C H A P T E R 9
Language I: Introduction to Language and Language Comprehension
Introduction
The Nature of Language
A Caution: Psycholinguistics Is English-Centered Background on the Structure of Language
A Brief History of Psycholinguistics Factors Affecting Comprehension In Depth: Neurolinguistics
Basic Reading Processes
Comparing Written and Spoken Language Discovering the Meaning of Unfamiliar Words Reading and Working Memory
Two Pathways for Reading Words
Understanding Discourse
Forming an Integrated Representation of the Text Drawing Inferences During Reading
Teaching Metacomprehension Skills
Individual Differences: Test Anxiety and Reading Comprehension Artificial Intelligence and Reading
287
P
R E V I E WIn Chapters 9 and 10, we’ll examine how we process language. Specifically, Chap- ter 9 emphasizes language comprehension in the form of listening and reading. In contrast, Chapter 10 will emphasize language production (speaking and writing), as well as bilingualism—a topic that encompasses both language comprehension and language production.
We’ll begin Chapter 9 by exploring the nature of language. In particular, we’ll look at the structure of language, a brief history of psycholinguistics, several factors that influence comprehension, and some neuroscience research on language.
Next, we’ll examine basic reading processes, beginning with a comparison of written and spoken language. Context is important when we need to understand the meaning of an unfamiliar word, and working memory also plays an important role in understanding sentences. This section also examines word recognition and implica- tions for teaching reading.
The last part of Chapter 9 moves beyond small linguistic units to consider dis- course, or language units that are larger than a sentence. Some important components of discourse comprehension include forming an integrated representation of a passage, drawing inferences that were not actually stated in the passage, teaching metacompre- hension skills, and the relationship between test anxiety and reading comprehension.
In addition, researchers in the field of artificial intelligence have designed programs that attempt to comprehend written language.
INTRODUCTION
Try to imagine a world without language. In fact, think how your life would change if you woke up tomorrow and language were forbidden. Even nonverbal gestures would be prohibited, because they are an alternate form of language (Beattie & Shovelton, 2003). Telephones, televisions, radios, newspapers, and books would all be useless.
Almost all college courses would disappear. You wouldn’t even be able to talk to your- self, so it would be impossible to reminisce, remind yourself about a task you must complete, or make plans for the future. Furthermore, our interpersonal and societal interactions would collapse.
Like many cognitive skills, language rarely receives the credit it deserves. After all, you simply listen to someone with a moving mouth and vocal equipment, and you understand the message this person is trying to convey. Equally effortlessly, you open your own mouth and sentences emerge almost instantaneously—an impressive testi- mony to the efficiency of our cognitive processes (Theme 2).
Another equally impressive characteristic of our language skills is our extraordinary ability to master thousands of words. In Chapter 8, we noted that the average North American has an estimated vocabulary of more than 20,000 to 40,000 words in his or
Introduction 289 her vocabulary (Baddeley, 1990; Saffran & Schwartz, 2003). Furthermore, the average
college-educated North American has a speaking vocabulary of at least 75,000 words (Bock & Garnsey, 1998; Wingfield, 1993).
Cognitive psychologists emphasize that human language is probably one of the most complex processes to be found anywhere on our planet (Gleitman & Liberman, 1995). The domain of language includes an impressive diversity of skills. Consider just a few of the abilities that you need in order to understand a spoken sentence: encoding the sound of a speaker’s voice, encoding the visual features of printed language, access- ing the meaning of words, understanding the rules that determine word order, and appreciating from a speaker’s intonation whether a sentence is a question or a state- ment. Furthermore, you manage to accomplish all these tasks while listening to a speaker who is probably producing three words eachsecond (Levelt et al., 1999; Vigliocco &
Hartsuiker, 2002). In fact, talking is so difficult that it should be an Olympic event—
except that most humans have mastered this athletic skill (Bock & Garnsey, 1998).
In addition, the productivity of language is unlimited. For example, consider only the number of 20-word sentences that you could potentially generate. You would need 10,000,000,000,000 years—or 2,000 times the age of the earth—to say them all (Miller, 1967; Pinker, 1993).
In Chapters 9 and 10, we will discuss psycholinguistics,an interdisciplinary field that examines how people use language to communicate ideas (Corballis, 2006). We use language in thousands of different settings, from courtrooms to comic strips. Further- more, language provides an excellent example of Theme 4 of this textbook, the inter- relatedness of the cognitive processes. In fact, virtually every topic we have discussed so far in this book makes some contribution to language processing. To illustrate this point, try Demonstration 9.1.
Demonstration 9.1
How Other Cognitive Processes Contribute to Language Look below at the list of chapters you have read so far. For each chapter, list at least one topic that is connected to language. The answers appear at the end of this chapter, on page 322.
Chapter 2: Perceptual Processes I: Visual and Auditory Recognition Chapter 3: Perceptual Processes II: Attention and Consciousness Chapter 4: Working Memory
Chapter 5: Long-Term Memory
Chapter 6: Memory Strategies and Metacognition Chapter 7: Mental Imagery and Cognitive Maps Chapter 8: General Knowledge
The two chapters on language should also convince you that humans are active information processors (Theme 1). Rather than passively listening to language, we actively consult our previous knowledge, use various strategies, create expectations, and draw conclusions. When we speak, we must determine what our listeners already know and what other information must be conveyed. Language is not only our most remark- able cognitive achievement, but it is also the most social of our cognitive processes.
The first of our two chapters on language focuses on language comprehension.
After an introductory discussion about the nature of language, we will examine basic reading, as well as the more complex process of understanding discourse. In Chap- ter 10, we will switch our focus from understanding to the production of language.
Chapter 10 considers two production tasks: speaking and writing. With a background in both language comprehension and language production, we can then consider bilingualism. Bilinguals—certainly the winners in any Olympic language contest—
manage to communicate easily in more than one language.
THE NATURE OF LANGUAGE
Psycholinguists have developed a specialized vocabulary for language terms; let’s now consider these terms. A phoneme(pronounced “foe-neem”) is the basic unit of spoken language, such as the soundsa, k,andth.The English language has about 40 phonemes (Groome, 1999; Mayer, 2004).
In contrast, a morpheme(pronounced “more-feem”) is the basic unit of meaning.
For example, the wordreactivatedactually contains four morphemes:re-,active,-ate,and -ed.Each of those segments conveys meaning. Many morphemes can stand on their own (likegiraffe). In contrast, some morphemes must be attached to other morphemes in order to convey their meaning. For instance,re-indicates a repeated action. As you might guess, the term morphologyrefers to the study of morphemes.
Another major component of psycholinguistics is syntax. Syntaxrefers to the grammatical rules that govern how we organize words into sentences (Owens, 2001).
A more inclusive and familiar term, grammar,encompasses both morphology and syntax; it therefore examines both word structure and sentence structure (Evans &
Green, 2006).
Semanticsis the area of psycholinguistics that examines the meanings of words and sentences (Carroll, 2004). The related term, semantic memory,refers to our organized knowledge about the world. We have discussed semantic memory through- out earlier chapters of this book, but especially in Chapter 8.
An additional important term is pragmatics,which is our knowledge of the social rules that underlie language use; pragmatics takes into account the listener’s perspec- tive (Carroll, 2004; Tomasello, 2003). For example, think how you would define the wordsyntaxto a 12-year-old child, as opposed to a college classmate. Pragmatics is an especially important topic when we consider the production of language (Chapter 10), but pragmatic factors also influence comprehension.
As you can see from reviewing the terms in this section, psycholinguistics encompasses a broad range of topics, including sounds, several levels of meaning, grammar, and social
The Nature of Language 291 factors. We begin by noting a problem with the current research in psycholinguistics.
Then we’ll consider additional aspects of the nature of language: some background about the structure of language, a brief history of psycholinguistics, factors affecting compre- hension, and neurolinguistics.
A Caution: Psycholinguistics Is English-Centered
In an important article, Elizabeth Bates and her coauthors (2001) emphasize a bias that operates in research about psycholinguistics. Specifically, most researchers in this discipline focus on how people understand and use English. As a result, some of the findings may apply only to English speakers, rather than to all humans. Current lin- guists estimate that 6,000 to 7,000 languages are spoken throughout the world, so the emphasis on just one of these languages is especially unfortunate (Fishman, 2006; Ku, 2006; Tomasello, 2003).
If your own first language is English, your ideas about language are probably English-centered. Therefore, you may react with surprise if you travel to another language community. For instance, I recall visiting Grenada, Spain, and hearing the tour guide (who appeared to be Spanish) describing the sights in Spanish and in Japanese to the tourists in her group. I’m embarrassed to report that I was startled to hear her translating Spanish into Japanese, without first passing through English.
Bates and her colleagues (2001) point out some differences among languages. For example, in English, word meaning does not depend on the relative pitch of the syl- lables in a word. In contrast, in Mandarin Chinese,mameans “mother” when the word is spoken at a single pitch. However,mameans “horse” when spoken in a tone that initially falls and then rises (Field, 2004).
Furthermore, Sesotho—a language spoken in southern Africa—uses the passive voice more than English does. In many languages, the nouns have a grammatical gen- der, although English nouns do not. Brain processing even differs as a function of a per- son’s language. A region in the frontal lobe is activated when English speakers listen to certain complex sentences; that region does not respond when German speakers listen to the translated versions of those sentences (Bornkessel & Schlesewsky, 2006).
In summary, languages differ widely from one another on numerous dimensions (Tomasello, 2003). Clearly, psycholinguists will need to conduct extensive research in many other languages if they want to determine which linguistic principles apply universally.
Background on the Structure of Language
Before we consider the history of psycholinguistics, we need to discuss a central con- cept in understanding language, called phrase structure. Phrase structureemphasizes that we construct a sentence by using a hierarchical structure that is based on gram- matical building blocks called constituents(Carroll, 2004). For example, suppose we have the following sentence:
The young woman carried the heavy painting.
We can divide this sentence into two broad constituents: (1) the phrase that focuses on the noun—“the young woman”—and (2) the phrase that focuses on the verb—“carried the heavy painting.” Each of these constituents can be further subdivided, creating a hierarchy of constituents with a diagram resembling an upside-down tree. These dia- grams, like the one in Figure 9.1, help us appreciate that a sentence is not simply a chain of words, strung together like beads on a necklace. Instead, we appreciate more complicated relationships among the elements of a sentence (Owens, 2001).
We often need the entire constituent to give us cues about the meaning of the words. For example, consider the wordpaintingin the sentence we just analyzed.Paint- ingcould be either a verb or a noun. However, from the context in whichpainting appears in the constituentthe heavy painting,we know that the noun version is appro- priate. The research indicates that people maintain a complete constituent in working memory while they process its meaning (Jarvella, 1971).
A Brief History of Psycholinguistics
Now that we have identified some central concepts in psycholinguistics, let’s consider some highlights in the history of this field. Early philosophers in Greece and India debated the nature of language (Chomsky, 2000). Centuries later, both Wilhelm Wundt and William James also speculated about our impressive abilities in this area (Carroll, 2004; Levelt, 1998). However, the current discipline of psycholinguistics can be traced to the 1960s, when psycholinguists began to test whether psychological research would support the theories of Noam Chomsky (McKoon & Ratcliff, 1998). Let’s consider Chomsky’s theory, the reactions to his theory, and subsequent theories about language.
FIGURE 9.1
An Example of Constituents.
The young woman carried the heavy painting
The young woman carried the heavy painting
the young woman the heavy painting
heavy painting
heavy painting the
carried
young woman
The Nature of Language 293
Chomsky’s Approach. People usually think of a sentence as an orderly sequence of words that are lined up in a row on a sheet of paper. Noam Chomsky (1957) caused great excitement among psychologists and linguists by proposing that there is more to a sentence than meets the eye (or the ear). Chapter 1 of this textbook identified Chomsky’s work on the psychology of language as one of the forces that led to the decline of behaviorism. The behaviorists emphasized the observable aspects of lan- guage behavior (Field, 2004). In contrast, Chomsky argued that human language abil- ities could be explained only in terms of a complex system of rules and principles represented in the minds of speakers (Chomsky, 2006). Chomsky is clearly one of the most influential theorists in modern linguistics (N. Smith, 2000).
Chomsky proposed that humans have innate language skills. That is, we have an inborn understanding of the abstract principles of language. As a result, children do not need to learn the basic, generalizable concepts that are universal to all languages (Chomsky, 2003, 2006; Field, 2004).
Of course, children need to learn many superficial characteristics of the language spoken in their community. For instance, children in Spanish-speaking communities will need to learn the difference betweenserandestar.Spanish linguistic space is carved up somewhat differently from that of English, where children learn only one form of the verbto be. Still, Chomsky argues that all children have a substantial, inborn language ability. This ability allows them to produce and understand sentences they have never heard before (Belletti & Rizzi, 2002; Chomsky, 2006).
Chomsky (1975) also proposed that language is modular;people have a set of spe- cific linguistic abilities that do not follow the principles of other cognitive processes, such as memory and decision making (Nusbaum & Small, 2006). We discussed a related concept, the phonetic module, in connection with speech perception on page 60.
Because language is modular, Chomsky (2002, 2006) argues, young children learn com- plex linguistic structures many years before they master other, simpler tasks, such as mental arithmetic.
In contrast to Chomsky’s theory, the standard cognitive approach argues that lan- guage is not modular. Instead, it is interconnected with other cognitive processes such as working memory. According to this alternative approach, we are skilled at language because our powerful brains can master many cognitive tasks. Language is just one of those tasks, having the same status as tasks such as memory and problem solving (Bates, 2000; Carroll, 2004; Tomasello, 2003).
In addition, Chomsky (1957, 2006) also pointed out the difference between the deep structure and the surface structure of a sentence. The surface structureis represented by the words that are actually spoken or written. In contrast, the deep structureis the underlying, more abstract meaning of a sentence (Garnham, 2005).
Two sentences may have very different surface structures, but very similar deep structures. Consider these two sentences: (1) “Sara threw the ball” and (2) “The ball was thrown by Sara.” Notice that the two surface structures are different. None of the words occupies the same position in the two sentences, and three of the words in the second sentence do not even appear in the first sentence. However, “deep down,” speakers of English feel that the sentences have identical core meanings (Harley, 2001).
Chomsky (1957, 2006) also pointed out that two sentences may have identical sur- face structures but very different deep structures; these are called ambiguous sentences.
For example, I live outside the small town of York in rural upstate New York. One day I drove past the announcement board outside the York Town Hall, and the message said:
“POP CAN DRIVE.” I was puzzled: Whose father is now allowed to drive, and why had he previously been prohibited from driving? To be honest, the alternate meaning (focus- ing on a community fundraiser) did not occur to me until the next day.
We will discuss ambiguity in more detail later in the chapter. However, context usually helps us resolve these ambiguities. Here are three additional ambiguous sen- tences, each of which has two meanings:
The shooting of the hunters was terrible.
They are cooking apples.
The lamb is too hot to eat.
Reactions to Chomsky’s Theory. Initially, psychologists responded enthusiasti- cally to Chomsky’s ideas about grammar (Bock et al., 1992; Williams, 2005). Not all the evidence for Chomsky’s theory was favorable, however. For example, the research failed to support Chomsky’s prediction that people would take longer to process sentences that required numerous transformations (Carroll, 2004; Herriot, 2003).
Furthermore, some of Chomsky’s theories have not been tested (Agassi, 1997).
Chomsky’s later theories have provided more sophisticated linguistic analyses. For example, Chomsky has placed constraints on the possible hypotheses that the language learner can make about the structure of language (Chomsky, 1981, 2000; Harley, 2001).
Chomsky’s newer approach also emphasizes the information contained in the individual words of a sentence. For example, the worddiscussnot only conveys information about the word’s meaning, but it also specifies the requirement thatdiscussmust be followed by a noun, as in the sentence, “Rita discussed the novel” (Ratner & Gleason, 1993).
Psycholinguistic Theories Emphasizing Meaning. Beginning in the 1970s, many psychologists became discouraged with Chomsky’s emphasis on the grammatical aspects of language (Herriot, 2003). These psychologists began to develop theories that empha- sized the human mind, rather than the structure of language (Tanenhaus, 2004; Treiman et al., 2003). They also emphasized semantics, or themeaningof language. In recent years, this focus on semantics has encouraged psychologists to explore how people understand the meaning of paragraphs and stories. We’ll look at some of that research later in this chapter.
Several psychologists have developed theories that emphasize meaning (e.g., Kintsch, 1998; Newmeyer, 1998; Tomasello, 2003). Here, we will briefly describe one representative theory, the cognitive-functional approach to language. The cognitive- functional approach(also called usage-based linguistics) emphasizes that the func- tion of human language in everyday life is to communicate meaning to other individuals.
As the name suggests, the cognitive-functional approach also emphasizes that our cog- nitive processes—such as attention and memory— are intertwined with our language comprehension and language production.
The Nature of Language 295
Michael Tomasello (2003) points out that young children have extremely power- ful cognitive skills and social-learning skills. During the years when they are mastering language, they will hear several million adult sentences. As we’ll see in Chapter 13, children analyze these sentences, and they use flexible strategies to create increasingly complex language (Kuhl, 2006).
Tomasello (1998a, 1998b) also emphasizes that adults use language strategically.
Specifically, we structure our language in order to focus our listeners’ attention on the information we wish to emphasize. For instance, look at Demonstration 9.2, which illustrates a concrete example of the cognitive-functional approach (Tomasello, 1998a).
Notice how each of those sentences emphasizes a somewhat different perspective on the same event. You’ll probably find that these different perspectives are reflected in the vari- ety of questions you generated. In short, the cognitive-functional approach argues that people can use language creatively, in order to communicate subtle shades of meaning.
We’ll explore the social use of language more thoroughly in Chapter 10.
Factors Affecting Comprehension
Beginning in the 1960s, psychologists began to examine a variety of linguistic factors related to language comprehension. In general, people have difficulty understanding sen- tences in these four conditions:
1. If they contain negatives, such asnot.
2. If they are in the passive rather than the active voice.
Demonstration 9.2
The Cognitive-Functional Approach to Language
Imagine that you recently saw an event in which a man named Fred broke a win- dow, using a rock. A person who was not present at the time asks you for infor- mation about the event. For each of the sentences below, construct a question that this person might have asked that would prompt you to reply with that spe- cific wording for the sentence. For example, the brief response, “Fred broke the window” might have been prompted by the question, “What did Fred do?”
1. Fred broke the window with a rock.
2. The rock broke the window.
3. The window got broken.
4. It was Fred who broke the window.
5. It was the window that Fred broke.
6. What Fred did was to break the window.
Source:Based on Tomasello, 1998a, p. 483.
3. If they contain nested structures, with a descriptive clause in the middle of the sentence.
4. If they are ambiguous.
Negatives. A sentence in a newspaper column reads. “Georgia rejected a challenge to a referendum that had barred same-sex unions.” This sentence requires several read- ings to understand the basic message: Will the state of Georgia prohibit same-sex unions? The research on negatives is clear-cut. If a sentence contains a negative word, such asnoornot,or an implied negative (such asrejected), the sentence almost always requires more processing time than a similar, affirmative sentence (Williams, 2005).
In a classic study, Clark and Chase (1972) showed a picture of a star above a plus sign. Then they asked people to verify statements, such as the following:
Star is above plus. *+
The participants responded quickly in this case, when the sentence was affirmative.
They responded more slowly if the sentence contained the negative formisn’t(for example, “Plus isn’t above star”). The participants also made fewer errors with affir- mative sentences than with negative sentences. Notice that these results are consistent with Theme 3 of this textbook: Our cognitive processes handle positive information bet- ter than negative information.
As you can imagine, readers’ understanding decreases as the number of negative terms increases. For example, people perform only slightly better than chance when they judge sentences such as, “Few people strongly deny that the world is not flat” (Sherman, 1976, p. 145). These findings have clear-cut practical applications in numerous areas, such as education, advertising, and creating political surveys (Kifner, 1994).
The Passive Voice. As we discussed earlier, Chomsky (1957, 1965) pointed out that the active and passive forms of a sentence may differ in their surface structure, even though they have similar deep structures. However, the active form is more basic; we need to add extra words to create the passive form of a sentence.
The active form is also easier to understand (Garnham, 2005; Williams, 2005). For example, Ferreira and her coauthors (2002) asked participants to determine whether each sentence in a series was plausible or likely. The participants were highly accurate in responding “No” to sentences in the active voice, such as, “The man bit the dog.” In contrast, their accuracy dropped to about 75% when the same sentences were con- verted to the passive voice, for example, “The dog was bitten by the man” (p. 13).
Most current writing-style manuals recommend using the active voice. For exam- ple, the current manual of the American Psychological Association (2001) points out that an active-voice sentence such as “Nuñez (2006) designed the experiment,” is much more direct and vigorous than “The experiment was designed by Nuñez (2006).”
Nested Structures. A nested structureis a phrase that is embedded within another sentence. For example, we can take the simple sentence, “The plane leaves at 9:41,” and insert the nested structure “that I want to take.” We create a more structurally complex
The Nature of Language 297 sentence: “The plane that I want to take leaves at 9:41.” However, readers often
experience a memory overload when they try to read a sentence that contains a nested structure (Gibson, 1998, 1999; Rayner & Clifton, 2002). You need to remem- ber the first part of the sentence, “the plane,” while you process the nested structure.
Afterward, you can process the remainder of the sentence. The memory overload becomes excessive when the sentence contains several nested structures. For exam- ple, you might find yourself stranded when you try to understand the following sentence:
The plane that I want to take when I go to Denver after he returns from Washington leaves at 9:41 in the morning.
The next time you write a paper, remember how these three factors can influence comprehension. Whenever possible, follow these guidelines: (1) Use linguistically pos- itive sentences, rather than negative ones; (2) use active sentences, rather than passive ones; and (3) use simple sentences, rather than nested structures.
Ambiguity. Suppose that you saw the following headline in your local newspa- per: “Bombing Rocks Hope for Peace.” As you might imagine, sentences are diffi- cult to understand if they contain an ambiguous word or an ambiguous sentence structure. Recall that we mentioned ambiguous sentences in connection with Chomsky’s approach to language. Now let’s consider how people manage to understand these sentences.
Psychologists have designed several methods of measuring the difficulty of under- standing a sentence with an ambiguous word (MacDonald, 1999; Rodd et al., 2002).
For example, one method measures the amount of time that the reader pauses on a word before moving his or her eyes to the next words in the sentence (Pexman et al., 2004; Rayner et al., 2005). People typically pause longer when they are processing an ambiguous word.
Psychologists have proposed many theories to explain how listeners process an ambiguous word (Rayner & Clifton, 2002; Van Orden & Kloos, 2005). Current research supports the following explanation: When people encounter a potential ambiguity, the activation builds up for all the well-known meanings of the ambigu- ous item. Furthermore, people are likely to choose one particular meaning (1) if that meaning is more common than the alternate meaning and (2) if the rest of the sen- tence is consistent with that meaning (Morris & Binder, 2001; Rayner & Clifton, 2002; Sereno et al., 2003).
Consider this potentially ambiguous sentence: “Pat took the money to the bank.”
Here, the “financial institution” interpretation ofbankwould receive the most activa- tion. After all, this is the most common interpretation ofbank,and the context ofmoney also suggests this meaning. Some minimal activation also builds up for other meanings ofbank(as inriverbankandblood bank). However, just a fraction of a second later, these alternative meanings are suppressed, and they are no longer active (Gernsbacher et al., 2001; Rayner & Clifton, 2002). This explanation of ambiguity would be consistent with the parallel distributed processing approach.
So far, we have considered ambiguous words. However, sometimes a sentence struc- ture is ambiguous, especially if it contains no punctuation (Rayner et al., 2003). Try reading this sentence:
1. “After the Martians invaded the town that the city bordered was evacuated.”
(Tabor & Hutchins, 2004, p. 432).
Did you find yourself reading along quickly, and then you were suddenly lost? You had wandered down the wrong path. An ambiguous sentence is especially difficult if you read a long string of words that seem consistent with your initial interpretation. In con- trast, you can correct your initial mistake more quickly with a shorter string of words.
If sentence 1 is still unclear, see if you can understand this shorter sentence:
2. “After the Martians invaded the town was evacuated.” (Tabor & Hutchins, 2004, p. 432).
Demonstration 9.3
Searching for Ambiguous Language
Ambiguity occurs quite often in the English language (Rodd et al., 2002). Per- haps the best source of ambiguous words and phrases is newspaper headlines.
After all, these headlines must be very brief, so they often omit the auxiliary words that could resolve the ambiguity. Here are some actual newspaper head- lines that colleagues, students, and I have seen:
1. “Eye drops off shelf”
2. “Squad helps dog bite victims”
3. “British left waffles on Falkland Islands”
4. “Stolen painting found by tree”
5. “Clinton wins budget; more lies ahead”
6. “Miners refuse to work after death”
7. “Kids make nutritious snacks”
8. “Local high school dropouts cut in half”
9. “Iraqi head seeks arms”
10. “Oklahoma is among places where tongues are disappearing”
For the next few weeks, search the newspapers you normally read, looking for ambiguous headlines. Try to notice whether your first interpretation of the ambiguous portion was a correct or incorrect understanding of the phrase. If you find any particularly intriguing ambiguities, please send them to me! My address is: Department of Psychology, SUNY Geneseo, Geneseo, NY 14454.
The Nature of Language 299 As Rueckl (1995) observes, “Ambiguity is a fact of life. Happily, the human cogni-
tive system is well-equipped to deal with it” (p. 501). Indeed, we can understand ambigu- ous sentences, just as we can understand negative sentences, sentences using the passive voice, and sentences with complex nesting. However, we typically respond more quickly and more accurately when the language we encounter is straightforward. Now that you are familiar with the concept of ambiguity, try Demonstration 9.3.
Neurolinguistics
Neurolinguisticsis the discipline that examines how the brain processes language (Treiman et al., 2003). Research in this area has become increasingly active in recent years, and it demonstrates that the neurological basis of language is impressively com- plex. Let’s consider three topics: aphasia, hemispheric specialization in language pro- cessing, and neuroimaging research with normal individuals.
Individuals with Aphasia. The initial investigations in neurolinguistics began in the 1800s, when early researchers studied individuals who had language disorders.
In fact, before the early 1970s, almost all the information that scientists had acquired about neurolinguistics was based on people with aphasia. A person with aphasiahas difficulty communicating, caused by damage to the speech areas of the brain. This damage is typically caused by a stroke, a tumor, or a serious infection (Saffran &
Schwartz, 2003). Figure 9.2 illustrates two especially relevant regions of the brain.
Broca’s areais located toward the front of the brain. Damage to Broca’s area typically leads to speech that is hesitant, effortful, and grammatically simple (Dick et al., 2001; Gazzaniga et al., 2002). For example, one person with Broca’s aphasia tried to describe the circumstances of his stroke:
Alright . . . Uh . . . stroke and uh . . . I . . . huh tawanna guy . . . h . . . h . . . hot tub and . . . And the . . . two days when uh . . . Hos . . . uh . . . huh hospital and uh . . . amet . . . am . . . ambulance. (Dick et al., 2001, p. 760)
Broca’s aphasiais primarily characterized by an expressive-language deficit—
or trouble producing language. These symptoms make sense. Broca’s region is one of the locations of the brain that manages motor movement; to produce speech, you must move your lips and tongue. However, people with Broca’s aphasia may also have some trouble understanding language (Dick et al., 2001; Martin & Wu, 2005).
For example, they may be unable to tell the difference between “He showed her baby the pictures” and “He showed her the baby pictures” (Jackendoff, 1994, p. 149).
Wernicke’s area is located toward the back of the brain. (“Wernicke” is pro- nounced either “Ver-nih-kee” or “Wer-nih-kee.”) Damage to Wernicke’s area typically produces serious difficulties in understanding speech, as well as language production that is too wordy and confused (Harley, 2001). People with Wernicke’s aphasiahave
such severe receptive-language problems that they cannot understand basic instruc- tions like “Point to the telephone” or “Show me the picture of the watch.” In contrast to the quotation above, here’s how a person with Wernicke’s aphasia tried to describe the circumstances of his stroke:
It just suddenly had a feffert and all the feffort had gone with it. It even stepped my horn. They took them from earth you know. They make my favorite nine to severed and now I’m a been habed by the uh stam of fortment of my annulment which is now forever. (Dick et al., 2001, p. 761)
The basic information about Broca’s aphasia and Wernicke’s aphasia has been known for about a century. However, the distinction between these two aphasias is much less clear-cut than researchers had once believed (Gazzaniga et al., 2002;
Martin & Wu, 2005).
Another problem is related to an issue we emphasized earlier in the chapter, on page 291: We should not rely exclusively on English-based psycholinguistic research (Bates et al., 2001). For instance, researchers who examined aphasia in English speak- ers concluded that individuals with Broca’s aphasia made many grammatical errors.
However, these same researchers concluded that individuals with Wernicke’s aphasia typically produced grammatically correct sentences.
Broca’s area Wernicke’s area FIGURE 9.2
Broca’s Area and Wernicke’s Area:Two Regions of the Brain That Are Commonly Associated with Aphasia.
The Nature of Language 301 Here is the complication: English speakers use the same grammatical form for a
noun, whether it is the subject or the object of a sentence. In contrast, languages such as German and Czech add letters to the end of a noun if it is the object of a sentence, rather than the subject. Interestingly, people with Wernicke’s aphasia who speak lan- guages like German and Czech often fail to add the appropriate endings to the nouns.
In short, we see thatbothkinds of aphasia can decrease a person’s grammatical accuracy—
once we examine some languages other than English (Dick et al., 2001).
Hemispheric Specialization. We noted at the beginning of this In Depth section that the early researchers examined people with aphasia. These scientists also noticed that individuals with speech disorders typically had more severe damage in the left hemisphere of the brain, rather than the right hemisphere. During the mid-1900s, researchers began a more systematic study of lateralization. Lateralizationmeans that each hemisphere of the brain has somewhat different functions.
If you’ve read about lateralization in a popular magazine—rather than an aca- demic journal—you may have seen something like, “Language is localized in the left hemisphere of the brain.” However, this statement is too strong. Yes, most neuro- linguistic studies find greater activation in the left hemisphere than in the right (Bates, 2000; Grodzinsky, 2000; Scott, 2005). Still, for about 5% of right-handers and about 50% of left-handers, language is either localized in therighthemisphere or is processed equally by both hemispheres (Kinsbourne, 1998).
The left hemisphere does indeed perform most of the work in language processing, for the majority of people. The left hemisphere is especially skilled at speech perception;
it quickly selects the most likely interpretation of a sound (Gernsbacher & Kaschak, 2003; Scott, 2005). Furthermore, the left hemisphere determines cause-and-effect rela- tionships (Gazzaniga et al., 2002). It also excels at reading, as well as at understanding meaning and grammar (Gernsbacher & Kaschak, 2003). In addition, high-imagery sen- tences activate the left hemisphere (Just et al., 2004).
For many years, people thought that the right hemisphere did not play a role in language processing. However, we now know that this hemisphere does perform some tasks. For example, the right hemisphere interprets the emotional tone of a message (Gernsbacher & Kaschak, 2003; Vingerhoets et al., 2003). It also plays a role in appre- ciating humor (Shammi & Stuss, 1999). In general, then, the right hemisphere is responsible for more abstract language tasks (Gernsbacher & Kaschak, 2003).
The left and right hemispheres often work together on tasks such as interpreting subtle word meanings, resolving ambiguities, and combining the meaning of several sentences (Beeman & Chiarello, 1998; Beeman et al., 2000; Grodzinsky, 2006). For example, suppose that you are one of the majority of individuals for whom the left hemisphere is dominant for language. Imagine that you see this ambiguous message that I once spotted on a bumper sticker:
SOMETIMES I WAKE UP GRUMPY.
OTHER TIMES I LET HIM SLEEP IN.
On seeing the phrase, “SOMETIMES I WAKE UP GRUMPY,” your left hemi- sphere immediately constructs a meaning in whichGRUMPYrefers to “I” (that is,
the owner of the car). After reading the next sentence, “OTHER TIMES I LET HIM SLEEP IN,” your right hemisphere searches for a less obvious interpretation, in which GRUMPYrefers to another person. Fortunately, when people have normal brain func- tions, both hemispheres work together in a complementary fashion (Gazzaniga et al., 2002). Before you read further, try Demonstration 9.4.
Demonstration 9.4
Reading Two Sets of Sentences A. Read the following set of sentences:
A grandmother sat at a table.
A young child played in a backyard.
A mother talked on a telephone.
A husband drove a tractor.
A grandchild walked up to a door.
A little boy pouted and acted bored.
A grandmother promised to bake cookies.
A wife looked out at a field.
A family was worried about some crops.
B. Now read the following set of sentences:
The grandmother sat at a table.
The young child played in a backyard.
The mother talked on a telephone.
The husband drove a tractor.
The grandchild walked up to a door.
The little boy pouted and acted bored.
The grandmother promised to bake cookies.
The wife looked out at a field.
The family was worried about some crops.
Now answer this question: The first set and the second set differ only with respect to the first word in each sentence. However, did you have a general feeling that the two sets differed in the overall meaning that they conveyed?
Source:Gernsbacher & Robertson, 2005, pp. 159-160.
The Nature of Language 303
Neuroimaging Research with Normal Individuals. During the past decade, researchers have increasingly used the fMRI technique to investigate language in humans. Functional magnetic resonance imaging (fMRI)is based on the princi- ple that oxygen-rich blood is an index of brain activity (Cacioppo & Berntson, 2005b;
Kalat, 2007; Mason & Just, 2006).
An fMRI is superior to a PET scan in detecting changes that occur very quickly.
An fMRI is also safer than a PET scan, because a PET scan requires an injection of radioactive material. However, one disadvantage is that fMRI values can be inaccurate when participants move their heads even slightly (Saffran & Schwartz, 2003). You can guess, therefore, that fMRIs are more suitable for language comprehension than for language production. (Try talking without moving any portion of your head!)
Research using the fMRI technique shows that several regions of the left tempo- ral lobe process semantic information. Therefore, processing the meaning of words is not confined to just a small region of the cortex. Furthermore, most people who have experienced damage to the left temporal lobe still understand the general meaning of a message. Typically, however, they make some relatively minor comprehension errors (Saffran & Schwartz, 2003).
In the previous discussion of hemispheric specialization, we emphasized that the right hemisphere plays an important role in language comprehension, even though the left hemisphere receives most of the publicity in the media. Morton Ann Gerns- bacher and David Robertson (2005) provide a good example of the subtlety of right- hemisphere processing. Specifically, they created several sets of sentences, such as the two kinds of sentences you read in Demonstration 9.4.
The first set of sentences began with “A,” whereas the second set began with “The.”
Would you expect such a subtle change to make a difference in the fMRI patterns? Gerns- bacher and Robertson (2005) found that the two sets of sentences produced virtually identical patterns of activation in the left hemisphere. In contrast, the right hemisphere responded very differently to the two sets. As Gernsbacher and Robertson emphasize, when a series of sentences uses the word “The,” it sounds more like a story in which the grandmother, the child, and other family members are connected with each other. In contrast, the string of sentences with the word “A” seems disconnected; the characters don’t seem like a cohesive unit. Impressively, the right hemisphere manages to respond differently to connected language than to disconnected language. The research on longer, more sophisticated language sequences is just beginning (Mason & Just, 2006).
In summary, the neuropsychology research highlights the complexity of our lan- guage skills. We’ve considered studies about individuals with aphasia, information about hemispheric specialization, and fMRI research. Neurolinguists still have a dis- tance to travel, but they have identified some important information about the coor- dinated brain regions that allow us to understand language.
In the first part of this chapter, we examined the basic structure of language, the history of psycholinguistics, factors affecting comprehension, and neurolinguistics.
Let’s summarize this background knowledge and then turn our attention to the impor- tant topic of reading.
Section Summary: The Nature of Language
1. Some of the central concepts in psycholinguistics are the phoneme, the mor- pheme, morphology, syntax, grammar, semantics, semantic memory, and pragmatics.
2. Because most psycholinguistics research focuses on English, we do not know whether the findings can be applied to other languages.
3. People use the information in constituents to determine meaning; working memory stores the constituent that people are currently processing, until they can determine its meaning.
4. According to Noam Chomsky, (a) language skills are innate in humans, (b) language is modular, and (c) the deep structure of a sentence captures its core meaning.
5. Many current psychologists emphasize the meaning of language, rather than linguistic structure; for example, the cognitive-functional approach to language emphasizes that we design our language so that listeners will pay attention to the information we want to emphasize.
6. Sentences are more difficult to understand if they (a) contain negatives, (b) use the passive voice, (c) include nested structures, and (d) are ambiguous.
7. Neurolinguistic research on adults with aphasia suggests that damage in Broca’s area usually leads to difficulty in producing language, whereas damage in Wernicke’s area usually leads to difficulty in understanding language; however, the distinction is not clear-cut.
8. The left hemisphere typically performs most components of language process- ing, such as speech perception, reading, and syntax processing; however, the right hemisphere performs tasks such as interpreting a message’s emotional tone, appreciating humor, and forming connections among sentences.
9. Research using fMRIs highlights a variety of brain regions that are responsible for language-related activities; for example, this research shows that the right hemisphere can distinguish between connected language and a series of uncon- nected sentences.
BASIC READING PROCESSES
Reading seems so simple to competent adult readers that we forget how challenging the task is for most children (Rayner et al., 2001). Take a minute to think about the impressive variety of cognitive tasks you perform when reading a paragraph like this one. Reading requires you to use many cognitive processes we have discussed in pre- vious chapters. For example, you must recognize letters (Chapter 2), move your eyes across the page (Chapter 3), use working memory to remember material from the sen- tence you are currently processing (Chapter 4), and recall earlier material that is stored in long-term memory (Chapters 5 and 6). You also need to use metacomprehension
Basic Reading Processes 305 to think about the reading comprehension process (Chapter 6). In some cases, you
must also construct a mental image to represent the scene of the action in the passage you are reading (Chapter 7). In addition, you must consult your semantic memory, your schemas, and your scripts when you try to understand the paragraph (Chapter 8).
In Demonstration 9.1 and throughout this book, we emphasize that the cognitive processes are interrelated (Theme 4). Reading is an important activity that requires virtually every cognitive process discussed in this textbook. Despite the complexity of the reading process, however, we are usually blissfully unaware of the cognitive effort that reading requires (Gorrell, 1999). For example we can silently identify an isolated word in 200 milliseconds, which is 1/5 of a second. In addition, we manage to read with impressive efficiency, typically at the rate of about 250 to 300 words per minute (Rayner et al., 2003; Wagner & Stanovich, 1996). Consistent with Theme 2, reading is remarkably efficient and accurate.
Here’s an additional reason to admire your reading skills: In English, we do not have a one-to-one correspondence between letters of the alphabet and speech sounds.
These irregular pronunciations make English more challenging than languages such as Spanish (Rayner et al., 2003).Try Demonstration 9.5 to illustrate this point. As we noted at the beginning of this chapter, most of the psycholinguistic research examines people whose language is English. We therefore cannot generalize this research to Spanish readers or readers whose language uses symbols to represent complete words (Wagner et al., 2006).
Demonstration 9.5
Noticing That Letters of the Alphabet Do Not Have a One-to-One Correspondence with Speech Sounds
Each of the words below has a somewhat different pronunciation for the letter sequenceea.Read each word aloud and notice the variety of phonemes that can be produced with those two letters.
beauty bread clear create
deal great heard
knowledgeable
react séance bear dealt
As you have demonstrated, this two-letter sequence can be pronounced in twelve different ways. Furthermore, each phoneme in the English language can be spelled in a variety of ways. Go back over this list of words and try to think of another word that has a different spelling for that phoneme. For example, theea(u)phoneme inbeautyis like theiewphoneme inview.
Source:Based on Underwood & Batt, 1996.
Let’s begin this section on basic reading processes by comparing written lan- guage with spoken language. Then we’ll explore how we discover the meaning of an unfamiliar word. We’ll also see how working memory plays a role in reading, and then we’ll consider the two pathways we use when recognizing words. A sepa- rate section in this chapter, on discourse processing, will examine how we understand larger units of language—such as sentences and stories—in both written and spo- ken language.
Comparing Written and Spoken Language
In Chapter 2, we explored several components of spoken language comprehension.
In this section onwrittenlanguage comprehension, we encounter a somewhat dif- ferent set of challenges. Reading and the comprehension of spoken language dif- fer in important ways (Ainsworth & Greenberg, 2006; Dahan & Magnuson, 2006;
Nelson et al., 2005; Rayner & Clifton, 2002; Saffran & Schwartz, 2003; Treiman et al., 2003):
1. Reading is visual and is spread out across space, whereas speech is auditory and is spread out across time.
2. Readers can control the rate of input, whereas listeners usually cannot.
3. Readers can re-scan the written input, whereas listeners must rely much more heavily on their working memory.
4. Writing is relatively standardized and error free, whereas variability, errors, sloppy pronunciation, and interfering stimuli are common in speech.
5. Writing shows discrete boundaries between words, whereas speech does not.
6. Writing is confined to the words on a page, whereas speech is supplemented by nonverbal cues and by additional auditory cues—such as stressed words and variations in pace—that enrich the linguistic message.
7. Children require elaborate teaching to master written language, whereas they learn spoken language very easily.
8. Adults who can read tend to learn new words more quickly when they appear in a written form, rather than a spoken form.
As you can imagine, these characteristics of written language have important impli- cations for our cognitive processes. For example, we can consult the words on a page when we want to make sense out of a passage in a book—a luxury we seldom have with spoken language. Despite the differences between written and spoken language, how- ever, both processes require us to understand words and appreciate the meaning of sen- tences. In fact, the research on individual differences highlights the similarity between the two comprehension processes. For adults, scores on reading comprehension tests are highly correlated with scores on oral comprehension tests; typically, the correlation is about +.90 (Rayner et al., 2001).
Basic Reading Processes 307
Discovering the Meaning of Unfamiliar Words
Chapter 2 examined how context aids both the visual recognition of letters and the auditory recognition of phonemes. Context also helps you recognize words. Specifically, you perceive familiar words more accurately when they are embedded within the mean- ingful context of a sentence (Rayner et al., 2003). Earlier in the chapter, we also saw that context helps to resolve the meaning of an ambiguous word.
In addition, context is vitally important when people want to discover the mean- ings ofunfamiliarwords (Rayner et al., 2003). Try Demonstration 9.6, which is an example of the passages used by Sternberg and Powell (1983) in their work on verbal comprehension.
Sternberg and Powell proposed that context can provide several kinds of informa- tion cues about the meaning of an unknown word. For instance, context can help us understand when and where this unknown item occurs. Consider the following sentence:
“At dawn, theblenarose on the horizon and shone brightly.”
This sentence contains several contextual cues that make it easy to infer the mean- ing ofblen.For instance, the phrase “at dawn” provides a cue about the time at which the arising of theblenoccurred. The wordaroselimits the possible candidates forblen to those things that move or appear to move. Other words and phrases in the sentence are equally helpful. With all these cues, an experienced reader can easily understand that the nonsense wordblenis a synonym for the familiar wordsun.
Demonstration 9.6
Figuring Out the Meaning of a Word from Context
Read the paragraph below. Then define, as precisely as possible, the words that are italicized.
Two ill-dressed people—the one a tired woman of middle years and the other a tense young man—sat around a fire where the common meal was almost ready. The mother, Tanith, peered at her son through theoamof the bubbling stew. It had been a long time since his lastceilidhand Tobar had changed greatly; where once he had seemed all legs and clumsy joints, he now was well-formed and in control of his hard, young body. As they ate, Tobar told of his past year, re-creating for Tanith how he had wandered long and far in his quest to gain the skills he would need to be permitted to rejoin the company. Then all too soon, their briefceilidhover, Tobar walked over to touch his mother’s arm and quickly left.
Source:Based on Sternberg & Powell, 1983.
Contextual cues are especially useful if the unknown word appears in several dif- ferent contexts. According to the research, words that appear in a rich context of different cues are more likely to be accurately defined (Sternberg & Powell, 1983).
As you might expect, the students in Sternberg and Powell’s study showed large individual differences in their ability to use contextual cues and to provide accurate definitions for the unfamiliar words. The students who were particularly good at this task also had higher scores on tests of vocabulary, reading comprehension, and gen- eral intelligence. (Incidentally, in the passage in Demonstration 9.6,oam means
“steam” and aceilidhis a “visit.”)
Reading and Working Memory
Working memory plays an important role during reading (Carpenter et al., 1995;
Carroll, 2004; Martin, 2007). The research shows that readers who have a relatively large working-memory span can quickly process ambiguous sentences (Miyake et al., 1994).
In addition, people with large working-memory spans are especially skilled in reading difficult passages and solving complex verbal problems (Haarmann et al., 2003; Long et al., 2006).
Working memory also helps us to understand complicated sentences (Carpenter et al., 1994, 1995; Just et al., 1996; Martin, 2007). People who can maintain many items in memory—while they unravel a sentence—are quick and accurate in under- standing complex sentences such as “The reporter whom the senator attacked admit- ted the error.”
This information about reading and working memory is an excellent illustration of Theme 4. The cognitive processes do not operate in isolation. Instead, reading skill depends heavily on other cognitive abilities, such as working memory.
Two Pathways for Reading Words
So far, our examination of reading in this textbook has emphasized how we identify alphabetical letters (Chapter 2), how our saccadic eye movements scan a line of text (Chapter 3), how we discover the meaning of an unfamiliar word, and how working memory plays a role in reading. Now we’ll address an important question about read- ing: How do we look at a pattern of letters and actually recognize that word? For exam- ple, how do you manage to look at the eleven letters in the fourth word in this paragraph and realize that it saysexamination?
For several decades, researchers debated whether readers actually “sound out”
words while reading a passage. Some researchers concluded that readers always sound out the words, and other researchers concluded that they never sound them out. In the current era, the debate is mostly resolved (Coltheart, 2005). You have probably com- pleted enough psychology courses to guess the answer: Sometimes readers sound out the words, and sometimes they do not. In fact, the dual-route approach to reading specifies that skilled readers employ both a direct-access route and an indirect-access route (Coltheart, 2005; Mayer, 2004).
Basic Reading Processes 309 1. Sometimes you read a word by a direct-access route;you recognize this word
directly through vision. That is, you look at the wordexamination,and the visual pattern is sufficient to access the word and its meaning. You are especially likely to use direct access if the word has an irregular spelling and cannot be “sounded out”—for example, the wordsoneorthrough.
2. Other times, you read a word by an indirect-access route;you recognize this word indirectly by sounding out the word. In more detail, as soon as you see a word, you translate the ink marks on the page into some form of sound, before you can access a word and its meaning (Rayner et al., 2003; Treiman et al., 2003). You are especially likely to use indirect access if the word has a regular spelling and can be sounded out—for example, the wordstenandcabinet.
Notice why this second process is indirect. According to this explanation, you must go through the intermediate step of converting the visual stimulus into a phonological (sound) stimulus. Think about whether you seem to use this intermediate step when you read. As you read this sentence, for example, do you have a speech-like representation of the words? You probably don’t actually move your lips when you read, and you cer- tainly don’t say the words out loud. But do you seem to have an auditory image of what you are reading?
Let’s discuss the research supporting each of these routes. Then we’ll consider the implications for teaching reading to children.
Research on the Dual-Route Approach. We’ll begin with a classic study that sup- ports the direct-access route. It demonstrates that people can recognize a word visu- ally, without paying attention to the sound of the word. Bradshaw and Nettleton (1974) showed people pairs of words that were similar in spelling, but different in sound, such asmown–down, horse–worse, andquart–part.In one condition, the participants were instructed to read the first word silently and then pronounce the second word out loud. Now, if they had been translating the first member of a pair into sound, the sound ofmownwould interfere with sayingdownout loud. However, the results showed that the participants experienced no hesitation in pronouncing the second word. This finding—and other similar studies—suggests that we do not silently pronounce each word during normal reading (Coltheart, 2005)
Now let’s shift to the research on the indirect-access approach. Many studies sug- gest that we often translate visual stimuli into sound during reading (Coltheart, 2005).
Furthermore, the sound coding may assist working memory, providing an additional advantage during reading (Rayner et al., 2003).
A study by Luo and his coauthors (1998) provides evidence for the indirect-access approach in adult readers. These researchers instructed college students to read a series of pairs of words and decide whether the two words were related or unrelated in mean- ing. A typical pair in the experimental condition wasLION–BARE.As you know, the wordBAREsounds the same as the wordBEAR,which is indeed semantically related toLION.The students frequently made errors on these pairs, because they incorrectly judged the two words as being semantically related. This error pattern suggests that they
were silently pronouncing the word pairs when they made the judgments. In contrast, they made relatively few errors on control-condition word pairs, such asLION–BEAN.
In this word pair, the second word looked like the wordBEAR, although it did not sound the same.
Word sounds may be especially important when children begin to read. Numer- ous studies demonstrate that children with high phonological awareness have superior reading skills. That is, the children who are able to identify sound patterns in a word also receive higher scores on reading achievement tests (Levy, 1999; Wagner &
Stanovich, 1996).
Perhaps you’re thinking that children may need to translate the printed word into sound. After all, children even move their lips when they read, but adults usually do not. Try Demonstration 9.7 and see whether you change your mind. Adults read
“tongue twisters” very slowly, which indicates that—at least in some circumstances—
they are indeed translating the printed words into sounds (Harley, 2001; Keller et al., 2003; Perfetti, 1996).
As we noted earlier, the dual-route approach has the definite advantage of flexibil- ity. This approach argues that the characteristics of the reading material determine whether access is indirect or direct. For instance, you may use indirect access the first time you see a long, uncommon word; you may use direct access for a common word (Bernstein & Carr, 1996).
The dual-route approach also argues that characteristics of the reader determine whether access is indirect or direct. Beginning readers would be especially likely to sound out the words, using indirect access. Experienced readers would be especially likely to recognize the words directly from print. Adults also vary in their reading styles.
Demonstration 9.7
Reading Tongue Twisters
Read each of the following tongue twisters silently to yourself:
1. The seasick sailor staggered as he zigzagged sideways.
2. Peter Piper picked a peck of pickled peppers. A peck of pickled peppers Peter Piper picked.
3. She sells seashells down by the seashore.
4. Congressional caucus questions controversial CIA-Contra-Crack connection.
5. Sheila and Celia slyly shave the cedar shingle splinter.
Now be honest. Could you “hear” yourself pronouncing these words as you were reading? Did you have to read them more slowly than other sentences in this book?
Basic Reading Processes 311 College students who are good readers typically use direct access, whereas college stu-
dents who are relatively poor readers typically use indirect access (Jared et al., 1999).
At present, the dual-route approach seems like an intelligent compromise. The dual-route approach is also consistent with brain-imaging research (Jobard et al., 2003).
Readers can identify words either directly or indirectly, depending on the characteris- tics of both the text and the reader.
Implications for Teaching Reading to Children. For many years, reading teachers and reading researchers debated about the most effective way to teach reading. In gen- eral, those who favored the direct-access approach also favored the whole-word approach. The whole-word approachargues that readers can directly connect the written word—as an entire unit—with the meaning that this word represents (Rayner et al., 2001). The whole-word approach emphasizes that the correspondence between the written and spoken codes in English is notoriously complex, as we saw in Demon- stration 9.5. Supporters therefore argue that children should not learn to emphasize the way a word sounds. Instead, the whole-word approach encourages children to identify a word in terms of its context within a sentence. The problem, however, is that even skilled adult readers achieve only about 25% accuracy when they look at an incomplete sentence and guess which word is missing (Perfetti, 2003; Snow & Juel, 2005).
In contrast, people who favor the indirect-access hypothesis typically support the phonics approach. The phonics approachstates that readers recognize words by try- ing to pronounce the individual letters in the word. If your grade school teachers told you to “sound it out” when you stumbled on a new word, they championed the phon- ics approach. The phonics approach argues that speech sound is a necessary interme- diate step in reading. It also emphasizes developing young children’s awareness of phonemes. According to the research, it’s clear that phonics training helps children who have reading problems (McGuinness, 2004; Perfetti, 2003; Snow & Juel, 2005). For example, a meta-analysis of thirty-four studies showed that phonological training pro- grams had a major impact on children’s reading skills (Bus & van IJzendoorn, 1999).
For many years, the debate between the whole-word supporters and the phonics supporters was feverish (McGuinness, 2004; Smith, 2004). In the current decade, how- ever, most educators and researchers support some form of a compromise: Children should be taught to use phonics to access the pronunciation of a word; they should also use context as a backup to confirm their initial hypothesis. Even the strongest phonics supporters would also agree that teachers should encourage children to recognize some words by sight alone.
Furthermore, educators typically favor some components of an approach called the whole-language approach (as opposed to the whole-word approach). According to the whole-language approach,reading instruction should emphasize meaning, and it should be enjoyable, to increase children’s enthusiasm about learning to read. Children should read interesting stories and experiment with writing before they are expert spellers. They should also use reading throughout their classroom experiences (Luria, 2006; McGuinness, 2004; Snow & Juel, 2005).
Before we leave this section on reading, however, we need to emphasize an impor- tant point. This discussion assumes that children and adults have had the opportunity
to learn how to read. In Canada and the United States, about 98% of adults have achieved basic literacy (Luria, 2006). However, the reality is that more than 800 mil- lion adults throughout the world are illiterate. Approximately two-thirds of these indi- viduals are women. Clearly, people who cannot read face tremendous disadvantages with respect to employment, health care, and everyday communication.
1. Reading is a challenging cognitive task that differs from understanding spo- ken language in many respects. For example, readers can control the rate of input and they can re-scan the text, and writing shows clear-cut boundaries between words.
2. Readers often use a variety of contextual cues to determine the meaning of an unfamiliar word.
3. Working memory helps readers decode ambiguous sentences and understand complex sentences.
4. Working memory plays a critical role in processing ambiguous or complicated sentences.
5. The dual-route approach argues that readers sometimes recognize a word directly from the printed letters (i.e., direct access), and sometimes they convert the printed letters into a phonological code to access the word (i.e., indirect access).
6. The whole-word approach emphasizes visual recognition of words, whereas the phonics approach emphasizes sounding out the word. Most educators and researchers favor a combination of these approaches.
7. The whole-languageapproach emphasizes language meaning, as well as inte- grating reading throughout the curriculum.
UNDERSTANDING DISCOURSE
We began this chapter with an overview of the nature of language; that overview consid- ered both linguistic theory and the biological basis of language. Then we explored basic reading processes. You’ll notice that these topics all focus on the way we process small units of language, such as a phoneme, a letter, a word, or an isolated sentence. In your daily life, however, you are continually processing connected discourse,or language units that are larger than a sentence (Bamberg & Moissinac, 2003; Treiman et al., 2003). You listen to the news on the radio, you hear a friend telling a story, you follow the instructions for assembling a bookcase . . . and you read your cognitive psychology textbook.
In Chapters 1 and 8, we considered Frederick Bartlett’s (1932) research, which focused on these larger linguistic units. Specifically, Bartlett demonstrated that peo- ple’s recall of stories becomes more consistent with their schemas after a long delay. For
