trees not yet banged are named not for any physical attributes but for their location with reference to the nearest tree that has provided grubs and worms. The whole system is duplicated with utterly unrelated words for fallen trees. They have, however, no word that simply means “tree.”

Since they use their knuckles and fingertips for counting, they can count only up to thirty-eight. (They count three knuckles on each finger but only two on the thumb.) Counting begins with the knuckle at the base of the left little finger, moves out to the tip, continues starting with the knuckle at the base of the ring finger, and so forth. Each hand thus provides nineteen units. Every knuckle and every fingertip has its own name, and those names are also the names of the numbers. They have, however, no names for the toes, and, while they do speak of the arms and the legs, they use one and the same word to name the ankles, knees, wrists, and elbows.

“Correct” social behavior among the Jiukiukwe is entirely a matter of doing and saying the right thing to the right relative under the right circumstances. Accordingly, the vocabulary and grammar of kinship are very large and complex. They have separate words for every possible degree of familial relationship. Not only, for instance, is there a special word for the oldest son of your mother’s next youngest sister, but there is yet another word for him should he have reached that estate through the death of some older brother. In either case, he is called by still another name until sunset on a day when he has found a dead fish. However, since all the Jiukiukwe are related to one another in some precisely nameable way, they have no need for words that mean things like “family” or “relative” or “kinship.”

There is a curious thing about the way they use their verbs. They have, of course, both passive and active forms, but they consider it a serious breach of etiquette amounting almost to sacrilege to use the active form when speaking of persons. In a child, the use of the active voice in the first person singular is taken somewhat less seriously, but it is still discouraged as a mark of arrogance or aggressiveness. Indeed, their words for “angry” and “insane” both contain an element of the ending that goes with the first person singular in that conjugation most often used by young children. They do not say: “I am eating my worm.” They say rather: “With regard to the worm unto me, there is an occasion of eating.” Animals and objects, however, are normally found as subjects of active verbs. The sun rises and the worm crawls, subject only to those forms available to that person who is saying those things and to whom he says them.

Their language sounds terribly complicated, and it is. It is every bit as complicated as English, or any other language, for that matter. All languages are complicated beyond hope of complete description. When it seems to us that German is less difficult to learn than Arabic, what we have noticed is not that German is less complicated than Arabic but that German is the more like English. Speaking his language is the most complicated thing a human being does, and should he undertake to go even further and learn to read and write it, he multiplies one infinitude of complications by another. It is an awesome marvel that anyone can do any of these things, never mind do them well.

Nevertheless, billions of people speak and understand a language. In fact, unless there’s something wrong, every human being there is speaks and understands at least one language. Every member of Homo sapiens ever born spoke and understood a language, unless, of course, he died too soon or was in some special way disabled. The ability to use language is included in the meaning of sapiens. We have no other way of being sapiens except through language. The Jiukiukwe may lack barbecue pits and some of our other things, but they are every bit as sapiens as the inhabitants of Manhasset. They have all it takes.

Still, they are different from the inhabitants of Manhasset in many ways. The material differences come easily to mind, since the Manhassetites have not only barbecue pits but much more, but there is a much more important difference than that. It is this: In the same circumstance, the Manhassetite will say, “I want food” and the Jiukiukwe will say, “As for me, there is hunger.” Every other difference is because of this difference; this is the difference between the Manhassetites and the Jiukiukwe, the difference from which all smaller differences flow.

The Manhassetites speak a language in which the typical statement takes the form of a sentence that names a doer and his deed. The most common elaboration also names the “object” of his deed. “I want food” displays exactly the typical structure of the most ordinary Manhassetite utterance. The structure may be modified and elaborated in many ways, some of them quite extensive and complicated, but it remains the enduring skeleton of the typical statement: A doer does something, often to something or someone. The continuous reappearance of this structure has taught all Manhassetites a particular view of the world and man’s place in it. They understand the world as a place where doers do things. That is why many of them will get raises next year and dig bigger barbecue pits.

The Jiukiukwe, on the other hand, have been taught by the basic structure of their language that doing is properly the business of the things in the world around them. Nor do they think of themselves, again because of their grammar, as the “objects” of the things that are done in the world. For the Jiukiukwe, the inanimate or animal doers of deeds do them at most “insofar as he is concerned,” as though he were, if not always an unaffected bystander, at least no more than accidentally related to what happens in the world. The Jiukiukwe are just there; the world does its things around them, sometimes “in their case.”

They will not get any raises next year, and you can easily see why they have no barbecue pits to enlarge. Technological change comes about when somebody does things to something. The Jiukiukwe have always lived, and will always live, exactly as they do today. Their technology will not change unless the basic structure of their language changes, although it may also be possible that the basic structure of their language would change should their technology change. There’s no way of knowing which must come first, if either, but it seems more likely that the language must change before the technology unless some imported technology should come along and eventually force a change in the language.

Imagine that some particularly eccentric or mildly demented Jiukiukwe should develop the rude habit of speaking in the active and saying things like “I will find worms.” He has now announced, to the Jiukiukwe way of thinking, some purported fact about the world and has at the same time subjected himself to considerable social disapproval. If he’s to get back into the good graces of his older sister’s father-in-law, and of everybody else, he had damn well better come up with the worms. The more the better. Then his arrogant statement might be re-understood and perhaps accepted indeed as a statement of fact about the world. In desperation, he might well discover that you can find more worms by prying off the bark with a sharp-edged stone than by banging the tree till the bark falls off. It might occur to him that some sharp stones are easier to hold and manipulate than others. Remember, he’s going to work hard; they’re all waiting for him to find worms and thus justify a statement in which he spoke of himself as one might speak of the sun or the moon. That’s serious. It won’t be long before he finds an obviously broken stone that works very well, and then it will come to him that he might bang some of those less efficient round stones together until they break and turn into good worm-diggers. Out of the active voice, a technology will be born. Before long, his relatives, noticing how plump and healthy he looks, will learn to copy both his magics, his verbs as well as his stones, and that will be the end of civilization as the Jiukiukwe know it.

When a Manhassetite faces a problem, he asks, drawing upon the basic structure of his language, “What shall I do?” He looks for an action to perform. The Jiukiukwe is unlikely even to think that he faces a problem, since that itself is a case of an agent doing something to something. Significantly, the Jiukiukwe language has no word for “problem.” “Problem” can be thought of only in a language that can also think of “solution,” and the relationship between the two is understood through a grammar that permits the idea of doers doing things. The Jiukiukwe do not think of a shortage of worms as a “problem,” a condition whose very name suggests that somebody might do something about something. They think of it as a “badness,” “a state of few worms in relation to us,” a condition in the world that just happens to affect them.

Although the Jiukiukwe seem to pay a heavy price for their grammar, they also take from it some advantages not available to the Manhassetites. They have, for instance, no warfare, because warfare not only arises from the willed deeds of agents but is itself a matter of willing agents doing things to each other. In any case, they don’t even have the individual analogues of war: hatred, envy, and competitiveness. The Manhassetites are a small subgroup of a large, warlike tribe, which is, in turn, only one of many tribes loosely associated into an enormous culture in which warfare is a permanent institution and even the root of much of its most vigorous enterprise. Naturally, hatred, envy, and competition are almost universal among individual Manhassetites. For the Jiukiukwe there is essentially only one doer of things, and that is something like the world itself, which does what it does neither out of will nor out of design. It has no intentions; it just happens. For the Manhassetites, there are as many possible doers as there are members of the species or nouns in the language, and their grammar encourages them to envision a universe in which conflict of intentions is simply a part of the fabric of reality.

An idea of reality is what we devise and perceive through our language; reality itself is probably something else again. Both the Jiukiukwe and the Manhassetites fancy that they know the real world, but what they know is some presumed order of things symbolized and suggested by the vocabularies and structures of their languages. They live by grammar, all men do. That’s why the grammars of all languages are so terribly complicated. There are no people, however “primitive,” who see the world as a simple place. In fact, the more “primitive” they are, the more complicated and elaborate the assumed underlying structure of reality in their languages. Furthermore, there seem to be no people who are content to have a language in which to consider only the world of sensible experience, and all languages are anchored mostly in other worlds rather than the one that we experience here and now.

It’s fun, and safe, to speculate on the origins of language. What makes it fun is obvious, but what makes it safe is that no dreary scholar will ever come along with the facts to prove what a fool you’ve been. If you speculate on Milton’s toilet-training or the social structures of the leaf-cutting ants, the very next mail will bring you a fat journal containing the definitive findings. The mystery of the origin of language is the linguist’s equivalent of the physicist’s mystery of the nature of the universe, if any, before the big bang. The evidence we need is utterly inaccessible, and there’s no way to draw valid conclusions from the evidence we have. Even experimentation is impossible. The story has been told, for instance, of several famous despots, that they had some newborn children raised from infancy by deaf-mute nursemaids to see what language they would speak. It turned out that children thus raised do not, after all, grow up speaking Hebrew, the supposed language of God and the angels. They grow up, of course, speaking nothing. Learning your language seems not at all analogous to developing one. Such an experiment would tell us nothing about the origin of language even if we could keep it going for thousands of years and watch generation after generation. The brains and even the vocal apparatus of those in whom language began were significantly different from ours, and our isolated infants would begin with an evolved propensity for language. It would surely be an interesting experiment, but it would tell us nothing about the origin of language.

It is just as much fun, however, and somewhat more useful to speculate about something that was probably not the origin of language, although we sometimes carelessly think that it might have been. There is a kind of everyday, commonsense notion about the origin of language that sticks in our heads and causes important misunderstandings. It is, of course, the notion that language must have begun as a way of naming things in the world in which we live. Those are exactly the things that need no naming.

The world in which we live is very tiny; it is as tiny for the Manhassetites as it is for the Jiukiukwe. That world can only be the world of immediate sensory experience, the world we can perceive in whatever way we can in this moment, which is now gone. The world of sensory experience is so tiny and so brief that, in a sense, we can’t do anything in it; we can only be in it. The world that was before this moment is immeasurably big, and so too the world that will be, to say nothing of the world that might have been or the world that may yet be or, the root of morality, the world that someone thinks should be. It is the main business of language to evoke such worlds.

The speechless animals live entirely in the world that is. It is sometimes said that some of the animals have a “language” with which they send signals to one another about the world that is. That’s nonsense. That we can uncritically accept such nonsense is testimony to the power of grammar. “Animals send signals” is a satisfying appearance of the basic structure of our language, and it harmonizes sweetly with our concomitant idea that the world is a place where most of what happens can be understood as the act of an agent who is doing something. The animals are not agents committing acts. When a zebra out on the edge of the herd sniffs a lion in the tall grass, he does not say to himself in any fashion, “I had better tell the others.” (Nor would you, for that matter.) He simply does what is appropriate for a successful zebra to do under those circumstances. His startled neighbors, startled by what he does whether they sniff lion or not, do likewise. That’s part of how they got to be grown-up zebras in the first place. The zebras who are slow to startle have a way of dropping out of the herd early in life. In a moment, the whole herd is in flight, but it cannot be properly said that a zebra has sent a message. It would be more accurate to say that the zebras have caught something from one another.

Human beings, too, catch things from one another, and usually with no use of language at all. We can imagine the hominid and speechless precursors of man wandering not far from the zebras. What need could they possibly have had to call the lion by a name? Just as surely as the zebras, they must have reacted appropriately to a whiff of lion. If they hadn’t, they wouldn’t have been around long enough to provide a future for all of us. What need, for that matter, would they have had to name the food they were eating or seeking, the food of the world of experience? Would they have had to be “told” to feed the young? To sleep at night? No animals need names for such things, because they do not have to “tell” of them. Such things, in their seasons, are just there. Language is for telling, not for naming. Nobody needs to be told that he is getting wet in the rain or that he is eating a banana.

Think what happens when you encounter an unexpected snake, the usual kind. There is no word in your head, no language at all, in the instant of automatic recoil. Then you say, if only in your head; “Oh, a snake.” The word “snake” makes you feel better, because it opens the gates for many other words. It occurs to you, because you have the words in which it occurs to you, that most snakes found in Manhasset are harmless and unaggressive and that you’re not really in danger. You have transformed a creature in the world of sense experience into a whole system of related ideas in a world that is not the world of sense experience. That little chill was real, and, for a moment, it remains, but you are not in flight. You are in another world, the world where snakes are made of discourse, not blood and bones and teeth. Your language gives you access to that world.

The Jiukiukwe are also afraid of snakes, and they have better reasons than we. The swamps are full of them, and none of them is harmless. Members of the tribe sometimes die of snakebite, and little children out looking for dead fish have occasionally been swallowed by anacondas. In the company of an anaconda, any forest animal is afraid, just like any Jiukiukwe. The difference, though, is that the Jiukiukwe have a language so that they don’t have to wait to be afraid until they actually find themselves in the company of the anaconda. They can be afraid ahead of time. In language they remember the anaconda of the past and take thought for the anaconda of the future. Those big snakes are not around to inspire fear; they are in other worlds. Language evokes them.

The Jiukiukwe use a part of the word for anaconda to describe those places where the snakes have been seen in the past. With the appropriate verb ending, of course, the same word is used to shoo children away from things. This is actually a form of technology, or, at least, social engineering, since it makes the children anxious and more than ordinarily observant when they are in places that carry that name. In spite of the obvious fact that there aren’t always anacondas to be found in those places, the language behaves as though there were. Some world other than the world of experience is projected by the use of an adjective. When we can project such an alternative world, we can also find, sometimes, a way of bringing it home and thus changing the world of immediate experience.

In language, man can project a world in which he doesn’t get soaked by every rain that falls. In that language, he can also reconstruct some past world in which he found himself under an overhanging cliff that kept him out of the rain. He can start looking for cliffs. He can eventually project a world in which some artificial cliff keeps him out of the rain and provides him with a quiet, comfortable afternoon in which to refine the invention of architecture. Such projections require language, not merely the naming of things.

People who have merely come up with a word for “wet” can do nothing more than stand around in the rain announcing to each other a sorry fact that needs no announcing. It won’t help them, either, to come up with a word for “dry.” What they need is a way to think about “dry” even while they are getting wet, a way to relate the two even when only one is present in the world of experience. They need “wet could be dry.” That’s grammar.

A collection of names for things in the world, however large, does not make a language. A language is only incidentally in the business of naming things. Its important business is to explore the way in which things are, or perhaps might be, related to one another. Building a shelter takes more than words for “dry” and “cliff.” It needs an idea of relationship, the idea of “under.” Then it needs another relationship, one that might be understood by something like “dry under made cliff.” To that, some designing mind must add not only “tomorrow” but “all tomorrows.” “Cliff” names something in the palpable world, and “dry” names not exactly a thing in the world but at least a physical condition. Those other words, however, “made” and “tomorrow” and “all,” name nothing in this world. They name some ways in which things can be related to each other.

How such words came to be is a profound mystery, every bit as vexatious as the mystery surrounding the birth of language itself. Perhaps, even probably, they are the same mystery, for there certainly could be no such thing as language without them. The ability to discern and name the relationships suggested, for instance, by words like “so” and “if” was more important to the history of man than fire or the wheel, neither of which would have been ours without “so” and “if.”

Even many of the words that we think of as the names of things in the world are really descriptions of relationships that exist only in the world of language. The Jiukiukwe, remember, have no word for “stone,” since for them stones as stones have no interest. To them, what we would call a pile of stones is a pile of treebangers, a name not intended to point at the things themselves but to their place in the life of the Jiukiukwe. To some other crowd of savages the same pile of stones might well be a pile of “weapons.” Stones are indeed found lying around on the ground, but weapons and treebangers are found in systems of culture.

Similarly, when the Jiukiukwe have a word for every possible family relationship, it’s not because they are naming precisely this or that person who actually exists before their very eyes. They are naming places in a complicated social design, not people. The social system exists in their minds, but it can exist there only because the language sets it forth for the mind to understand. They would say that we are primitive in our appreciation of family life because we can name so few relationships, just as we say that they are primitive because they use the same word for “knee” and “elbow.” From their point of view, we are just that backward because we call all the sons of our mothers by the one word “brother.”

People who can remember only one line of Wittgenstein are likely to remember his terse suggestion that we ought to keep our mouths shut about matters whereof we can tell nothing. This dictum provides some interesting corollaries. One of them is that in matters where we do keep our mouths shut, we can have no knowledge. If, for instance, there should exist something that a language has no way of expressing, then that thing does not “exist” for the culture that speaks the language. For all their compulsive attention to their relatives, the Jiukiukwe, as you know, have no word for “family.” They are all related. Since we are not, we need the word “family” as a term of distinction. In like fashion, every word isolates the thing named from everything else that there is, whether it be the name of some object in the world of experience or some idea of relationship in the world of language. In effect, everything that gets a name gets it because we can perceive that everything else is not that thing. Even the simple abstraction “stones,” therefore, is for us just as much a recognition of a relationship as “treebangers” is for the Jiukiukwe. The relationship of “stones” to everything that is not “stones” we can perceive. Of those relationships that we do not, for one reason or another, perceive, we cannot speak, and thus we can have no knowledge.

Our knowledge is made up of the stories that we can tell, stories that must be told in the language that we know. (Even mathematics is a “language” that states propositions and tells stories. It’s a very elaborate form of “play” language. That’s why it’s such fun for those who speak it well.) Where we can tell no story, we can have no knowledge. That’s probably why we all have amnesia about infancy. As infants, we hadn’t the language with which to transform the world of experience into the more durable form of knowledge and memory. When our forebears had no language and no knowledge, the best they could do was sniff the lion in the long grass. That wasn’t too bad, of course, and thousands of other kinds of creatures have done quite well with nothing more. But when we found language and became Homo sapiens (how could those events have been anything other than simultaneous?), we could tell ourselves the story of the last time we passed this way and lost Gbloog to a lion. Even more important, we could invent the story of the next time and either be prepared for a lion or decide to take the long way around. To deal with the lion here and now, we don’t need language, just a good nose and strong legs. Language can deal with the lion that was and the lion that will be, even the lion that may be. Creatures who can speak can survive even if they have bad noses and short, stumpy legs.

In those times, linguistic ability must have brought an even better hope of survival than physical prowess. The silly boob who just couldn’t get into his head the implications of the future conditional, accordingly, was just a hair more likely to be eaten by lions than his lame but more intellectual brother. A new element had appeared in the process of natural selection. Nowadays, we notice that there is very little danger of being eaten by lions. Nevertheless, the destiny that waits in the long grass for the silly boob who can’t get his language straight is not a good one, although, of course, it does last longer than the business of being eaten by a lion. It’s more like being eaten by a worm, slowly.

Chapter 3: A Bunch of Marks