In everyday speaking, a “language” is a system of communication used by people from a specific region or culture. Mandarin, Hindi, Spanish, and English, for example, are languages—the four with the most native speakers.

In psychology, however, the word “language” usually does not refer to a specific language, but to language in general: the ability to communicate using the words and rules of a language. Language is a communication system in which sounds, or non-auditory symbols such as hand gestures, have meaning (Hauser, Chomsky, & Fitch, 2002). In psychology, the branch of the field that studies how people communicate meaningfully through the use of language is known as psycholinguistics.

Language is not the only way that organisms communicate. Bacteria communicate by sending chemical signals (Pai & You, 2009). Animals communicate about their personal territory by leaving scents; that’s what your dog is doing when stopping at every tree in the neighborhood. However, these are not examples of language, which has two properties that distinguish it from these other forms of communication:

314

Now let’s see how this unique mechanism of communication, language, is organized.

Language is highly structured. You can’t make any sounds you want and expect to be understood. You must follow language’s rules—and there are a lot of them, at multiple levels of organization. This sounds hard, yet it turns out to be easy. One of our mind’s most amazing powers—one you probably took for granted before now—is the ability to effortlessly follow multiple rules, at multiple levels of language organization, simultaneously. Once you consider these levels of organization, you will more deeply appreciate the power of the human mind.

315

LEVELS OF ORGANIZATION. Suppose you want to talk about your experiences in psychology (Figure 8.2). At the highest level of organization of language, you’re engaged in a conversation. Conversations have rules (Grice, 1975). If a friend says, “I have been feeling really depressed lately,” and you reply, “Don’t you love my new hat!” then you are violating a rule—the implicit rule that people should cooperate with each other when conversing.

Conversations are structured by different points of view, or conversational “positions,” that people adopt (Harré & van Langenhove, 1999). At one moment, you may provide information to someone. A moment later, you may ask a question; your position shifts from “information provider” to “question asker.” Different positions bring different obligations, such as to provide information or to listen carefully.

In any conversation, the meaning of statements depends on the setting in which they occur (Austin, 1962). Suppose, at a dinner table, you say, “Salt.” Its meaning depends on the setting. It could be a command (to pass the salt), comment (that the food’s too salty), or answer (if someone has asked what you just added to your food).

Moving down one level (Figure 8.2), language is organized into sentences. Sentences bring remarkable power. With them, you can communicate about not only your immediate environment, but also those millions of miles (“It must be cold on Pluto”) and billions of years away (“The Big Bang must have been really loud”). Sentences can even refer to things that never existed (“When conversing, how do unicorns avoid poking each other’s eyes out?”).

At the next level down are the main parts of sentences, which are phrases. The syntax of a language is the set of rules that determine whether a particular series of phrases forms a sentence that is grammatically correct. Some arrangements of phrases are correct (“Psych professors tell the funniest jokes,” “Psych professors wonder why people differ”), whereas others are not (“Psych professors wonder the funniest jokes”). We discuss syntax in detail below.

One more level down, phrases are made up of words. Individual words can convey meaning by themselves: “No,” “Eureka.”

Moving down another level, words contain parts that convey meaning. “Funniest,” for example, contains two parts: “funny” plus “est,” which conveys extremity. “Unhappiest” contains three parts: the emotion (happy), extremity (“est”), and a reversal of the emotional state (“un”). These parts are morphemes. A morpheme is the smallest part of a language that conveys meaning. Morphemes can be as small as a single letter—for example, the letter “s” after “professor” (Figure 8.2) conveys plurality.

Finally, at the lowest level of analysis, language consists of sounds. Phonemes are the smallest units of sound that convey meaning in a language. Here, too, language is organized; some sounds patterns are common in a language, whereas others are not. Different languages use different sounds. Hawaiian does not contain some sounds found in English (e.g., the sounds of the letters B or Z). The Chinese language Mandarin uses pitch (how high or low your tone of voice is) to convey meaning to a much greater extent than does English (e.g., Klein et al., 2001).

Let’s now focus on one of these levels: syntax.

316

SYNTAX AND SENTENCES. Consider this sentence: “Colorless green ideas sleep furiously.” Odd though it may seem, this statement is a sentence. “Furiously green sleep ideas colorless” is not a sentence, but “Colorless green ideas sleep furiously” is.

Noam Chomsky, a linguist at the Massachusetts Institute of Technology, composed the sentence to show how syntax works. Syntactic rules, Chomsky explained, are like rules of mathematics. In a mathematical formula, you can plug in any numbers and obtain a valid answer; the area of a circle is π × radius² for circles of any size. Similarly, with syntactic rules, you can plug in any noun and verb phrases and get a syntactically valid sentence. If “Crotchety old Ida sleeps fitfully” is a sentence, so is “Colorless green ideas sleep furiously.”

Chomsky set out to discover the rules of syntax, that is, the rules that govern the grammatical correctness of sentences. The following sentences give you a sense of these rules:

“Psych professors tell the funniest jokes.”

“The funniest jokes are told by psych professors.”

“Jokes—the funniest ones—are told by professors of psych.”

“It’s psych professors who tell the jokes that are the funniest.”

The sentences differ in their wording. Yet, at a deeper level, they’re the same; they express the same concept. The first sentence says it most simply, with a noun phrase (“psych professors”) followed by a verb phrase (“tell the funniest jokes”). The other sentences simply shift these components around; in the second sentence, for example, the noun phrase shifts from the sentence’s beginning to its end. In the study of language, these shifts are called transformations.

You now can see what rules of syntax do. They specify how a core sentence such as “Psych professors tell the funniest jokes” can be transformed. A transformational grammar is the set of rules governing how components of a sentence can be shifted around to create other sentences that are grammatically correct. (The words “syntax” and “grammar” have overlapping meanings. Both refer to rules that govern language use. “Grammar” includes rules of syntax that determine whether a series of words is a correct sentence, as well as rules of style such as “do not split infinitives.”)

These are facts about language. Here’s an amazing fact about psychology: All native speakers of a language know how to use its rules of transformational grammar, even if they cannot say what those rules are. You know “Psych professors tell the funniest jokes” is a sentence and “Psych professors wonder the funniest jokes” is not, even if you can’t identify the grammatical rule that the latter sentence violates. Furthermore, you used grammatical rules correctly at an early age, before you ever took a course on English grammar. How did you learn the syntax of your native language? This is a question of language acquisition.

317

Everybody acquires language—and fast! Children know about 13,000 words before they start school (Pinker, 1999). They also can understand and produce sentences in active and passive voice, and present and past tense, before they learn what “voice” or “tense” mean. Three types of theories have been proposed to explain how children acquire language skills.

REWARDS. The first theory is famous for being more wrong than right. The learning psychologist B. F. Skinner (1957; see Chapter 7) believed that children learn language by experiencing environmental rewards. If a child looks at a cookie and says, “Cookie,” the behavior is rewarded: Adults praise the child (and may give her the cookie). Because people tend to repeat behaviors that, in their experience, have been rewarded, rewards could drive language acquisition.

But there’s a problem with Skinner’s theory: It cannot explain how children learn grammar even when they are not rewarded for doing so (Chomsky, 1959)—and they usually are not. Parents rarely reward grammar use; “in general,” researchers find, “parents [seem] to pay no attention to bad syntax” (Brown, 1973, p. 412). They reward statements that are factually correct, even if grammatically flawed. Yet children learn grammar.

INNATE MECHANISMS. In a second type of theory, Chomsky (1965, 1980; Hauser et al., 2002) argued that language ability is innate. We don’t “learn” grammar any more than we “learn” breathing or digestion; each is done automatically by evolved biological mechanisms (see Chapter 4). Chomsky proposed, specifically, that people inherit a mental mechanism—a “language-acquisition device” (Chomsky, 1965, p. 32)—dedicated to processing the syntax of language. Without any instructions from parents, as in the case of the isolated twins in our chapter opening, children’s language mechanism extracts meaning from the speech sounds they hear.

There, of course, is some role for learning as well—children learn the specific language spoken in their society. But, in Chomsky’s theory, the overall capacity to acquire language is inherited. Language, then, is an instinct (Pinker, 1994), that is, a behavioral tendency people possess due to biological inheritance. Birds fly, fish swim, and people talk. Even a brief experience with other speakers—a “triggering experience” (Anderson & Lightfoot, 1999, p. 698)—will switch on the inherited language mechanism.

According to Chomsky, humans’ inherited language mechanism contains a universal grammar, which is a set of rules that enable people to understand and produce sentences. It is “universal” in that all human beings possess the grammatical ability. Chomsky therefore predicts that all human languages will have common features, including sentences containing noun and verb phrases.

318

Chomsky’s proposal has been accepted widely, but not universally (see This Just In). In a third type of theory, learning plays a larger role than Chomsky envisioned, but the learning is more complex than the reward process Skinner proposed.

THIS JUST IN

STATISTICAL LANGUAGE LEARNING. The third approach to the study of language acquisition is called statistical language learning (Rebuschat & Williams, 2012; Romberg & Saffran, 2010). It proposes that children acquire language by learning patterns of sounds and words they hear frequently, in other words, that are statistically common. According to statistical learning theory, then, the key mental mechanism in language learning is a general ability to distinguish between frequent and rare events.

Here’s the type of question statistical language learning theory can answer. If you say “pretty baby” to a baby, how does the child know you said “pretty baby” and not “pri tee-bay bee” (i.e., three words, the second of which is formed by the letters “tyba”). The explanation, according to statistical language learning theory, is that the baby hears some sounds more frequently than others. In English, “pre” occurs at the beginning of words far more often than “tyba,” and “pre” is always followed by at least one syllable to form a word. The child’s mind thus automatically guesses that “pre” and “ty” go together—merely because they have gone together so frequently in the past (Saffran, 2003). Evidence suggests that not only individual words, but also grammatical rules, can be learned through experiences (Elman, 1991; Rebuschat & Williams, 2012).

319

Statistical language learning represents “a new view of language acquisition” (Kuhl, 2000). Like Skinnner’s approach, it emphasizes the role of experience in learning language. But unlike Skinner, it identifies an inherent mental ability that infants possess that enables them to learn language: the ability to identify the frequency with which various speech sounds are heard. Statistical language learning differs strikingly from the approach of Chomsky. Unlike Chomsky, who proposed that knowledge of grammatical structure is innate (i.e., an inherited product of human evolution), statistical language learning theorists propose that grammar is learned through experience. This emphasis on experience suggests that, in language acquisition, culture plays a substantial role. To acquire a language, you have to coordinate your behavior with that of others in your culture (Chater & Christiansen, 2010). To communicate, you have to communicate like them.

The cases of Natasha in Siberia and of the twins in San Diego, from this chapter’s opening, can be addressed from either an innate or a statistical learning perspective. Chomsky would say that the twins, who developed their own language, provided each other with “triggering experiences” that turned on their innate language mechanisms. Statistical learning theorists would suggest that the twins learned to coordinate their communications with each other. Both perspectives would predict that the isolated Natasha would fail to develop human language. She lacked a triggering experience—communication with other people. And, sadly, growing up with dogs, she learned to coordinate her vocal sounds with theirs.

CULTURAL OPPORTUNITIES

Let’s shift levels of analysis, from the mind to the brain. What parts of the brain are most involved in language use? We’ll look at some early discoveries in this field of study, and then at recent brain-imaging evidence.

321

EARLY DISCOVERIES. In the nineteenth century, French physician Paul Broca had an unusual patient. A man named LeBorgne had, Broca explained, “all the freedom of his intelligence and his movements … but he cannot speak” (Schiller, 1979, p. 174). He could understand speech; when asked a question, he gestured in ways that revealed his understanding. But he could not put his thoughts into words. LeBorgne suffered from aphasia, an impairment of language abilities that occurs while other mental abilities remain intact.

When LeBorgne died in 1861, Broca performed an autopsy and examined his brain (Schiller, 1979). There was disease-related damage in a region of the left frontal lobe (Figure 8.4) that, today, is known as Broca’s area. Because the patient (1) could not speak and (2) had left-side frontal lobe damage, Broca concluded that this region of the brain is needed to produce speech (Rorden & Karnath, 2004). The study of subsequent patients confirmed his conclusion.

People with damage to Broca’s area may be able to produce some meaningful speech, but their verbal output, produced with difficulty, consists only of individual words rather than grammatical sentences. When asked to describe the weather in a sentence, a patient might say merely, “Weather … sunny” (Geschwind, 1970).

Later in the nineteenth century, a German physician and researcher, Carl Wernicke, observed patients who could not understand language (Geschwind, 1970). Their hearing was intact, but words made no sense to them. Like Broca, Wernicke examined his patients’ brains after their deaths. He found damage in an area different from the one identified by Broca: a region in the left temporal lobe, near the part of the brain that processes sound, which has come to be known as Wernicke’s area (Figure 8.5).

Because patients with damage in Wernicke’s area cannot understand language, they also cannot produce meaningful sentences. They are able to easily combine nouns and verbs according to rules of syntax, but their sentences are meaningless semantically. A patient might say, “I was over in the other one, and then after they had been in the department, I was in this one” (Geschwind, 1970, p. 941).

322

BRAIN-IMAGING EVIDENCE. These early studies of language and the brain told a simple story. The brain, it appeared, contained two specialized language devices: one that produced grammatical language (Broca’s area) and one that understood it (Wernicke’s area). Damage one area, and the corresponding language ability (production or understanding) is impaired.

However, brain damage is not the only relevant source of evidence. Another—unavailable in Broca’s and Wernicke’s day—is brain imaging. Brain images taken while people (primarily healthy volunteers with no brain damage) produce and listen to language reveals that the simple story about language and the brain was too simple (Bookheimer, 2002; Martin, 2003). The Broca-and-Wernicke story has, in three ways, “turned out to be wrong” (Fedorenko & Kanwisher, 2009, p. 2):

The original belief, that damage to Broca’s and Wernicke’s areas disrupts language, is correct. But the idea that these two brain regions are uniquely responsible for language is not. Broca’s area and Wernicke’s area are parts of the brain needed for normal language use, but they are not independently responsible for language.

So far, we’ve discussed communication among human beings. What about our furry friends?

Animals definitely can communicate. Bees, for example, make movements that signal the presence of food to other bees. But is this language? Recall that language is a system in which the language-user can generate novel communications by combining symbols that are arbitrarily related to the objects to which they refer. Can animals do this?

If you observe animals, it doesn’t look as if they are doing this. But maybe that’s because they haven’t been taught properly. The psychologist Herbert Terrace decided to find out what would happen if an animal received language training, as humans do (Hess, 2008).

Terrace raised a newly born chimpanzee, which he named Nim Chimpsky (after Noam Chomsky), in a human setting: a New York City apartment populated by humans. Terrace’s goal was to pit Chimpsky versus Chomsky. If Chimpsky learned language, he would disprove Chomsky’s contention that language stems from a uniquely human innate brain mechanism.

Terrace knew Chimpsky would never speak like a human; chimps lack the physical ability to produce human speech sounds. But they do have manual dexterity. Terrace therefore tried to teach Nim sign language. A sign language is a language in which bodily movements, especially of the hands and fingers, are used to convey information. Nim was tutored in sign language day after day, year after year.

What happened? After four years of instruction, Terrace concluded that Nim never developed true language use (Hess, 2008). He could imitate signs his teachers made. However, unlike a human child, Nim never generated novel combinations of signs to express ideas. Nim is not alone in the animal world; no animal has ever displayed true language use.

Human language abilities require a wide range of cognitive skills (Fitch, Hauser, & Chomsky, 2005; Pinker & Jackendoff, 2005), and Nim, and other animals, do possess some of them. For example, the skills of language acquisition include using symbols (words) to refer to objects in the world, remembering the meaning of the symbols, and combining them. Many animal species can perform some of these skills. A chimp named Washoe, for example, learned hand signals that symbolized “please” and “drink,” and combined them into “please drink” (Gardner & Gardner, 1969). An African Grey Parrot learned speech sounds that referred not only to objects, but also to categories of objects (Pepperberg, 1991).

324

Remarkable as these animal abilities are, however, they are not the same as human language (Hauser et al., 2002). Animals fail to display language skills that are common among humans. For example, animals know which members in their pack are the dominant ones, but their vocalizations don’t reflect this knowledge. They signal others with sounds but, when doing so, don’t take into account the knowledge and goals of the animals to whom they’re communicating, as humans do. Most important, animal communication is not generative; animals do not use rules of syntax to generate an indefinite variety of sentences, as even young human children do. “Nearly a century of intensive research” shows that “no species other than humans has a comparable capacity to recombine meaningful units into an unlimited variety of larger structures” (Hauser et al., 2002, p. 1576).