Fieldwork demands constant attention to details. And it can be hard in the jungle to continue to pay close attention day after day to anything, whether language or any other important part of life. Each day requires disciplined routine.
During the rainy season, there are frequent all-night downpours. I learned that it took only a couple of hours for a heavy rain to sink my boat. My motor was bolted onto the stern and it weighed about 150 pounds, so I could not remove it at the end of each day and put it on dry land. The motor stayed on the boat. But when it rained, the weight of the motor pulled the boat down just enough so that all the rainwater rushed aft. In an Amazonian rainstorm, it doesn’t take long for enough water to accumulate there to push the stern under the water and sink the boat—even though my boat could hold one ton of cargo.
So when I heard rain arrive about midnight, I knew that if it was a strong, hard rain, I would have to get up about 3 a.m. and walk in the downpour to my boat to bail the water out of the stern. This was paying attention to detail, part of the disciplined routine I tried to follow. But it was so hard to get out of a warm, comfortable hammock at 3 a.m. and go out into a driving rain, worrying about snakes and other animals, including Pirahã dogs, and walk through the village down to my boat. I knew I had to and I always did—except for one time.
It was pouring rain, but when I woke up I just couldn’t get myself to walk down to where my boat was moored, even though it was only about a hundred feet away. I told myself that the rain wasn’t that bad and that my boat would, after all, hold more than a thousand pounds of water before sinking.
As usual, about 5 a.m., I got up to begin planning my day. I noticed a smell of gasoline. Deep down I think that I knew something was wrong, but I didn’t want to admit this to myself. So I went about business as usual and was making my coffee when Xioitaóhoagí yelled out, “Hey Dan! Come see your boat!” I ran out of the house and down the path to the river. There was gasoline floating on the surface of the water. I saw the nylon rope that moored my boat stretched tight—in a nearly vertical line into the water. I walked to the edge and looked down. At the end of the rope, in some thirty feet of water, I could see my boat, sunroof still up.
I was one hundred miles by water from the Transamazon Highway. My boat was my only way out. I had no idea whether I could get it out of the water, whether I could get it running again, or what I would do if I couldn’t get it working. A group of Pirahã men and women came running to help. I got some twelve-foot-long ironwood two-by-four boards that were left over from constructing my house and I thought of a plan.
Several of us tugged on the boat until we moved it up a few feet to a submerged ledge on the bank. Then, with straining and red faces, we worked it into a still shallower area where it was only a few feet below the surface. I gave men two-by-fours and explained that we needed to use these as levers to work it bit by bit up the bank. After a couple of hours, we worked it up until the rims of the boat were just above the surface of the water. At that moment, without any prompting from me, women jumped in with gourds and started bailing out the water. We finally got about two-thirds of the water out of it. I tied its bow and stern to the shore and inserted a siphon hose into the built-in gas tank. I was able to drain out most of the water that had entered the tank. Since water is heavier than gasoline, I got the milky water-gas mixture out until pure gasoline began to exit the hose. I had about one-fourth of my gasoline left. Maybe it would be enough to get me to the road. But the pressing problem was to see if the motor would run. If not, I wouldn’t need the gasoline anyway.
The first step was to remove both carburetors and dismantle them, drying them out and coating the inside with rubbing alcohol. Then I removed and dried the spark plugs. Next I took a syringe and injected three cubic centimeters of alcohol into each of the motor’s cylinders. Finally, I pulled to start the motor. It started up on the third pull. Alcohol in the cylinders, though a bit of a risk for explosion, can really spark the gasoline. I took off and quickly got up to full speed, careful to stay within sight of the village, in case the motor died. Once the motor was warmed up I knew it would dry out the remaining water. I was pretty proud of myself.
Except that then I remembered that if I had simply gotten up for about fifteen minutes of light work during the night, none of this would have been necessary. Details. I read biographies of explorers and realized that success depended on hard work, planning, and attention to details. This attention to details would be a challenge as I began the study of Pirahã words, a far more demanding task than cleaning out a couple of Johnson carburetors.
And the analysis of the language was more important, even though not as urgent at the time, than fixing my boat. Pirahã’s importance for our understanding of human language ranges far beyond its sounds. It is in the grammar that the more profound challenges to most modern theories of the nature, origin, and use of human language lie. I was now coming to realize that Pirahã grammar was a particularly hard nut to crack for Chomsky’s hypothesis that specific grammatical principles are innate, as well as for much of his theory’s account of how the components of grammar work and fit together. Since the stakes in our conclusions about this issue are so high for our understanding of human language and the human mind, it is important that we work through all of this carefully.
The place to begin, at least following linguistic tradition in discussing the grammar of a language, is with the words. Sentences are built from words and stories are built from sentences. So linguistic studies tend to follow this order in discussions of grammars of different languages.
One of the first word groups I was interested in recording, because of its usefulness and because of what I expected to be its simplicity, was the set of words for body parts: hand, arm, eye, foot, butt, and so forth.
As usual, I was working with Kóhoibiíihíai.
“What is this?” I asked, pointing at my nose.
“Xitaooí.”
“Xitaooí,” I repeated, perfectly I thought.
“Xaió, xitaopaí,” he said.
Aargh, I thought. What is that -paí business doing at the end of the word?
So, naively, I asked, “Why are there two words for nose?”
“There is one word, xitaopaí,” came the exasperating answer.
“Just xitaopaí?”
“Right, xitaooí,” he said.
It took a long time to figure this out, but the -paí at the end of a body-part word (and it can occur on all of those, but on no other words in the language except body-part words) means something like “my own.” So xitaooí means just “nose” but xitaopaí means “my own nose.”The Pirahãs could no more tell me that than the average English speaker could tell me what to means in I want to go. Why isn’t it just I want go? Linguists have to figure this kind of thing out for themselves.
Aside from this, Pirahã nouns are for the most part very simple. They have no other prefixes or suffixes, they do not have plural or singular forms, and they don’t have any tricky features, like irregular forms and so on.
The lack of grammatical number in Pirahã is unique among the world’s languages, according to British linguist Greville Corbett’s book-length survey of grammatical number in the world’s languages, though now-extinct languages or earlier stages of spoken languages appear to have lacked number too. Thus there is no distinction between dog and dogs, man and men, and so on. It’s as though every Pirahã word were like the English words fish and sheep in having no plural.
So a sentence like Hiaitíihí hi kaoáíbogi bai -aagá is vague in various ways. It could mean “The Pirahãs are afraid of evil spirits,” or “A Pirahã is afraid of an evil spirit,” or “The Pirahãs are afraid of an evil spirit,” or “A Pirahã is afraid of evil spirits.”
This unique lack of grammatical number could follow from the immediacy of experience principle in the same way that the lack of counting did. Number entails a violation of immediacy of experience in many of its uses—as a category it generalizes beyond the immediate, establishing larger generalizations.
Although Pirahã nouns are simple, Pirahã verbs are much more complicated. Each verb can have as many as sixteen suffixes—that is, up to sixteen suffixes in a row. Not all suffixes are always required, however. Since a suffix can be present or absent, this gives us two possibilities for each of the sixteen suffixes—216, or 65,536, possible forms for any Pirahã verb. The number is not this large in reality because some of the meanings of different suffixes are incompatible and could not both appear simultaneously. But the number is still many times larger than in any European language. English only has in the neighborhood of five forms for any verb*—sing, sang, sung, sings, singing.* Spanish, Portuguese, and some other Romance languages have forty or fifty forms for each verb.
Perhaps the most interesting suffixes, however (though these are not unique to Pirahã), are what linguists call evidentials, elements that represent the speaker’s evaluation of his or her knowledge of what he or she is saying. There are three of these in Pirahã: hearsay, observation, and deduction.
To see what these do, let’s use an English example. If I ask you, “Did Joe go fishing?” you could answer, “Yes, at least I heard that he did,” or “Yes, I know because I saw him leave,” or “Yes, at least I suppose he did because his boat is gone.” The difference between English and Pirahã is that what English does with a sentence, Pirahã does with a verbal suffix.
The placement of all the various suffixes on the basic verb is a feature of grammar. There are sixteen of these suffixes. Meaning plays at least a partial role in how they are placed. So, for example, the evidentials are at the very end because they represent a judgment about the entire event being described.
The role of a verb in a sentence is crucial. So word structure is important to sentence structure. The meaning of each verb determines most of what needs to be in a simple sentence. Think about the English verb die. The meaning of this verb is what makes the sentence John died Bill sound bad. “To die” is something that happens to a single individual. If you know what to die means in English, you know that there are too many nouns in John died Bill, because dying is not something that you do to someone else. But we can say that “John caused Bill to die” or, more simply, “John killed Bill,” by adding to the meaning of die the meaning of cause. Thus John becomes responsible for someone else’s dying in the kill or cause to die sentence (both include the semantics of cause to die), so that John died Bill is ungrammatical but John killed Bill is fine. Changing the meaning structure, either by following the English options of adding other words, such as cause to, or selecting a related but nonidentical form, such as kill, alters the meaning of the entire sentence. As we study the role of verbs in sentence formation further, we see that most of the syntax of a sentence is little more than a projection of the meaning of the verb (some linguistic theories make this an explicit part of their theoretical apparatus).
Although I initially embedded my discussions of Pirahã grammar in Chomsky’s generative grammar, it became clearer and clearer over the years that this theory really had little enlightening to say about the Pirahã language, especially when culture seemed to play a role in the grammar.
According to Chomsky’s theory, what sets humans apart from all other terrestrial forms of life is the ability to use grammar. It is not the ability to communicate, since Chomsky recognizes that many other species communicate.
Certainly we have to know how to form sentences and compute the meanings of sentences that we hear or speak—some grammatical knowledge is thus vital for human speech. But since humans are not the only creatures that communicate, grammar cannot be crucial to communication per se. To live is to communicate. All living things, plants and animals and bacteria, communicate.
What makes the give-and-take of information within and across species possible? That is, what makes communication possible? There is a two-word answer: meaning and form. This is essentially what the great Swiss linguist, Ferdinand de Saussure, underscored with his concept of the linguistic sign—linguistic units are composites of form and meaning.
A bee communicates the meaning that food is near by the form of dancing. An ant communicates the meaning that a picnic is under way (though it might not use that term) by secreting chemicals, the form of ant communication. A dog communicates lack of aggression by specific forms—wagging its tail, barking, licking, and so on. And humans communicate meanings by the forms of making sounds or gestures.
But form is not all there is to human communication. Surely human communication differs from that of other species by more than a larger set of sounds, gestures, or words. There must be more to human communication than that. We are able to discuss much more complex matters and a much wider range of subjects than any other species. How do we do that? Two ways. The first and most obvious way is that we are smarter than other species. Human brains are the highest cognitive accomplishment of nature on this planet, so far as we know. The expression of this greater complexity of human thinking and communication requires tools that go beyond the tools available to other species. Linguists vary as to what they think these tools are, though there is wide consensus about several of them. My own vote for the most important tool goes for what the late linguist Charles Hockett labeled “duality of patterning.” There are different ways to conceive of this. But basically humans organize their sounds into patterns and then organize these sound patterns into grammatical patterns of words and sentences. This layered organization of human speech is what enables us to communicate so much more than any other species, given our larger, but still finite, brains.
We can illustrate the organization of sounds looking at an example similar (but not identical) to one we have already seen, using the simple words pin, pan, bin, and spin. Pin is formed by the sequence p + i + *n.*Think of each of these three positions of letters as “slots” and the letters themselves (p, i, n) as “fillers.” The slots represent the horizontal, or linear, organization of the word from left to right on the page or first to last as spoken sounds emerge from the mouth. The fillers are the vertical organization of the word. If we add a unit to the linear organization, we get a longer word, such as spin, by adding s to the front of pin. If we change things around in the vertical organization, we get different words of the same size, like pan from pin, when we substitute an a for the i in pin, and so on.
This is more complex than it might seem, though, because not just any filler or extension of a word is possible. We can add an s to pin to get spin, for example, but we cannot add a t, to get tpin. We can replace i with e to get pen, but we cannot place an s there to get psn, at least not if we want to form English words. This sound-based organization of the language is referred to as phonology. The physical nature of the individual sounds used in the organization is, roughly, phonetics. This is the first part of duality—the organization of sounds into words.
I should add immediately, however, that humans are resourceful, and that if for some reason they are unable to, or choose not to, use speech sounds, another channel of communication, sign language, is available. In sign languages the forms corresponding to sounds in spoken language are gestures or signs. Linguists have discovered that although the physical nature of gestures is obviously different from the physical nature of sounds, the organization of these elements into words and larger units, such as phrases and sentences, follows similar principles. Thus we can have a conception of phonology that includes both gestures and sounds.
Whether we use gestures or sounds, we need more than just words to have a grammar. Since grammar is essential to human communication, speakers of all human languages organize words into larger units—phrases, sentences, stories, conversations, and so forth. This form of compositionality is called grammar by some and syntax by others. No other creature has anything remotely like duality of patterning or compositionality. Yet all humans have this.
The Pirahãs certainly do. So consider the Pirahã sentence Kóhoi kabatií kohóaipí (Kóhoi eats the tapir). The Pirahãs put the object before the verb, a pattern we find in many languages, so kabatií means “tapir” and kohóaipí means “eat.” This shows us that Pirahã organizes its phonemes into words and its words into sentences. So the Pirahã language has duality of patterning and compositionality. It is hard to imagine a human language without these.
The most crucial component of language to my way of thinking, though, is meaning. Meaning is the gyroscope of grammar. I like this gyroscope metaphor because it expresses the belief of a large number of linguists, including me, that a slight difference in meaning, like the slight motion of a gyroscope, can lead to a large difference in the attitude of the rocket or the form of the sentence.
In other words, language is about meaning. We begin with a meaning and we encase it in grammar. All of grammar is guided by meaning. But what is meaning? That question has bothered thinkers for millennia. At the risk of biting off much more than I can chew, here’s my sketch of the core parts.
Philosophers and linguists talk about meaning in terms of its two parts, sense and reference. Reference is the use of language by the speaker and the hearer to agree on a specific object that they are talking about. So when two people in conversation use, say, the nouns boy, Bill, you, these words refer to entities in the real world. We know the boy or the person named Bill or who “you” is when we talk (or there will be severe miscommunication until both the hearer and the speaker agree on who or what they are referring to).
On the other hand, there are nouns that do not refer to anything. When I say that “John rode the unicorn,” it is pretty clear that unicorn doesn’t refer to anything in the real world. Likewise, if I say that “I will keep tabs on you,” tabs doesn’t actually refer to any object in this expression—it is part of an idiom. And there are things other than nouns that refer to things; for example, in I had built a house, had built includes a reference to a point of past completion. In The house is yellow, yellow refers to a particular color quality. There is disagreement on what it means to refer to things (some linguists deny that verbs and adjectives can refer) or how important this property is for defining parts of speech.
The other basic component of meaning is sense. We can understand sense as having two subparts. First, it includes the way that speakers think about entities, actions, and qualities—all those things we use in our speech. (What do I have in mind when I say “big,” for example, in big butterfly, versus big loss or big elephant?) Second, sense is about the relations between words and the ways that they are used.
Think of what break means in examples like John broke his arm, John broke the ice in the frigid conversation, John broke the sentence down for me, or John broke into the house, for example. The only way that we can know what break means is to know how it is used. And using a word means selecting a particular context, a set of background assumptions shared by the speaker and the hearer, including how particular words should be used, and the other words that the word in question is used with.
That is meaning in a nutshell: the way a word or a sentence is used, the way it relates to other words and sentences, and what speakers agree that a word or a sentence points to in the world. And the Pirahãs, like all humans, mean things when they speak. But that doesn’t mean that we all use the same meanings. Like all humans, what Pirahãs mean when they talk is severely circumscribed by their values and beliefs.
We learn, therefore, when we study words from any language, that we must understand each word at various levels simultaneously. We must understand a word’s cultural relevance and use. We must understand its sound structure. And we must understand how the word is used in context, in specific sentences and stories. Most linguists agree on these three levels of understanding the word. But Pirahã has also taught us something else. It has taught us that not only can the meaning of individual words be the result of culture, such as the closely related words for friend and enemy, but that the very sounds of the words, whether they are whistled, hummed, and so on, can themselves be determined by culture—and this latter lesson, which is abundantly illustrated in other languages, has not been discussed much in the linguistic literature. Pirahã gives us an extremely clear example for future linguistic investigations.