Language Module 8.3 PDF

module 8.3 Language After studying this module, you should be able to: Discuss attempts to teach language to nonhumans. Describe and evaluate possible explanations for how humans are specialized to learn language....

module 8.3 Language After studying this module, you should be able to: Discuss attempts to teach language to nonhumans. Describe and evaluate possible explanations for how humans are specialized to learn language. I want an apple. Distinguish between types of language I’d like an apple. impairment after brain damage. Describe how children develop language. Give me an apple Explain the role of eye movements in reading. May I have an apple? Every species of animal has ways of communicat- ing, but only human language has the property of productivity, the ability to combine words into new ▲ Figure 8.20 According to transformational grammar, we transform a given deep sentences that express an unlimited variety of ideas structure into any of several sentences with different surface structures. (Deacon, 1997). People constantly invent new sen- tences that no one has ever said before. the sentence “Never threaten someone with a chain saw.” The surface structure of You might ask, “How do you know that no one that sentence maps into two deep structures, as shown in ▼ Figure 8.21. has ever said that sentence before?” Well, of course, no one can be certain that a particular sentence is new, but we can be confident that many sentences are new (without specifying which ones) because Nonhuman Precursors to Language of the vast number of possible ways to rearrange Researcher Terrence Deacon once presented a talk about language to his 8-year- words. Imagine this exercise (but don’t really try it old’s elementary school class. One child asked whether other animals have their unless you have nothing else to do with your life): own languages. Deacon explained that other species communicate but without Pick a sentence of more than 10 words from any the productivity of human language. The child persisted, asking whether other book you choose. How long would you need to keep reading, in that book or any other, until you found the exact same sentence again? Deep Structure No. 1: Deep Structure No. 2: You are holding a chain Some deranged person is In short, we do not memorize all the sen- saw. Don’t threaten to holding a chain saw. Don’t tences we use. Instead, we learn rules for making use it to attack someone! threaten him! and understanding sentences. The famous linguist Noam Chomsky (1980) described those rules as a transformational grammar, a system for converting a deep structure into a surface structure. The deep structure is the underlying logic or meaning of a sentence. The surface structure is the sequence of words as they are actually spoken or written (see ▲ Figure 8.20). According to this theory, when- ever we speak, we transform the deep structure of the language into a surface structure. Two surface structures can resemble each other without representing the same deep structure, or they can represent the same deep structure without resembling each other. For example, “John is easy to please” has the same deep structure as “Pleasing John is easy” and “It is easy to please John.” These sen- ▲ Figure 8.21 The sentence “Never threaten someone with a chain saw” has one tences represent the same idea. In contrast, consider surface structure but two deep structures, corresponding to different meanings. 274 / chap ter 8 Co gnition and L anguage Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it. animals had at least a simple language with a few words and short sentences. How much do these gestures resemble lan- No, he replied, they don’t. guage? Washoe and other chimpanzees trained in Then another child asked, “Why not?” (Deacon, 1997, p. 12). Deacon this way used their symbols almost exclusively to paused. Why not, indeed? If language is so useful to humans, why haven’t other make requests, not to describe, and rarely in origi- species evolved at least a little of it? And what makes humans so good at learn- nal combinations (Pate & Rumbaugh, 1983; Terrace, ing language? Petitto, Sanders, & Bever, 1979; Thompson & Church, One way to examine humans’ language specialization is to ask how far an- 1980). By contrast, a human child with a vocabulary other species could progress toward language. Beginning in the 1920s, several of 100 words or so links them into short sentences and psychologists reared chimpanzees in their homes and tried to teach them to frequently uses words to describe. However, Washoe talk. The chimpanzees learned many human habits but understood only a few did show some understanding. She usually answered words and their few attempts to speak were extremely inarticulate. “Who” questions with names, “What” questions with Much of the problem is that chimpanzees make their sounds while inhal- objects, and “Where” questions with places, even ing, whereas humans speak while exhaling. (Give it a try. Can you say your when she specified the wrong name, object, or place name while inhaling?) However, chimpanzees do make hand gestures in na- (Van Cantfort, Gardner, & Gardner, 1989). ture. R. Allen Gardner and Beatrice Gardner (1969) taught a chimpanzee named More impressive results have been reported Washoe to use the sign language of the American deaf (Ameslan). Washoe even- for another species, the bonobo chimpanzee, Pan tually learned the symbols for about 100 words, and other chimps learned to paniscus. Bonobos’ social behavior resembles that communicate with other visual symbols (see ▼ Figure 8.22). of humans in several regards: Males and females © Ann Premack © Ann Premack a b © Ann Premack © Ann Premack c d ▲ Figure 8.22 Psychologists have tried to teach chimpanzees to communicate with gestures or symbols. (a) A chimp arranges plastic chips to request food. (b) Another chimp in her human home. (c) Kanzi, a bonobo, presses symbols to indicate words. (d) A chimp signing toothbrush. m o d u l e 8. 3 Language / 275 Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  ✓ concecpkt © Elizabeth Rupert, Language Research Center, Georgia State University 14. Based on the studies with bonobos, can you offer advice che about how to teach language to children with impaired language learning? Answer instead of providing direct reinforcements for correct responses. 14. Start language learning when a child is young. Rely on imitation as much as possible Human Specializations for Learning Language Humans are clearly more adapted for language than any other species, includ- ▲ Figure 8.23 Kanzi, a bonobo, points to answers ing bonobos. Why do we learn language so easily? on a board in response to questions he hears through earphones. Experimenter Rose Sevcik does not hear Language and General Intelligence the questions, so she cannot signal the correct answer. Did we evolve language as an accidental by-product of evolving big brains? Several observations argue strongly against this idea. Dolphins and whales have even larger brains but do not develop a flexible communication system resem- form strong attachments, females are sexually re- bling human language. Some people with brain damage have less total brain sponsive outside their fertile period, males contrib- mass than a chimpanzee but continue to speak and understand language. ute to infant care, and adults often share food. Like Also, some children, up to 7 percent by some estimates, have normal intel- humans, they stand comfortably on their hind legs, ligence in other ways but noticeable limitations in language. For example, they and they often copulate face-to-face. Several bono- don’t understand the difference between “Who was the girl pushing?” and “Who bos have learned to press keys on a board to make was pushing the girl?” (Leonard, 2007). People with a particular gene have even short sentences, as in Figure 8.22c and ▲ Figure 8.23. greater language impairments despite otherwise normal intelligence (Fisher, Unlike Washoe and other common chimpanzees, Vargha-Khadem, Watkins, Monaco, & Pembrey, 1998; Lai, Fisher, Hurst, Vargha- bonobos sometimes use symbols to describe events, Khadem, & Monaco, 2001). They do not fully master even simple rules, such as without requesting anything. One with a cut on his how to form plurals of nouns. hand explained that his mother had bitten him. At the opposite extreme, consider Williams syndrome, a genetic condition However, unlike children, bonobos almost never characterized by mental retardation in most regards but surprisingly good use of use expressions of possession, such as “mine!” (Lyn, language relative to their other abilities (Meyer-Lindenberg, Mervis, & Berman, Greenfield, Savage-Rumbaugh, Gillespie-Lynch, & 2006). One child, when asked to name as many animals as he could, started with Hopkins, 2011). “ibex, whale, bull, yak, zebra, puppy, kitten, tiger, koala, dragon...” Another The most proficient bonobos seem to compre- child could sing more than 1,000 songs in 22 languages (Bellugi & St. George, hend symbols about as well as a 2- to 21/2-year-old 2000). However, these children prefer 50 pennies to 5 dollars and, when asked child understands language (Savage-Rumbaugh et to estimate the length of a bus, give answers such as “3 inches or 100 inches, al., 1993). They also show considerable understand- maybe” (Bellugi, Lichtenberger, Jones, Lai, & St. George, 2000). They often show ing of spoken English, following such odd com- problems in attention and planning (Greer, Riby, Hamiliton, & Riby, 2013). mands as “bite your ball” and “take the vacuum Evidently, language ability is not the same as overall intelligence. cleaner outside” (Savage-Rumbaugh, 1990; Savage- Rumbaugh, Sevcik, Brakke, & Rumbaugh, 1992). Language Learning as a Specialized Capacity They passed a test of responding to commands is- Susan Carey (1978) calculated that children between the ages of 1½ and 6 learn an sued over earphones, to make sure the experimenter average of nine new words per day. But how do they infer the meanings of all those sitting nearby wasn’t using “Clever Hans”-type sig- words? A parent points at a frog and says “frog.” How does the child guess that nals, as discussed in Chapter 2 (see Figure 8.23). the word means “frog” rather than “small thing,” “green thing,” or “this particular Why have bonobos been more successful with frog”? Indeed, why does the child assume the sound means anything at all? language than common chimpanzees? Apparently, Noam Chomsky has argued that children must begin with preconceptions. bonobos have a greater predisposition for this type Chomsky and his followers suggest that people are born with a language acqui- of learning. Also, they learned by observation and sition device, a built-in mechanism for acquiring language (Pinker, 1994). One imitation, which promote better understanding line of evidence for this theory is that deaf children who are not taught a sign than the formal training methods that previous language invent one of their own and try to teach it to their parents or to other studies used (Savage-Rumbaugh et al., 1992). deaf children (Goldin-Meadow, McNeill, & Singleton, 1996; Goldin-Meadow & Finally, the bonobos began their language experi- Mylander, 1998). Further evidence is that children learn to use complex gram- ence early in life. matical structures, such as “Is the boy who is unhappy watching Mickey Mouse?” 276 / chap ter 8 Co gnition and L anguage Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Wernicke’s area: and muscle control. If you hear a highly emotional Brain damage leading to story, activity increases in brain areas important for Wernicke’s aphasia usually includes this area. emotion (Chow et al., 2013). That is, understanding language means relating it to everything else your brain does. Although Broca’s area, Wernicke’s area, and the surrounding areas are important, they can do nothing without connections to all of the brain. ✓ concecpkt 15. Brain-damaged patient A che speaks fluently but is hard to understand, and she has trouble understanding other people’s speech. Patient B understands most speech, but he speaks slowly and inarticulately, and he leaves out nearly all prepositions, conjunctions, and Broca’s area: Brain damage word endings. Which kind of aphasia does each leading to Broca’s patient have? aphasia usually includes this area. Answeraphasia. 15. Patient A has Wernicke’s aphasia. Patient B has Broca’s ▲ Figure 8.24 Brain damage that produces major deficits in language usually includes the left-hemisphere areas shown here. However, the deficits are severe only if the damage includes these areas but extends beyond them. Language Development even though they don’t hear that kind of expression very often. To pick up that Brain specializations facilitate language learning, but kind of grammar so quickly, children must have a predisposition of some sort. we still have to learn. Children’s language learning is Still, the exact nature of that predisposition is uncertain. amazing. Nearly every child learns language, even if the parents know nothing about how to teach it. Language and the Human Brain What aspect of the human brain enables us to learn language so easily? Studies Language in Early Childhood of people with brain damage have long pointed to two brain areas as particu- Table 8.2 lists the average ages at which children larly important for language. People with damage in the frontal cortex, including reach various stages of language ability (Lenneberg, Broca’s area (see ▲ Figure 8.24), develop Broca’s aphasia, a condition character- 1969; Moskowitz, 1978). Remember, these are aver- ized by difficulties in language production. Serious language impairment occurs ages, and children vary considerably. Progression only if the damage extends beyond Broca’s area, and even into the interior of the through these stages depends largely on maturation brain, but what matters here is the nature of the impairment, not the exact loca- (Lenneberg, 1967, 1969). Parents who expose their tion of the damage. The person speaks slowly and inarticulately and is no better children to as much language as possible increase with writing or typing. Someone with Broca’s aphasia is especially impaired with the children’s vocabulary, but they hardly affect the using and understanding grammatical devices such as prepositions, conjunc- rate of progression through language stages. Hearing tions, and word endings. For example, one person who was asked about a dental children of deaf parents are exposed to much less appointment slowly mumbled, “Yes... Monday... Dad and Dick... Wednesday spoken language, but they too progress almost on nine o’clock... 10 o’clock... doctors... and... teeth” (Geschwind, 1979, p. 186). schedule. People with damage in the temporal cortex, including Wernicke’s area (see Deaf infants babble as much as hearing infants Figure 8.24), develop Wernicke’s aphasia, a condition marked by impaired recall do for the first 6 months and then start to decline. At of nouns and impaired language comprehension, despite fluent and grammati- first, hearing infants babble only haphazard sounds, cal speech. Difficulty with nouns and impaired comprehension fit together: If but soon they start repeating the sounds they have you cannot remember what something is called, you will have trouble process- been hearing. By age 1 year, an infant babbles mostly ing a sentence based on that word. Because these people omit or misuse most sounds that resemble the language the family speaks nouns, their speech is hard to understand. For example, one patient responded (Locke, 1994). to a question about his health, “I felt worse because I can no longer keep in mind One of an infant’s first sounds is muh, and that from the mind of the minds to keep me from mind and up to the ear which can sound or something similar has been adopted by be to find among ourselves” (Brown, 1977, p. 29). many of the world’s languages to mean “mother.” However, language depends on far more than just Broca’s area and Infants also make the sounds duh, puh, and buh. Wernicke’s area. If you hear a story about sights and sounds, activity increases In many languages, the word for father is similar to in the brain areas responsible for vision and hearing. If you hear a story with dada or papa. Baba is the word for father in Chinese much movement, activity increases in the areas responsible for body sensations and for grandmother in several other languages. In m o d u l e 8. 3 Language / 277 Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it. effect, infants tell their parents what words to use for important concepts. Table 8.2 Stages of Language Development By age 1½, most toddlers have a vocabulary of about 50 words, but they seldom link words together. Age Typical Language Abilities (Much Individual Variation) A toddler says “Daddy” and “bye-bye” but not 3 months Random vocalizations. “Bye-bye, Daddy.” In context, parents can usually discern considerable meaning in these single-word 6 months More distinct babbling. utterances. Mama might mean, “That’s a picture of Mama,” “Take me to Mama,” “Mama went away 1 year Babbling that resembles the typical sounds of the family’s and left me here,” or “Mama, I’m hungry.” Toddlers language; probably one or more words including “mama”; language also communicate extensively by gestures (Behne, comprehension much better than production. Carpenter, & Tomasello, 2014; Kraljevic, Cepanec, & 1 1/2 years Can say some words (mean about 50), mostly nouns; few or no phrases. Simlesa, 2014). So do adults, of course. (Try to explain the concept of spiral without moving your hands!) 2 years Speaks in two-word phrases. Toddlers also sometimes combine a word with a gesture, such as pointing at something while say- 2 1/2 years Longer phrases and short sentences with some errors and unusual ing “mine” or pointing at a hat and saying “mama” constructions. Can understand much more. to indicate mama’s hat (Iverson & Goldin-Meadow, 3 years Vocabulary near 1,000 words; longer sentences with fewer errors. 2005). The word and gesture constitute a primitive kind of sentence. Children who convey much in- 4 years Close to adult speech competence. formation by gesture alone or word plus gesture at age 1½ are likely to develop better-than-average vocabulary and complex sentence structure by age 3½ (Rowe & Goldin-Meadow, 2009). Much of the evidence for this conclusion comes from people who learn a sec- By age 2, children start producing telegraphic ond language. Adults can memorize the vocabulary of a second language, but phrases of two or more words, such as “more page” children are far better at mastering the pronunciation and somewhat better at (read some more), “allgone sticky” (my hands are acquiring the grammar (Huang, 2014). People who start a second language after now clean), and “allgone outside” (someone has early childhood only rarely approach the level of a native speaker. Even those closed the door). Note the originality of such phrases. who start after the first couple years of life are at a disadvantage (Abrahamsson It is unlikely that the parents ever said “allgone sticky”! & Hyltenstam, 2009). However, researchers find no sharp age cutoff when lan- By age 2½ to 3 years, most children gener- guage suddenly becomes more difficult. The ease of learning a second language ate sentences but with some idiosyncrasies. Many declines steadily from early childhood through the 60s and 70s (Vanhove, 2013). young children have their own rules for negative sentences. A common one is to add no or not to the beginning or end of a sentence, such as, “No I want ✓ concecpkt 16. Why do psychologists believe that even very young to go to bed!” One little girl formed her negatives just by saying something louder and at a higher pitch. If che children learn rules of grammar? she shrieked, “I want to share my toys!” she meant, Answer creating such words as womans and goed. “I do not want to share my toys.” She had learned this 16. Children show that they learn rules of grammar when they overgeneralize those rules, rule by noting that people screamed when they told her not to do something. My son Sam made nega- tives for a while by adding the word either to the end of a sentence: “I want to eat lima beans either.” He Children Exposed to No Language had heard people say, “I don’t want to do that either.” or Two Languages When young children speak, they apply gram- Would children who were exposed to no language make up a new one? In rare matical rules, although of course they cannot state cases, an infant who was accidentally separated from other people grew up in those rules. For example, they apply the rules of a forest without human contact until discovered years later. Such children not English to produce such sentences as “the wom- only fail to show a language of their own but also fail to learn much language ans goed and doed something,” or “the mans getted after they are given the chance (Pinker, 1994). However, their development is their foots wet.” We say that children overregularize so abnormal and their early life so unknown that we should hesitate to draw or overgeneralize the rules. My son David invented conclusions. the word shis to mean “belonging to a female.” (He Better evidence comes from studies of children who are deaf. Children who apparently generalized the rule “He–his, she–shis.”) cannot hear well enough to learn speech and who are not taught sign language Clearly, children are not just repeating what they invent their own sign language (Senghas, Kita, & Özyürek, 2004). Observations have heard. in Nicaragua found that sign language evolved over the decades. Deaf peo- People have an optimal period for learning ple learned sign language and taught it to the next generation, who, having language in early childhood (Werker & Tees, 2005). learned it from early childhood, elaborated on it, made it more expressive, 278 / chap ter 8 Co gnition and L anguage Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it. taught the enhanced sign language to the next generation, and so on (Senghas & Coppola, 2001). If a deaf child starts to invent a sign language and no one responds to it, ✓ concecpkt 17. What are the advantages and because the child meets no other deaf children and the adults fail to learn, the che disadvantages of bilingualism? child gradually abandons it. If a deaf child does learn sign language, it can be a bridge to later learning a spoken language. On average, the children who are Answer language. best at sign language are also best at reading English or other written languages language as well as someone who is learning only one (Andrew, Hoshooley, & Joanisse, 2014). A deaf child who is given no opportunity languages than one, and probably will not master either to learn sign language until age 12 or so struggles to develop signing skills and Disadvantages: The individual takes longer to learn two never catches up with those who started earlier (Harley & Wang, 1997; Mayberry, and perhaps improves the ability to control attention. Lock, & Kazmi, 2002). This observation is our best evidence for the importance 17. Advantages: The individual can speak with more people, of early development in language learning: A child who doesn’t learn any lan- guage while young is permanently impaired at learning one. Many children grow up in a bilingual environment, learning two lan- Understanding Language guages. You might guess that a bilingual person represents the languages in The English language has many words with am- different brain areas. However, the research shows that both languages acti- biguous meanings. For instance, peck can mean vate the same areas (Perani & Abutalebi, 2005). Those who are bilingual from one-fourth of a bushel or to strike with a beak. Rose early infancy devote more brain areas to language, including parts of both can mean a flower or the past tense of the verb to left and right hemispheres, and they develop wider connections in the brain, rise. Desert can mean a dry stretch of land or (with but the same brain areas participate in both languages (Hull & Vaid, 2007; a different accent) to abandon someone. In context, Luk, Bialystock, Craik, & Grady, 2011; Mechelli et al., 2004; Perani & Abutalebi, however, listeners usually understand the meaning. 2005). If the brain represents two languages in the same places, how do We become even more aware of context when we bilingual people keep their languages separate? They don’t, at least not compare languages. Mandarin Chinese draws no dis- completely (Thierry & Wu, 2007). They often get confused when they switch tinctions for noun number or verb tense, and it lacks between languages (Levy, McVeigh, Marful, & Anderson, 2007; Linck, Kroll, & articles such as a and the. Thus, the sentence for “A man Sunderman, 2009). is buying an apple” is also the sentence for “The men Bilingualism has two disadvantages: Children take longer to master two bought apples,” “A man will buy apples,” and so forth. languages than one, and their vocabulary lags behind that of someone who Despite this ambiguity, listeners ordinarily understand speaks only one language. Bilingual people often take longer than average to the meaning in context. If the context is insufficient, a think of a word (Bialystok, Craik, & Luk, 2008). The primary advantage of bilin- speaker adds a word such as tomorrow or yesterday. gualism is obvious: People who know another language can communicate with The Malay language has one word for “you and I” more people (see ▼ Figure 8.25). A second advantage, as mentioned earlier, is and a different word for “someone else and I.” English that bilingual people learn to control their attention more effectively (Engel de translates both words as “we.” The Malaysians wonder Abreu et al., 2012; Gold, Kim, Johnson, Kryscio, & Smith, 2013). Researchers still how listeners understand this ambiguous word, just debate the size of that effect, however. as English speakers wonder how the Chinese get by without indications of number or tense. Understanding a Word Context not only determines how we interpret a word, but also primes us to hear an ambiguous sound one way or another. For example, a computer generated a sound halfway between a normal s sound and a normal sh sound. When this intermedi- ate sound replaced the s sound at the end of the word embarrass, people heard it as an s sound. When the same sound replaced sh at the end of abolish, people heard the same sound as sh (Samuel, 2001). We also use lip-reading more than we realize to understand what we hear. If lip movements do not match the sound, we strike a compromise be- tween what we see and what we hear (McGurk & Jim West/Alamy MacDonald, 1976). In one study, students listened to a tape recording of a sentence with a sound missing (Warren, 1970). The sentence was, “The state gover- ▲ Figure 8.25 Children who grow up in a bilingual or multilingual environment nors met with their respective legislatures convening gain in their ability to communicate with more people, and possibly gain in their in the capital city.” However, the sound of the first s in ability to control their attention. the word legislatures, along with part of the adjacent m o d u l e 8. 3 Language / 279 Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it. i and l, had been replaced by a cough or a tone. The students were asked to listen to the recording and try to identify the location of the cough or tone. None of the 20 students identified the location correctly, and half thought the cough or tone interrupted one of the other words on the tape. Even those who were told that the s sound was missing insisted that they clearly heard the sound s. The brain uses the context to fill in the missing sound. Just as we hear the word legislatures as a whole, not as a string of separate letters, we interpret a sequence of words as a whole, not one at a time. Suppose you hear a tape-recorded word that is care- fully engineered to sound halfway between dent and tent. The way you perceive it depends on the context: © James Kalat 1. When the *ent in the fender was well camou- flaged, we sold the car. 2. When the *ent in the forest was well camou- ▲ Figure 8.26 In England, a football coach is a bus full of soccer fans. In the United flaged, we began our hike. States, it’s the person who directs a team of American football players. Most people who hear sentence 1 report the word dent. Most who hear sentence 2 report tent. Now consider two more sentences: That store sells horseshoes. 3. When the *ent was noticed in the fender, we sold That store sells alligator shoes. the car. You interpret horseshoes to mean “shoes for horses to wear,” but you interpret 4. When the *ent was noticed in the forest, we alligator shoes as “shoes made from alligator hide.” Your understanding of the stopped to rest. sentences depends on your knowledge of the world, not just the syntax of the For sentences 3 and 4, the context comes too late to sentences. help. People are as likely to report hearing dent in Here is another example: one sentence as in the other (Connine, Blasko, & I’m going to buy a pet hamster at the store, if it’s open. Hall, 1991). Consider what this means: In the first I’m going to buy a pet hamster at the store, if it’s healthy. two sentences, after *ent, the person heard only two intervening syllables before hearing fender or Nothing about the sentence structure told you that it refers to the store in the forest. In the second pair, five syllables intervened. first sentence and a hamster in the second sentence. You understood because Evidently, when you hear an ambiguous sound, you you know that stores but not hamsters can be open, whereas hamsters but not hold it briefly in an “undecided” state for the context stores can be healthy. to clarify it. Beyond a certain point, it is too late. In short, understanding a sentence depends on your knowledge of the world Although a long-delayed context cannot help and all the assumptions that you share with the speaker or writer of the sen- you hear an ambiguous word correctly, it does help tence. Sometimes, you even have to remember where you are because the you understand its meaning. Consider the following meaning of a word differs from one place to another (see ▲ Figure 8.26). sentence (Lashley, 1951): Now consider this sentence: While Anna dressed the baby played in the crib. Quickly: Whom did Anna dress? And who played try Rapid righting with his uninjured hand saved from loss the contents of the capsized canoe. in the crib? The addition of a comma would simplify the sentence, but even without it, English grammar prohibits “baby” from be- your self it If you hear this sentence spoken aloud so that spell- ing both the object of dressed and the subject of played. If the baby ing provides no clues, you are likely at first to inter- played in the crib (as you no doubt answered), Anna must have dressed her- pret the second word as writing, until you reach the self. Nevertheless, many people think Anna dressed the baby (Ferreira, Bailey, & final two words of the sentence. Suddenly, capsized Ferraro, 2002). When speaking or writing, it is important to try to imagine ways canoe tells you that righting meant “pushing with a in which people might misunderstand you. paddle.” Only the immediate context can influence what you hear, but a delayed context can change the word’s meaning. Limits to Our Language Understanding Some grammatical sentences are almost incomprehensible. One example is a Understanding Sentences doubly embedded sentence—a sentence within a sentence within a sentence. A singly embedded sentence is understandable, though difficult: Making sense of language requires knowledge about the world. For example, consider the following sen- The dog the cat saw chased a squirrel. tences (from Just & Carpenter, 1987): The squirrel the dog chased climbed the tree. 280 / chap ter 8 Co gnition and L anguage Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  AR SUG not p su g cyan oison ta ar sod ide, Sucr ble iu No m ose t , Sucrose, Not sodium table cyanide, sugar not poison ▲ Figure 8.27 Most students preferred Kool-Aid made with sugar labeled “sugar” instead of sugar labeled “not cyanide,” even though they had placed the labels themselves. People don’t always trust the word not. (Based on results of Rozin, Markwith, & Ross, 1990) In the first sentence, “the cat saw the dog” is em- notice encourage you to buy the product? Hardly! I bedded within “the dog chased a squirrel.” In the was once on an airplane that turned around shortly second, “the dog chased the squirrel” is embedded after departure because one of its two engines failed. within “the squirrel climbed the tree.” So far, so good, The attendant told the passengers what was hap- but now consider a doubly embedded sentence: pening, but until she said, “Please don’t panic,” we didn’t realize there might be a reason to panic. If you The squirrel the dog the cat saw chased climbed ask someone for a favor and the person responds, the tree. “No problem,” how do you react? The expression “no Doubly embedded sentences overburden our mem- problem” implies there was almost a problem, or ory. In fact, if your memory is already burdened with maybe there is a bit of a problem. other matters, you may have trouble understanding In one clever experiment, students watched an a singly embedded sentence (Gordon, Hendrick, & experimenter pour sugar into two jars. The students Levine, 2002). were then told to label one jar “sucrose, table sugar” Double negatives are also difficult to under- and the other “not sodium cyanide, not poison.” stand. “I would not deny that...” means that I agree. Then the experimenter made two cups of Kool-Aid, “It is not false that...” means that something is true. one from each jar of sugar, and asked the students People often misunderstand such sentences. Have to choose one to drink (see ▲ Figure 8.27). Of the 44 you ever seen a multiple-choice test item that asks who expressed a preference, 35 wanted Kool-Aid “Which of the following is not true...” and then one made from the jar marked “sucrose,” not from the of the choices has a not in it? With such items, confu- one that denied having poison (Rozin, Markwith, & sion is almost certain. Ross, 1990). Triple negatives are still worse. Consider the following sentence, which includes four negatives (emphasis added): “If you do not unanimously find from your consideration of all the evidence that Reading there are no mitigating factors sufficient to preclude Students of language distinguish between phonemes the imposition of a death sentence, then you should and morphemes. A phoneme is a unit of sound, such sign the verdict requiring the court to impose a sen- as f or sh. Machines that talk to you, such as a GPS, tence other than death.” In Illinois some years ago, take a written word, break it into phonemes, and judges used to read those instructions to a jury to ex- pronounce the phonemes. For some words with ir- plain how to decide between a death penalty and life regular spellings, the machine’s pronunciation may in prison. It means that if even one juror sees some be wrong or hard to understand. A morpheme is a reason to reject the death penalty, the jury should unit of meaning. For example, the noun thrills has recommend prison instead. Do you think many ju- two morphemes (thrill and s). The final s is a unit of rors understood? meaning because it indicates that the noun is plural With a single negative, people often don’t fully (see ▼ Figure 8.28). Harp has one morpheme, and accept the meaning of the word not. Suppose a pack- harping has two, but harpoon has just one, as it is aged food says, “Contains no rat pieces!” Does that not derived from harp. Morphemes help us break an m o d u l e 8. 3 Language / 281 Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it. Phonemes (units of sound): Word Recognition Expertise develops from many years of practice, en- SHAMELESSNESS abling someone to recognize complex patterns at a glance. Because you have been reading for hours a day, Morphemes (units of meaning): almost every day since childhood, you have developed expertise at reading. You may not think of yourself as an ▲ Figure 8.28 The word shamelessness has nine phonemes (units of sound) and expert, because we usually reserve that term for some- three morphemes (units of meaning). one who is far more skilled than others. Nevertheless, you recognize words instantaneously, like an expert who recognizes chess patterns at a glance. Consider the following experiment: The inves- tigator flashes a letter on a screen for less than a quarter-second, shows an interfering pattern, and asks, “Was it C or J?” Then the experimenter flashes an entire word on the screen under the same KM Graphic/Shutterstock.com conditions and asks, “Was the first letter of the word C or J?” (see ▼ Figure 8.29). Which question is easier? Most people identify the letter more accurately when it is part of a word than when it is presented by itself (Reicher, 1969; Wheeler, 1970). This is known as the Japanese hiragana word-superiority effect. In further research, James Johnston and James Chinese characters McClelland (1974) briefly flashed words on the Illustration of Japanese hiragana writing and Chinese screen and asked students to identify one letter at a writing. marked position in each word (see ▼ Figure 8.30). On some trials, the experimenters told the students to try to see the whole word. On other trials, they showed unfamiliar word into meaningful parts. For exam- the students exactly where the critical letter would ap- ple, we can see reinvigoration as re-in-vigor-ation, pear on the screen and told them to focus on that spot meaning the process of increasing vigor again. and ignore the rest of the screen. Most students iden- Readers of English and other European lan- tified the critical letter more successfully when they guages are accustomed to the idea that a letter looked at the whole word than when they focused on or combination of letters represents a phoneme. just the letter itself. This benefit occurs only with a real However, in the Japanese hiragana style of writing, word, like COIN, not with a nonsense combination, each character represents a syllable. In Chinese, like CXQF (Rumelhart & McClelland, 1982). each character represents a morpheme and ordinar- You may have experienced the word-superiority ily a whole word. effect yourself. To pass time on long car trips, people sometimes try to find every letter of the alphabet on the billboards. It is easier to spot a letter by reading ✓ concecpkt 18. How many phonemes are in words than by checking letter by letter. What accounts for the word-superiority ef- che the word thoughtfully? How fect? According to one model (McClelland, 1988; many morphemes? Rumelhart, McClelland, & the PDP Research Group, 1986), our perceptions and memories are repre- Answer sented by connections among “units” correspond- morphemes: thought-ful-ly. 18. It has seven phonemes: th-ough-t-f-u-ll-y. It has three ing to sets of neurons. Each unit connects to other &!@ #)$#(}X )$#(}X ▲ Figure 8.29 Either a word or a single letter flashed on a screen and then an interfering pattern. The observers were asked, “Which was presented: C or J?” More of them identified the letter correctly when it was part of a word. 282 / chap ter 8 Co gnition and L anguage Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  1. The stimulus itself C O I N RIVER * SVLEJ 2. Letter detectors in a b the brain make tentative identifications. ▲ Figure 8.30 Students identified an indicated letter 3. Their input goes to a better when they focused on an entire word (a) than C? system that identifies a word. on a single letter in a designated spot (b). C-O-I-N “Yes! That’s a word!” O? Yes! Yes! units (see Figure 8.31). Any activated unit excites I? Yes! ▲ Yes! some of its neighbors and inhibits others. Suppose N? 4. It sends feedback to confirm units corresponding to the letters C, O, I, and N are the tentative letter identifications. moderately active. They excite a higher-order unit corresponding to the word COIN. Although none ▲ Figure 8.31 According to one model, a visual stimulus activates certain of the four letter units sends a strong message by letter units, some more strongly than others. Those letter units then activate itself, the collective impact is strong (McClelland & a word unit, which in turn strengthens the letter units that compose it. For Rumelhart, 1981). The perception COIN then feeds this reason, we recognize a whole word more easily than a single letter. excitation back to the individual letter-identifying units and confirms their tentative identifications. This model helps explain our perception of Reading and Eye Movements ▼ Figure 8.32. You see the first and third words as Reading requires eye movements, of course. When BIRD and PROOF, not BIPD and RROOF. You psychologists monitored eye movements, they dis- see the second and fourth words as DRIVE and covered that a reader’s eyes move in a jerky fashion. FRIDAY, not DRIVF and EIRDAY. Why? After all, the You move your eyes steadily to follow a moving ob- R in BIRD looks the same as the P in PROOF, and ject, but when scanning a stationary object, such as the E in DRIVE looks the same as the F in FRIDAY. a page of print, you alternate between fixations, when When you tentatively perceive a word, the feedback your eyes are stationary, and saccades (sa-KAHDS), strengthens the perception of the units that would quick eye movements from one fixation point to an- make the word, and not those that would make a other. You read during fixations, not during saccades. meaningless string of letters. Word recognition For an average adult reader, most fixations last about can become more complex. Consider the following 200 to 250 milliseconds. Fixations are briefer on fa- sentence: miliar words like legal than on harder words like The boy cuold not slove the porblem so he aksed luau or words with more than one meaning like lead for help. (Rodd, Gaskell, & Marslen-Wilson, 2002). Because Most readers “recognize” the words could, solve, BIP D problem, and asked, although of course they read faster if all words are spelled correctly (Rayner, White, Johnson, & Liversedge, 2006; White, Johnson, Liversedge, & Rayner, 2008). When we read, we pro- cess the context so that even out-of-place letters activate identification of the correct words. (This tendency can pose a probelm for prooofreaders, who sometimes fail to notice a misspellling!) DRIVF ✓ concecpkt 19. What evidence indicates that PROOF che FRIDAY we do not read a word one let- ter at a time? Answer of order. misspelled word even when certain letters are out different context. Also, a reader sometimes “recognizes” a ▲ Figure 8.32 The combination of possible letters enables us to identify a word. Word recognition in turn to be one letter in one context and another letter in a 19. Ambiguous letters, such as those in Figure 8.32, appear helps to confirm the letter identifications. m o d u l e 8. 3 Language / 283 Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it. saccades last only 25 to 50 ms, a normal reading pace If your reading span were limited by how many letters can fit into the fovea of is about four fixations per second (Rayner, 1998). your retina, you would read fewer letters as the print gets larger. In fact, you do at How much can someone read during a fixation? least as well, maybe even better, with larger print (up to a point). On average, the limit is about 11 characters at a time. The results vary from one language to another. In Japanese and Chinese, To demonstrate, focus on the letter i marked by an where each character conveys more information than English letters do, read- arrow (↓) in these sentences: ers see fewer letters per fixation (Rayner, 1998). That is, the limit depends on ↓ how much meaning one can attend to at once. In Hebrew and Farsi, which are written right to left, readers read more letters to the left of fixation and fewer 1. This is a sentence with no misspelled words. to the right (Brysbaert, Vitu, & Schroyens, 1996; Faust, Kravetz, & Babkoff, 1993; ↓ Malamed & Zaidel, 1993). 2. Xboc tx zjg rxunce with no mijvgab zucn. Reading is a strategic process of pausing longer on difficult or ambiguous words and sometimes looking back to previous words. In fact, of all eye move- If you permit your eyes to wander back and forth, you ments while reading, about 10 to 15 percent are backward movements. One notice the gibberish in sentence 2. But as long as you app you can get for your computer claims to increase reading speed by elimi- dutifully keep your eyes on the fixation point, the sen- nating those backward movements. A device monitors your eye movements, tence looks all right. You read the letter on which you so that whenever you move your eyes forward, every word you have already fixated plus about three or four characters (including read goes blank. You therefore cannot go back. That procedure may indeed spaces) to the left and about seven to the right. The increase your reading speed, but sometimes you need to look back. Consider rest is too blurry to make it out. Therefore, you see — this sentence: ce with no m—or possibly—nce