<<

CHAPTER6

Psycholinguistics

HERBERT H. CLARK AND MIJA M. VANDER WEGE

Psycho is the study of the processes Modern psycholinguistics is diverse in its by which people use . In conversa- perspectives, theories, approaches, and goals. tion, people engage in actions that range from At its center is how people process language- producing and interpreting to steering from producing speech sounds and under- the course of the conversation--determining standing to participating in discourse. what topics are taken up when. In , But it also includes first and people apply many of the same processes, acquisition, aphasia, speech disorders, read- but by using a skiii that has taken years to ing, and many other issues. Unlike many areas learn. In writing, authors compose, edit, and of , psycho linguistics has borrow- rewrite to engineer just the right experience ed heavily from other disciplines-linguistics, for their readers. When we think of language philosophy, computer science, sociology, and use, we tend to focus on words, phrases, and anthropology. It has also drawn upon broad sentences, but these are often parts of com- evidence-laboratory experiments, field ex- posite actions that include pointing and other periments, linguistic intuitions, computer sim- gestures as well. ulations, large corpora of conversations, clin- Psycholinguistics was launched in 1900 ical case studies, and much more. There is with the publication of 's Die no royal road to knowledge in the study of Sprache (Language) as the first two vol- psycholinguistics. umes of his monumental Volkerpsychologie. In this chapter, we focus on the core of Wundt' s enterprise was broad, and it led to psycholinguistics-the elements we believe such distinguished works as Karl Buhler's make it a field. Our goal is not to review the Sprachtheorie (Language Theory) in 1934. field, but to frame it. It is to describe the foun- By the middle of the 20th century, psycholin- dational issues and principles. We begin with guistics had run into rough weather and, at (why people use language in least in America, had almost disappeared. the first place), then take up speaking and lis- In the 1960s, it was revived with Noam tening, and finally tum to the mental repre- Chomsky's (1957, 1965) vision of language sentations necessary for using language. and linguistics, where it often got narrowed to the study of the ''psychological reality of COMMUNICATION linguistic structures." By its hundredth birth- day, psycholinguistics had matured into a field To use language-to speak or listen, to read in its own right. or write-is to take action (Austin, 1962;

209 !'·

210 Psycholinguistics

Levinson, 1983; Searle, 1969, 1975b; Sacks, what they are told. In plays, actors recite lines l Schegloff, & Jefferson, 1974). People choose already memorized. But in spontaneous con- to speak or not to speak, and they try, or do versation, speakers decide what they want to not try, to attend to, identify, understand, and talk about, plan their own words, and - react to what others say. Psycholinguistics is duce them. Managing all three processes- about the social and cognitive processes by especially while under time pressure-is a which people carry out these actions. delicate act of juggling. Spontaneous speak­ ing is clearly different from reading aloud, repeating back, and reciting. But how are they Language Settings alike, and how are they different? And how do Language gets used in a wide range of settings listening, writing, and reading change with the (Clark, 1996). It atises as spoken language setting? in personal and nonpersonal settings (e.g., One setting is basic, and that is face-to-face face-to-face vs. lectures), institutional set- conversation (Clark, 1996; Fillmore, 1981). It tings (courts, church, etc.), fictional settings is the only setting that is universal to all the (movies, plays), and private settings (talking world's peoples, about a sixth of whom are to oneself). It comes as written language in illiterate. It is the setting in which all of the just as many settings-personal letters, news- world's evolved before the spread paper stories, institutional letters and labels, of literacy. It is the only setting that does fictional novels and contic strips, and private not require specialized skills such as reading, notes to oneself. With the invention of new writing, or oratory. It is the setting in which communication technologies, there seems to children acquire the rudiments of their first be no end to the settings in which people use language; learning from books and television language. comes later. Other settings can be viewed as The processes that people use in these set- secondary to, or derivative from, face-to-face tings range just as widely as the settings them- conversation. People understand what they selves. It is self-evident that speaking and read, for example, largely by treating printed listening are different from writing and read- language as if it were a representation of spo- ing. Speaking requires the execution of vocal ken language. sounds, words, and phrases in a tight temporal So, psycholinguistics must account first pacing. Writing, in contrast, requires a manual and foremost for face-to-face conversation. skill, learned over years of training, which It must go beyond reading aloud, repeat- can be done at any pace and with as much ing back, and reciting, and understanding editing and rewriting as needed. Listening re- this speech. It must account for how people quires the aural skill of identifying sounds, plan, speak, listen, and gesture-how they words, and phrases as they are produced in communicate-in the give-and-rake of spon- time. Reading, in contrast, depends on a visual taneous dialogue. Eventually, it must account skill, also learned over years of training, which for all language settings, but these accounts can be done at any pace, with as much reread- differ from setting to setting. ing as needed. Speaking, listening, reading, and writing Language in Joint Activities themselves change radically with the setting. Take speaking, for example. On television, People use language to do things. In all but news anchors read aloud what is already writ- one of the settings we .have reviewed, people ten. In weddings, the bride and groom repeat use language to do things with others. Using Communication 211

language is inherently social, and that is Alan and Beth must coordinate on its partic­ nowhere more evident than in face-to-face ipants, timing, and content. (1) Who the par- conversation-the primary setting. But what ticipants are gets established when Alan ad- is dialogue for? To answer this question, we dresses Beth with "hi" and she acknowledges . draw on 30 years of close of spon- (probably by nodding and meeting Alan's taneous conversation recorded in a variety gaze). (2) The time they start is established of settings (e.g., Atkinson & Heritage, 1984; when Alan says "hi," and Beth, with her nod Button & Lee, 1987; Drew & Heritage, 1992; and eye-gaze, agrees. (3) The content of their Sacks et a!., 1974; Schegloff, Jefferson, & basic activity-its public goal-gets estab- Sacks, 1977). lished in two steps as Alan proposes the pur- Joint ·activities are activities that two or chase of four size-C flashlight batteries, and more people can only carry out by coordinat- Beth agrees to the proposal by turning to ing with each other (see Clark, 1996). Such get them. Each piece of the dialogue is de- examples include one person helping another signecl to coordinate a piece of the basic joint person put on his or her coat; four musi- activity. cians playing a string quartet; people playing Simple as this example is, it illustrates a game of football or chess or poker; a per- several points. First, people distinguish basic son buying goods from a clerk in a store; two from coordinating activities. If Alan were lawyers negotiating a contract; and two people asked what he did in the drug store, he would gossiping. The participants in each activity, answer, "I bought four batteries," not "I talked as distinguished from bystanders, assume par- to the server" (even though he did). The pur- ticular roles (e.g., dealer vs. players in poker) chase was primary, and the talk was only as they presuppose or establish common goals secondary-in support of the purchase. Sec- (e.g., completing the poker game) and even ond, people coordinate on basic activities in pursue their own private agendas. Joint activ- increments. Alan and Beth first establish the ities have coordinated beginnings, ends, and participants and starting time ("Hi" plus the subsections, and the participants have con- nod), then a prerequisite for Alan's order ventional andnonconventional procedures for ("Do you have size C flashlight batteries?" achieving this coordination (e.g., dealing plus "Yes, sir"), and then Alan's order proper cards, saying "I raise you ten," etc.). Finally, ("I'll have four please" plus her turning away). people often engage in more than one joint Third, the participants' actions depend turn- activity at the same time or intermittently by -turn on the actions of the other participants. (e.g., gossiping and eating dinner). Beth, for example, could have refused Alan's Dialogue is a means of coordinating "Hi" with "Uh, wait a minute" or "Sorry, I'm actions in joint activities. Take this brief ex- busy." Or she could have said "No, sir" instead change at a drug store counter between Alan, a of "Yes, sir," and Alan would have followed customer, and Beth, the server (Merritt, 1976, up with another direction. These features are p. 324); characteristic of joint activities.

(1) Alan Hi. Do you have uh size C flashlight batteries? Joint Projects Beth Yes, sir. Alan I'll have four please. Each increment to a joint activity takes coordi- Beth [turns to get) nation. Alan cannot advance his business with The basic joint activity is a business trans- Beth without her agreement, and vice versa, action, the purchase of batteries. To succeed, and that normally requires actions from both. ..

,I 212 Psycholinguistics

A common way to reach agreement is via call from Jane to Kate (see Clark, 1994): adjacency pairs, as shown here: Joint project Example I. Alan Do you have uh size C flashlight batteries? 2. Beth Yes, sir. 1. Summons Jane (rings telephone) 2. Response Kate Miss Pink's office An adjacency pair consists of two utterances, 1. Greetings Kate hello 2. Greetings by two speakers, in which the first utterance Jane hello is of a type (e.g., a question) that makes an 1. Question Kate who is it? 2. Answer oh it's Professor Worth's utterance of a second type (e.g., an answer) Jane secretary, from Pan- conditionally relevant as the next utterance American College (Schegloff & Sacks, 1973). Once Alan has 1. Assertion Jane oh it's Professor Worth's asked his question, it is conditionally relevant secretary, from Pan- for Beth to answer it Adjacency pairs must American College be spoken communicative acts, so if Beth had 2. Assent Kate m nodded instead of saying "yes," that would 1. Request Jane could you give her a no longer be an adjacency pair. The following message for me 2. Promise Kate certainly pair of actions would not be an adjacency pair either: 1. Promise Kate I'll tell her 2. Acknowledgment Jane thank you

1. Alan I'll have four please. 1. Thanks Kate thank you very much 2. Beth [turns to get} indeed 2. Acknowledgment Jane right In this chapter we use the term projective pair 1. Good-bye Kate bye bye to cover both spoken and nonspoken pairs of 2. Good-bye Jane bye conditionally relevant actions. A projective pair is really a minimal joint There are many other types as well. project (Clark, 1996). When Alan says, "Do People can create larger joint projects by you have uh size C flashlight batteries?" he combining minimal ones, and there are three proposes, or projects, a joint action for Beth main ways of achieving this--chaining, em- and him to carry out: She is to tell him whether bedding, and pre-sequencing. she has size C flashlight batteries. When Beth Chaining says, "Yes, sir," she takes up Alan's proposal I. is illustrated in the telephone and tells him what he wants to know, com- call from Jane to Kate in these three turns: pleting her part of the projected joint action. Question 1 Kate who is it? The result is a ntinimal joint project with two Uptake 1 =Assertion 2 Jane oh it's Professor Worth's parts: secretary, from Pan- American College 1. Proposal A proposes a joint project for A and B Uptake 2 Kate m 2. Uptake B takes up A's proposal The first two turns constitute one minimal This schema also provides a rationale for Alan joint project-a proposal (Kate's question) and Beth's second pair of actions-"I'll have plus its uptake (Jane's answer). But Jane's an- four please" and "[turns to get]." swer itself initiates a second joint project She Minimal joint projects come in great vari- proposes that Kate assent to her claim of being ety. Here are examples from a single telephone Professor Worth's secretary, and Kate takes Communication 213 her up with "m" ("yes"). So Jane's utterance taken up with an offer, as here: I, is both the uptake in one joint project and the Question l =Pre-request 2 Customer Do you have the I, proposal of a second, linking the two joint pecan danish ' i, projects together in a chain. today? !': 2. Embedding is illustrated in this ex- Uptake I Server Yes we do. Uptake 2 =Offer 3 Would you like change between Susan, a waitress, and Jean, one of those? L a customer (Merritt, 1976): Uptake 3 =Request 4 Customer Yes, please. 'i I'· Uptake 4 Server [turns to get] lj!! Question l Susan What'll ya have girls? il The entire sequence may get compressed into ::,! Question 2 Jean What's the soup of the day? two turns, as in this phone call to a liquor store: !Iii

Uptake 2 Susan Clam chowder Question 1 =Pre-request 2 Susan Do you have a price on a fifth Uptake 1 Jean I'll have a bowl of clam chowder of Jim Beam? and a salad with Russian Uptake 1 Manager Yes, I do. dressing Uptake 2 =Offer 3 Manager It's five dollars and fifty-nine When Susan asks, "What'll ya have girls?" cents. she projects an answer such as "I'll have a It may be compressed even further: ham sandwich." But Jean doesn'thave enough information to answer, so she initiates a sec- Question 1 =Pre-request 2 Susan Can you tell me ond sequence with "What's the soup of the what time you day?" The result is one minimal joint project close? Uptake l=Uptake2 Manager Nine. (question 2 +uptake 2) embedded within an- other (question 1 +uptake 1). The embedded Not only are there pre-requests, but pre- sequence is called a side sequence (Jefferson, questions ("Can I ask you something?"), pre- 1972) or insertion sequence (Schegloff, announcements ("Did you hear what hap- 1972). pened?" or "You know what?"), pre-narratives 3. Pre-sequencing is illustrated in an ex- ("Did you hear the joke about the three Irish- change we have already examined: men?"), and other pre-sequences. It takes the strategic use of chaining, em- Question 1 =Pre-request Alan Do you have uh size bedding, and pre-sequencing to navigate C flashlight larger joint activities. Pre-sequences, for ex- batteries? ample, can be used to project subsections of a Uptake 1 Beth Yes, sir. joint activity-jokes, announcements, request Request 2 Alan I'lt have four please. sequences, and more. They can also be used Uptake2 Beth [turns to get] to project entire joint activities. When Jane rings Miss Pink's telephone, she is propos- When Alan asks, "Do you have uh size C ing not just a local acknowledgemef\t, but an flashlight batteries?" he is projecting a local entire conversation, and when Kate allswers answer of yes or no. At the same time, he is "Miss Pink's office," she takes up both pro- pre-figuring, or projecting, a second exchange posals at once. When Alan says "Hi" to Beth, in which he will request some of those bat- he is proposing not only a greeting, but also teries (see Schegloff, 1980). Alan's first utter- a business transaction, and when she takes ance is taken to be not only a question, but also him up on it, she agrees to both. So, although a pre-request. Indeed, the pre-request may be conversations work tum-by-tum (Sacks eta!., 214 Psycholinguistics I '! 1974), the participants use these strategies for to get Kate to accept her assertion that she projecting broader joint activities. is Worth's secretary. 2. Directives. The point of a directive is to Speech Acts get addressees to do things. When Alan People appear to create dialogues one ut- says, "Do you have size C flashlight bat- terance at a time. By tradition, these utter- teries;' he is trying to get Beth to tell him ances are called speech acts-acts performed something. Directives include questions, in speaking. The philosopher John Austin requests, orders, commands, and even (1962) introduced this idea and distinguished hints. among several types of speech acts. When 3. Commissives. The point of a commissive Alan says, "I'll have four please," he performs is to commit the speaker to a future ac- four speech acts (among others): tion. When Jane says, "I'll tell her," she is committing herself to giving Kate's mes- phonetic act 1. The of making the sounds in sage to Miss Pink. Commissives include "I'll have four please"; promises, offers, and other actions. utterance act 2. The of producing a token of 4. Expressives. The point of an expressive is the sentence "I'll have four please"; to express a certain feeling to addressees. 3. The illocutionary act of ordering four bat- When Kate says, ''Thank you," she is ex- teries from Beth; pressing gratitude toJane. Expressives also 4. The perlocutionary act of trying to get include greetings ("hi"), farewells ("bye"), Beth to sell him four batteries. apologies ("sorry"), and congratulations. I. I The very term "speech act" focuses on speak- A fifth category, called declarations, is a spe- ers and speaking-as if listeners and under- . cialized class performed by speakers in their !': standing were incidentaL And most of those official roles in social institutions. Examples who followed Austin have focused on illocu- include a judge sentencing a prisoner, a ref- tionary acts, even though the other levels are eree saying "foul" in a tennis match, or a poker also important. player saying "I raise you five." Everyday illocutionary acts can be classi- Viewed this way, illocutionary acts are fied by their public point or purpose. Accord- best classified by their role in minimal joint ing to one proposal (Bach & Harnish, 1979; projects. Alan's utterance of "Do you have Searle, 1969, 1975b), they fall into four major size C flashlight batteries" is a question, a type categories: of directive, because it projects an answer as 1. Assertives. The point of an assertive is to uptake. Other illocutionary acts project other get addressees to accept or reactivate a types of uptake:

Type of Act A's Proposal B's Projected Uptake

I. Assertives A expresses a belief for B to accept B accepts A's belief 2. Directives A directs B to do an act B commits to doing that act 3. Commissives A commits to doing an act for B B accepts A's commitment 4. Expressives A expresses a feeling forB to accept B accepts A's feeling

certain belief. When Jane says, "Oh it's That is, speakers use illocutiona!y acts to per- Professor Worth's secretary," she is trying formperlocutionary acts, by which they try to Communication 215 get addressees to take on obligations (as with a specific interpretation in mind, and it is up directives) or to accept the speakers' beliefs, to the manager to recognize it. The problem commitments, or feelings (as with assertives, with such a view is that it leaves no role for commissives, and expressives). Speakers nor- the manager. mally expect addressees to complete the pro- Indirect speech acts are better viewed cess with their uptake. People in conversation as pre-sequences (Gibbs & Mueller, 1988; engineer these social exchanges~the accep- Schegloff, 1988). When Susan asks, "Do you tance of beliefs, commitments, feelings, and accept credit cards?" she is initiating a nego­ obligations-so as to coordinate their basic tiation about what she is to be taken to mean, joint activities. and it takes the manager to complete the nego- If illocutionary acts are partly defined by tiation. The manager can reply, "Yes, we do," their role in minimal joint projects, then ad- as in 2, and let Susan initiate the next step in dressees may help determine how they are to the negotiation with "Which ones?'' The man- be classified. When a woman named Susan ager can also shortcut the process by offering called up restaurants and asked, "Do you ac- the information he or she believes Susan will cept credit cards?", she got the first answer ask for, "We accept MasterCard and Visa," 40% of the time and the second 14% of the as in 3. Finally, the manager can answer her time (Clark, 1979): question and shortcut the process, as in "Yes, we accept MasterCard and Visa" (which man- (2) Susan Do you accept credit cards? agers did 33% of the time). To succeed, man- Yes, we do. Manager A agers must try to infer Susan's larger plans, (3) Susan Do you accept credit cards? Manager B We accept MasterCard and Visa. and they clearly did. One manager replied, "Uh, yes, we accept credit cards. But tonight In 2, managers construed Susan as asking a we are closed." Another replied, "Uh-uh. yes/no question, but in 3, managers construed We're not open anyways." Both inferred that her as requesting a list of credit cards. In ef- she intended to eat at the restaurant that night. fect, she left the interpretation up to the man- Some pre-requests are so conventional that agers, because she couldn't correct them to they don't seem to allow such a negotiation. the opposite interpretation ("No, I mean ...") It seems impossible to treat "Can you tell me without offending them. What she was taken the time?'' or "Do you have the time?'' merely to mean-the illocutionary act she was con- as yes/no questions. Yet addressees do have strued as performing-was detennined not options. When Susan asked other businesses, just by her words, but by the manager's up- "Can you tell me what time you close?'' some take. This conclusion may seem paradoxical managers replied, "Six," but others replied, (how can what speakers are taken to mean be "Yes, at six," treating the yes/no question ex- shaped by their addressees?), but it falls neatly plicitly. "Yes, at six" is heard as more po- out of the view oflanguage use as joint action. lite because it explicitly deals with both the Traditionally, questions such as "Do you yes/no question and projected request (Clark accept credit cards?" and "Can you tell me & Schunk, 1980). Uptake plays a role in even what time you close?'' have been called indi­ the most conventional pre-sequences. rect requests (Gordon & Lakoff, 1971; Searle, 1975a). In this view, when Susan asks, "Do Common Ground you accept credit cards?" (literally, a yes/no question), she is indirectly requesting a list of Joint activities are carried out against the par- credit cards. The assumption is that Susan has ticipants' common ground. Common ground 216 Psycholinguistics refers to participants' mutual knowledge, Joint activities are governed by the partici- beliefs, assumptions, and awareness (Clark, pants' common ground. When Alan buys bat- 1996; Clark & Marshall, 1981; Lewis, 1969; teries from Beth at the drug store counter, the Stalnaker, 1978). There are two main types of two of them start with a large body of pre- common ground: communal common ground suppositions-their initial common ground. and personal common ground. They presuppose that they are clerk and cus- Communal common ground is based on tomer at a drug store counter, that certain prac- the communities that people belong to. Sup- tices hold at Philadelphia drug store counters, pose Kenneth and Jane meet and establish and that they both speak English. They may that they both speak English, live in San be wrong, but that is what they presuppose Francisco, and play classical piano. English (see Fussell & Krauss, 1992). As they pro- speakers, San Franciscans, and classical pi- ceed, they take actions to add to that common anists are three communities of shared exper­ ground. They try to update the current state of tise, and we all belong to many such com- their activity-what they have comntitted to munities. The expertise of a community may so far and what is left to do. In their first ex- be based on nationality, residence, education, change, Alan and Beth establish as common occupation, employment, hobby, language, ground that the store sells size C batteries, and religion, politics, ethnicity, club, subculture, in their second, that Alan is comntitted to buy- cohort, or gender. Once Kenneth and 1ane ing four. Joint activities would fail without the establish joint membership in a community, orderly maintenance of common ground. they can take as common ground all the ex- pertise that people in these communities take Grounding for granted. As English speakers, they can pre- Using language is itself a joint activity. When suppose basic English vocabulary and gram- Alan speaks to Beth, the two of them must es- mar. As San Franciscans, they can presup- tablish (a) that she is attending to him, (b) that pose the geography, , and politics of San she is identifying his words and gestures, Francisco. As classical pianists, they can pre- (c) that she understands what he means, and suppose classical composers, techniques of (d) that she is considering taking him up. In playing, and musical genres. general, two people, A and B, have to coordi- Personal common ground is based in- nate their actions at four levels (Clark, 1996; stead on the personal experiences people have Paek, 2000):

Level A's Action B's Action

1. Channel A makes sounds, gestures for B B attends to A's sounds, gestures 2. Signal A produces a signal forB B identifies A's signal 3. Intention A means something for B B understands what A means 4. Project A proposes a joint project B considers the joint project shared with each other. At the drug store Indeed, the two of them try to establish, as counter, Alan and Beth perceive each other common ground, the beliefthat they have suc- standing there, looking at objects, and hearing ceeded at each of these levels well enough for the cash register work. They also talk, point, current purposes, a process called grounding and hand things to each other. Personal com- (Clark & Brennan, 1991; Clark & Schaefer, mon ground is built up from joint perceptual 1989; Clark & Wilkes-Gibbs, 1986). In con- experiences and joint communicative actions. versation, people ordinarily try to ground Communication 217 everything that gets said. They realize. tacitly, signal levels and asking specifically about that a minor misunderstanding or mishearing Johnstone's identity ("senior?"). Only once now may lead to greater troubles later. Atthur has said "yes" is she willing to go How people ground varies with the level. on to her answer. If she hadn't succeeded At the channel level, Alan and Beth may ex- at the channel level, she might have asked, change eye gaze as evidence of joint attention "What?" and Arthur would have repeated the (Goodwin, 1981; Kendon, 1967). At the sig- question. nal and intention levels, Alan looks for posi- Grounding is carried out by and for the I speaker and addressees, and that does not tive evidence from Beth, and she tries to pro- ! vide it. One type of evidence is Beth's uptake, guarantee success for overhearers. In one ex- as here: periment (Schober & Clark, 1989), two peo- ple, whom we will call Ann and Ben, con- Alan Do you have uh size C flashlight batteries? versed freely as Ann got Ben to arrange 12 Beth Yes, sir. Tangram figures (abstract, block-like depic- tions of people) in a particular order. A third When she replies "Yes, sir," she provides evi- person, whom we will call Oscar, sat nearby dence not only that she has attended to Alan's but wasn't allowed to speak, and also tried utterance, but that she believes she has identi- to arrange the 12 figures in that order. That fied and understood it. She also shows that she made Oscar an overhearer. The three were has construed it as a yes/no question. Alan ac- separated by barriers, unable to see each other, cepts all this evidence by going on to say, "I'll and all began as strangers. The figures were have four please." If she had responded, "My not easy to describe. In one case, Ann began, is Beth," he would have evidence of a "Then number twelve, is (laughs) looks like a, failure to understand, and he might repeat his a dancer or something really weird. Urn. and, question. Other times Beth can assert her un- has a square head." Ann and Ben then took derstanding with acknowledgements, or con­ several turns to ground that description, of- tinuers, like "uh huh," "yes," and "mhm," of­ ten using information that Ben presented (e.g., ten called back-channel responses (Schegloff, "and a big fat leg?"). Ben was much more ac- 1982; Yngve, 1970). curate than Oscar in arranging the figures. He Grounding is a two-way process, and ad- made errors 5% of the time, whereas Oscar dressees often initiate repairs when they fail at made errors 22% of the time. Why was Oscar the channel, signal, intention, or project level. so bad? When Ann and Ben grounded their A common strategy is for addressees to initi- descriptions, they were opportunistic in using ate side sequences, as here: information they happened to share. That of- ten left Oscar in the dark. (4) Arthur can I speak to Jim Johnstone please? In summary, there would be no language as we know it if people didn't engage in joint Barbara senior? activities. For two people to play cards, or Arthur yes. move a table, or transact business, they need to Barbara yes--- coordinate their individual actions, and they use dialogue to do that. They use projec- Although Barbara identified Arthur's utter- tive pairs to carry out minimal joint projects, ance, she doesn't understand to which Jim which they combine via chaining, embed- Johnstone he is referring. She implies all this ding, and pre-sequencing, to create larger joint by presupposing success at the channel and projects. The act of communication is itself a 218 Psycholinguistics joint activity, and to coordinate that requires Planning Units l grounding. Speakers cannot formulate an expression or I gesture without some plan. But where do these I plans come from? And what are the plans SPEAKING about? If people use dialogue for advancing basic joint activities, their plans must derive, In face-to-face settings, the current speaker in part, from these activities. produces words and gestures while the others The major units of planning are easiest, to try to attend to, identify, understand, and illustrate in narratives. In a study by Chafe consider them. Although the two processes (1980), people were shown a short movie, of speaking and listening are not autonomous, without dialogue, about farm workers pick- researchers have traditionally investigated ing pears in an orchard and were then asked them separately. Here we consider speaking, to describe what happened. The following is or how speakers work their way, as Levelt an excerpt from one narrative: (1989) put it, ''from intention to articulation." The three main steps en route are conceptual­ izing what to say,formulating how to say it, (5) (a) (.85) A-nd he (.35) sees this three pear and articulating the result. (.20) these three baskets In speaking, speakers begin with at least of pears, (b) (.15) and then sees this man up in the some idea of what they want to do at the mo- (.50) tree, ment. When Jane is asked, "Who is it?'' she (c) and decides (.45) that he'd like must decide, "Do I want to say who I am, some pears. and if I do, how do I want to identify my- (d) And at first looks like he's going to take one or two. self?" Speakers normally begin with incom- (e) (.60) Then decides that he'd (.15) much plete plans, and they often change their minds rather take a whole mid-utterance. basket, (f) (.55) puts the basket on the bike, Speakers cannot express just anything. (g) (.90) tsk a-nd kind of struggles .. What they decide to say (their conceptualiza- (h) cause it's much too big tions) must be expressible in the language they for him. (i) And the bike is much too big are speaking, and they must be able to formu- for him. late the right expressions in time. In English, (j) (1.85) The-n he's riding .. across this .. the ideas of motion and manner can be ex- great (.25) expanse, pressed in a single verb, run ("go fast"), but (k) (.2) and (1.15) a girl comes, [continues] in Japanese they must be expressed in two words. Speakers ofEnglish and Japanese must conceptualize what they say with these targets Pauses (in seconds) are marked in parentheses in mind and then follow through with the right ["(.60)"], slight breaks in tempo by double formulation (Slobin, 1996). periods (" .. "), and prolonged words by dashes Speakers must also coordinate their actions ("a-nd"). Narratives like this show evidence with their addressees. They often signal de- of three levels of planning. lays, describe mistakes, prolong words, and hedge expressions that do not quite fit-all, Intonation Units apparently, to help addressees attend to, iden- tify, and understand their speech and gestures. An important level of planning is the intona- Speakers devote a pall of speaking to manag- tion unit, represented by each line of the ex- ing the process of communicating itself. cerpt. As the name suggests, an intonation unit Speaking 219

has a single intonation contour, or melody, written paragraphs (Chafe, 1979; Gee, 1986). with a .distinctive ending such as a rising or Sections are defined in part by their prosody: falling pitch. There is good evidence that these They tend to begin at a higher pitch and end are units of planning. They must be planned as with a falling-pitch glide. In our excerpt, one a whole for their intonation contours to come section begins at line a, and another at line out right. And they often have entry prob­ j. Sections require even more planning than lems. Speakers need extra planning time be- intonation units, for they display more severe fore starting them, and often reformulate them entry problems. Line a, for example, begins before continuing fluently. In Chafe's (1980) with longer and more frequent disftuencies, pear stories, 88% of the intonation units were "(.85) A-nd (.15) he (.35) sees," than all the preceded by pauses that averaged 1 second in other lines in the section, and so does line j, length. Many also had false starts, prolonged "(1.85) The-n (.2) he's." words, and other disfluencies at or near their Sectionsrepresentanotherlevelofconcep- beginnings. tualization. In narratives, sections have a sin- Intonation units also represent unified con­ gle topic or theme that reflects a single place, ceptual plans. They tend to be single, finite time, and set of characters, and they begin at clauses, that is, clauses with verbs that have discontinuities in the event being described. tense, as in lines a, d, h, i, j, and k. When The sentential and intonation units they con- they are not finite clauses, they are usually tain tend to fall into parallel structures. Nar- constituents of a clause, such as the predicate rators cannot plan the whole narrative before- phrases in lines b, c, e, f, and g. In narra- hand, so they must keep track of where they tives, they are often introduced with and, and are as they create each section, sentential unit, then, but, or so, signaling a continuation of and intonation unit. the story. In the pear stories, intonation units Unlike narratives, dialogues are created averaged six words long and lasted an aver- when, by the participants working together, age of 2 seconds. They represent what Chafe so many of their plans are local. In Jane's (1980) called idea units-single events or fo- telephone call to Kate, illustrated earlier, Jane cal points in the larger event being described. asks, "Could you give her a message for me?'' and once Kate decides to comply, she plans Sentential Units "Certainly" and produces it. She cannot plan These consist of one or more intonation units "Certainly" until she has understoodJ ane's re- (an average of four in the pear stories) that quest. Yet local plans are part of larger plans. end with a tenninal contour reserved for sen- Kate's local plan to comply with Jane's re- tences. In the previous excerpt, intonation quest is part of her larger plan to pass infer- units are marked with commas, and senten- marion to Miss Pink. When Alan says "Hi" to tial units with periods. Lines a to c represent Beth at the drug store counter, his local plan is one sentential unit, line d another, lines e to h to greet her, but only as the initial move in his a third, and so on. Unlike intonation units, larger plan to buy batteries. Most local plans sentential units vary enormously in length. are derived from larger joint activities. Conceptually, they appear to represent a sin- Intonation units, sentential units, and sec- gle center of interest in the larger event being tions are planning units even in dialogues. described (Chafe, 1979, 1980). As noted earlier, people in conversation pro- ceed largely by means of projective pairs. Sections These proposals and uptakes each normally Sentential units, in tum, are strung together to occupy single turns and are often sentential create sections, which correspond roughly to units of one or more intonation units (Ford & ! I I.I I 220 Psycholinguistics ,,I I Thompson, 1996). Jane's "Could you give her just one (as in lines h and i). With two argu- liJi : 11 !'' a message for me?" and Kate's "Certainly" ments, he can mention the bartender and glass ]! •' :i are each single intonation units. Together, a (as in lines c and d), or the beer and glass (as in proposal and its uptake constitute a type of lines e through g). Even without mentioning section, and these can be combined, through the bartender, he can imply the presence of an chaining, embedding, and pre-sequencing, to agent (as in lines e and h) or not (as in lines f, form larger sections. So, although the plan- g, and i). ning units in dialogues look much like those in The propositions expressed are also deter- narratives, they emerge from the participants mined by choice. Instead of bartender, acting together. Burton could have used barman or guy be­ hind the bar. Instead of beer, he could have Perspective used brew or lager or suds. Instead of glass, he could have used stein or schooner. Or he People taking part in joint acllvJ!les must could have added modifiers, as in tall glass, coordinate the content of their actions-the very dirty glass, or glass with a picture of the ideas, beliefs, and assumptions they are pre- President on it. Each choice reflects a different senting. Speakers initiate this process by their perspective. choice of words, phrases, clauses, and ges- Burton must get Charlotte to understand tures. Many of these choices have to do with his perspective. Recall the experiment de- perspective, broadly defined. scribed earlier in which Ann got Ben to ar- Suppose that Burton, who is speaking to range 12 Tangram figures in an order (Schober Charlotte, wants to describe a scene in which & Clark, 1989). Ann and Ben repeated the a bartender filled a glass with beer. Here are task six times with new arrangements of the some of his options: same figures. The first time through, it took Example arguments them many words ( 112 on average) to estab- (6) a. The bartender filled bartender, glass, beer lish a jointly acceptable perspective, as in this the glass with beer. b. The glass was filled example: with beer by the bartender. (7) Ann All right, the next one looks like a person c. The bartender filled bartender, glass who's ice skating, except they're the glass. sticking two arms out in front. d. The glass was filled Ben Uh huh, okay. by the bartender. Ann Got that one? e. The glass was filled beer, glass Ben Yeah. with beer. f. The glass filled Here Ann and Ben agreed on the descrip- with beer. tion "person who's ice skating, ~xcept they're g. The beer filled sticking two arms out in front." (Another pair the glass. h. The glass was filled. ·glass agreed on the description "person dancing" i. The glass filled. for the same figure.) By the sixth time through, ittook Ann and Ben only 16 words on average, I. Propositions as shown here: Burton must choose the propositions he (8) Ann The ice skater wishes to express. To form a clause, he must Ben M-hm. include a verb. As arguments of that verb he can include three arguments (as in lines a and Ann simplified the perspective to "ice skater" b), two arguments (as in lines c through g), or based on the perspective she and Ben had Speaking 221 grounded earlier (see also Chantraine & ground, and new information, which refers to Hupet, 1994; Hupet, Seron, & Chantraine, information not yet part of common ground 1991; Krauss & Weinheimer, 1964, 1966, (Clark& Haviland, 1977; Prince, 1981). Con- 1967). When Ann was given a new partner, sider Burton's two choices: Carl, she had to return to a fuller perspective j. and ground it from the beginning (Wilkes- (6) What the bartender did was bartender, glass fill the glass. Gibbs & Clark, 1992): k. The bartender filled a glass. bartender, glass

(9) Ann All right, the second one looks like a person With 6j, Burton assumes that it is a given- that's ice skating, kind of. They've got a diamond for a head and then they've already common ground-that the bartender gOt two arms sticking out to the right did something, but not what it was. He adds and a leg in back, and a leg- the new information that it was "fill the glass." Carl To the right or to the left? To the- bartender Ann To the left, sorry. With the accent on in 6k, he as- Carl I got it. sumes that it is given that someone filled the glass, but not who it was. He adds the new in- 2. and Predicate formation that it was the bartender. Burton's Even if Burton mentions the bartender and choice of given and new information deter- glass, he must decide which to make the sub- mines not only the of his utterance, ject. In 6c, the bartender is the subject, and but its intonation. what the bartender did is the predicate, but One choice that depends on common in 6d, the glass is the subject, and what hap- ground is the choice between definite and in­ pened to it is the predicate. Many languages definite descriptions. Burton would tell also mark a topic (what the utterance is about) Charlotte, "The bartender is filling the glass and comment (what is said about the topic), for you" if he thought she could infer the iden- but English does not. Normally, the topic in tity of the glass from their common ground English is the subject. (e.g., he had just given her glass to the bar- tender). But he would reply, "The bartender 3. Figure and Ground is filling a glass for you," if he thought she Burton can also choose between saying ''The could not infer its identity. The general rule glass filled with beer" and "The beer filled the is this: Definite descriptions require the ref- glass" (6fandg). In the first, he views the glass erents to be inferable from current common with respect to the beer, treating the glass as ground; indefinite descriptions do not. There- figure and the beer as ground. In the second, fore, Burton can say, "I got in my car and he does the reverse. Using another verb, he has grabbed the steering wheel," and assume the same choice with (a) "The bartender filled Charlotte will infer "The steering wheel be- the glass with beer" versus (a') "The bartender longs to the car." He can also say, "I walked poured beer into the glass" (see Talmy, 2000). into the room; the chandeliers were burning brightly," and she will infer "There are chan- 4. Given and New Information deliers in the room." Inferences like these are Bnrton is exquisitely sensitive to Charlotte's called bridging inferences, and Burton de- state of mind. As we noted earlier, he keeps signs his utterance to make bridging easy, a track of their current common ground and de- point we return to later: signs his utterance against it. Most utterances Speakers choose perspectives as part of lo- divide into given information, which refers to cal plans. Suppose Charlotte asks, "What is information inferable from current common the bartender filling my glass with?" It would 222 Psycholinguistics

be natural for Burton to reply, "Beer," "With appropriate perspective is an important part beer," or perhaps, "He's filling it with beer." It of planning. would be odd to reply, "It's being filled with beer," or "Beer is filling it," or even "It is fill- ing up with beer." The natural replies retain Functional Processing Charlotte's perspective: the propositions, sub- It was once believed that speakers produce ut- ject and predicate, figure and ground, and terances one word at a time by association- given and new information of her question. from left to right. One of the revolutions of The other replies replace her perspective. In the 20th century was to overturn that idea. an experiment by Level! and Kelter (1982), It was replaced by the theory that speakers confederates phoned Dutch merchants and formulate utterances from the whole to its asked the Dutch equivalent of lines lOa, b, parts, from the top down. Speakers begin with c, ord: a message-a selection of propositions un- (lO) a. What time does your shop close? der a particular perspective, or enough mate- b. At what time does your shop close? rial for about one clause. They then proceed c. What time does your shop close, because I in three overlapping stages (Bock & Level!, have to come into town especially for this, you see? 1994; Level!, 1989): (a) functional process- d. At what time does your shop close, because ing, (b) positional processing, and (c) phono- I have to come into town especially for logical . Much of the evidence for this, you see? the top-down view comes ·from a surpris- ing source: slips of the tongue (Dell, 1986; Although the perspectives in lOa and lOb dif- Fromkin, 1971, 1973; Garrett, 1980). fer only slightly ("What time" vs. "At what In functional processing, speakers select time"), the merchants tended to retain that per- the lexical concepts needed for their message spective. They preferred "Five" over "At five" and assign them to grammatical functions ap- for question lOa, and the reverse for JOb. With propriate to their perspective. Suppose Alan the extra clause in lOc and IOd, merchants wants to tell Barbara that Ben has been of- were more likely to give full answers, such as fered a job in engineering. For this message "We close at five," which are appropriate to Alan needs six lexical concepts, roughly, "the either perspective. Retaining a perspective person speaking," "believe," "male person in is the easiest, and therefore expected, thing focus," "officially propose," "technical pro- to do. fession," and "paying position." These lexical One reason for a respondent to change per- concepts, called lemmas, are each associated spective is to take issue with the speaker, as with a word form, or . The lemma "of- in this example: ficially propose," for example, is associated (11) Jim how old, were most of the children,- . with the lexeme offer. Once Alan has formu- Kay well uh only a few of them, were lated each lemma, he must retrieve the corre- children in fact, . urn . I was teaching sponding /, think, he, offer, engineer­ , ing, and job. Many types of slips of the tongue arise at In changing perspective, Kay implies dis- this stage (Bock & Levelt, 1994; Dell, 1986). agreement with Jim's presupposition about Kay's students. The general rule is this: To 1. Semantic Substitutions. One speaker retain a perspective is to presuppose agree- produced "the the Ca - . the the Protestants, ment; one way to imply disagreement is to seem just as bad at this." He intended to acti- change perspective. All in all, selecting the vate the lemma "Protestant religious group," Speaking 223 but instead activated "Catholic religious of collocation substitution is that of chamber­ group," a closely related lemma. Slips like this maid for chamber music. lead to the substitution of one semantically Speakers must assign the lemmas and related word for another, not only Catholic lexemes they select to syntactic functions for Protestant, but in other examples such as (Bock & Levell, 1994). Their message spec- high for low, cherries for grapes, and Chinese ifies the perspective for the current clause for Japanese. (which propositions are to be expressed, what 2. Blends. Another speaker referred to a is subject, , and indirect object, what container "that they swishle swizzle things is figure and what is ground, what is given around in." Apparently, he activated the lem- and what is new), and these detennine the mas for swish and swizzle simultaneously functional assignments. Alan's message spec- (both fit his message) and combined the cor- ifies two main propositions: "x thinks that y," responding two lexemes to form the blend where y is "z offers u to v." He assigns I to the swish/e. Other attested blends include momen­ role x, he to the role v, engineering job to the taneous from momentary and instantaneous, role u, and leaves z unspecified. Also, he as- stougher from stiffer and tougher, and hi/aries signs I to the subject of the main clause, and for hilarity and hysterics. he to the subjectofthe embedded clause. And, 3. Sound-related substitutions. Another Alan detennines that I and he are given infor- speaker said, "because she'd laughed so much mation in focus of attention, therefore mak- she'd burnt a couple burst a couple of stitches." ing them pronouns and that engineering job She selected the lemma for burst, but retrieved is new information, therefore making it in- the sound-related lexeme burnt instead. Other definite. The result is a functional assignment attested examples include sympathy for sym­ something like this: [I think that [he be offered phony, bodies for bottles, and garlic for gargle engiueering job]]. (Bock & Levell, 1994). Certain slips of the tongue can arise in the 4. Tip ofthe tongue. Another speaker said, process of functional assignment: "and can you assess can you . keva- what's 6. Word interchanges. One speaker wanted the word, . connect them ... " Apparently, he to say "writing a letter to my mother," but had selected the lemma for connect, but could said "writing a mother to my letter," exchang- not retrieve the lexeme by the time he needed ing the lexemes mother and letter. Although it. Hence the initial attempt "keva-" followed the speaker retrieved mother and letter, he as- by the comment "what's the word." English signed them to the wrong arguments. He must has special words for use at such moments, as have activated both words at the same time to when another speaker said, "you don't mean be able to exchange them. Words may also be the Hussey thingummy and whatsit." anticipated, as in "the sky is in the sky" (for 5. Collocation substitutions. As it hap- "the sun is in the sky"), or perseverated, as in pened, in our example Alan has trouble "the class will be about discussing the class" retrieving the right lexemes: (for "discussing the test'). The word substi- (12) Alan I think he was offered an engineering tuted almost always has the same form class degree, engineering -job, after the as the intended word (such as noun for noun, first slump, and verb for verb). He is trying to retrieve the stock phrase of en­ 7. Phrasal interchanges. One speaker gineering job for the two lemmas "technical wanted to say "I got into a discussion with this profession" and "paying position," but instead guy," but produced "I got into this guy with a retrieves a stock phrase with the same first discussion." He exchanged not just two words, word, engineering degree. Another example however, but two entire phrases, a discussion 224 Psycholinguistics

and this guy. He must have plauued these placed the lighter constituent "yesterday phrases before inserting them into their appro- morning" (2 words long) first. priate slots in the construction of "I got into x Many other types of slips of the tongue withy." Another speaker intending "they must occur at this stage (Bock & Levell, 1994; Dell, be too tight for you" produced "you might be !986). too tight for them." He must have switched the 8. interchanges. One speaker, lemmas "third person plural" and "first per- intending to say "Singer sewing machine," son" and only then selected the lexemes you produced "Singing sewer machine." He kept and them. Otherwise, he would have produced sing and sew in the right order, but added the "you might be too tight for they." inflections -er and -ing to the wrong stems. Another example is "he go backs to" for "he Positional Processing goes back to." Morpheme accommodation. Once speakers have selected the lexemes and 9. One 1Speak- their functional assignment, they need to order er, intending to say "Mr. Keene, tracer of lost the lexemes for articulation. The first step is to persons," said "Mr. Keene, loseroftracedper­ assemble the lexemes, in their assigned func- sons." At the functional level, he exchanged, tional roles, into constituents. Alan assembles not the words tracer and lost, but the verb I and think into one major constituent, and stems trace and lose. Then, at the positional he, offered, and engineering job into another, level, he added -er to trace to form tracer and and he places them in this order. (If he had made lose into a past participle to form lost. reversed the order, he would have said, "He Speakers also select a or an to fit the word was offered an engineering job, I think.") He that follows it, even if that word is itself in er- then adds the right inflections, making "be + ror, such as the speaker who misproduced "a " into was, and he spells out the func- meeting arathon" for "an eating marathon." tion words, making the indefinite article into 10. Mis-derivations. One speaker produc- an to agree with engineering job. ed "these are oral contraception," another Speakers at this stage are sensitive to "I've just gave given you," and another "he the weight of each constituent. When they think thinks that Ella's worried." These speak- have a choice, they prefer to place heavier ers planned the right words, "contracept + constituents later than lighter ones (Arnold, nominal suffix," "give + past participle," and Wasow, Losongco, & Ginstrom, 2000; "think+ singular," but in deriving the words, Behaghel, 1909/1910; Hawkins, 1994; added the wrong inflections. Wasow, 1997). Consider this example, noting the order of constituents in brackets: Phonological Encoding

(13) the first European conference on astronomy at Once speakers have selected the words, as- Leicester, . reported [yesterday morning], - [on signed them to functional positions, assem- overnight observations of the behaviour of the bled them in the right order, and filled in the object, - . known as A six uhu two one one inflections and function words, they are ready zero], (l.lla.28) to spell out the phonetic segments. They do this, not one intonation unit at a time, but one Ordinarily, the speaker would have said, "A short constituent at a time. Once again, the reported [on some observations] [yesterday evidence comes from slips of the tongue. morning]." But she anticipated that her de- scription of the observations would require 11. Sound interchanges. These include the a heavy constituent (17 words long), so she anticipation of an upconting sound, as in Speaking 225 leading list for reading list, the perseveration Phonological encoding, therefore, works of a previous sound, as in beefneedle for beef one short phrase at a time. It assembles these needle, or Liverpool lullapie for Liverpool phrases according to their phonetic segments, lullaby. The classic "spoonerism" is an ex- syllables, and meter, regardless of what they change of two sounds, as in lork yibrary for mean. And when it makes errors, it makes York Library, speer bill for spill beer, and flow them out of the elements in these plans. The snurries for snow flurries. Speakers can in- final product is a motor program that works terchange consonants (e.g., p and b), vowels the tongue, lips, larynx, jaw, and lungs. There (e.g., ee and oo), consonant clusters (e.g., fl is an analogous process that creates a motor and sn), and what are called the rimes of two program to work the hands, arms, eyes, face, syllables (e.g., -eer and -ill). and torso in gestures. Although less is known about this process, it is linked in both time Generally, speakers produce more antici- and content to the functional, positional, and pations (leading list) than perseverations (beef phonological processes for speech. Speakers' needle). According to a model developed by gestures are closely tied to the content and Dell, Burger, and Svec (1997), this is because timing of the words they use (see the follow- speakers are focused more on the future of ing sections). their speech planning than on the past. When people have to say tongue twisters such as "chef's sooty shoe soles," people tend to per- Primary and Collateral Speech severate more often than anticipate words or People are not automatons. They are normally sounds. After practice, while the overall error aware of what they are doing, able to reflect rate drops, the errors tend to be anticipations on what they have just done and are about rather than perseverations. People who speak to do, and if they don't like what they see, more slowly (e.g., children and people with they change directions. People are no different brain damage) also tend to focus more on the when they are speaking. They normally mon­ past and produce more perseverations. itor what they are about to say and have just Sound interchanges work in remarkably said, and what their addressees are doing and regular ways. The two elements involved al- saying, and if they don't like what they see, most always come from content words (nouns, they change directions (Levell, 1983, 1989). verbs, adjectives, adverbs) and not function Taking actions based on self-awareness adds words (articles, prepositions, etc.). They al- a second track to utterances. The distinction most always come from adjacent words (as in is between primary and collateral signals York Library) or even the same word (as in (Clark, 1996). aminal for animal). They tend to be similar Spontaneous speech is replete with actions phonetically and metrically, and in homolo- not found in idealized speech. The following gous parts of words. The y and l in York Li­ is one example (Svartvik & Quirk, 1980): brary are similar types of consonants-what are called liquids-and both are in the ini- (14) Reynard well,. I mean this. uh Mallet said tial position of accented syllables. Therefore, Mallet was uh said something sound exchanges stand in contrast to word ex- about uh you know he felt it would be a good thing if uhh . if changes. Word exchanges come from homol- Oscar went, ( 1.2.370) ogous locations in phrases and are similar in and function. Sound exchanges come This utterance is full of supplementary from homologous locations in words and are features-repeats ("if uhh if'), repairs similar in sound and meter. ("Mallet said Mallet was"), fillers ("uh"), 226 Psycholinguistics

prolonged syllables ("uhh"), and editing ex- about uh you know he felt. .." Reynard pressions ("I mean," "you know"). These ac- points out that he is changing "said some- tions each appear to reflect a difficulty in de- thing about" to the more accurate "felt." ciding what to say or how to say it. Still, 2 Fillers such as uh and urn (Clark, 1994, they allow Peter, the addressee, to identify 1996; Clark & Fox Tree, 2001). Speakers what Reynard really wants to say. Concep- use these to signal delays in speaking. In tually, Reynard's utterance divides into two "Mallet was uh said something about ..." parts. The primary signals reflect the official Reynard signals a delay with "uh" while business of the conversation at the moment, he rephrases "was ..." to "said somethirig namely: about." 3. Discourse markers such as well, now, oh, (14') Reynard well, Mallet said he felt it would be a good thing if Oscar went like, and so (Fox Tree & Scl:u'ock, 1999; Schiffrin, 1987; Schourup, 1982; Under- The collateral signals are about the on-going hill, 1988). Speakers use these to indicate performance itself. changes in direction and other such things. Supplementary features typically divide With "well" in example 14, Reynard indi- into two types: problems and solutions. Take cates that he isn't giving a direct answer to Reynard's "it would be a good thing if uhh . the question he had been asked. if Oscar went." By the time Reynard reached 4. Back-channel responses or continuers thing, he apparently had a problem-perhaps such as uh-huh, yeah, andm-hm (Goodwin, he didn't quite know what to say next. 1986a; Schegloff, 1982; Yngve, 1970). Peter, his addressee, may have inferred the Speakers use these to acknowledge they problem, but the problem itself remained hid- have heard or understood their partuer well solution den. All Peter heard was Reynard's enough for current purposes. to the problem. Reynard took four actions: 5. Certain gestures, including certain head (a) Before suspending his speech, he produced nods, eye gaze, smiles, grimaces, and if to commit himself to producing an if-clause; pointing (Bavelas, Chovil, Lawrie, & (b) he produced uh to signal that he was de- Wade, 1992; Bavelas & Chovil, 2000; laying the resumption of his speech; (c) he Goodwin, 1981, 1986b; Goodwin & prolonged uh to signal that he was continu- Goodwin, 1986). Speakers use these to ac- ing an ongoing delay; and (d) upon resuming knowledge what is being said and other- speech, he repeated if to restore continuity to wise coordinate with their partners. the if-clause. These actions are each collateral strategic silences and overlaps signals to help Peter deal with the delay with 6. Certain the least effort. (Goodwin, 1981; Schegloff, 1987). Speak- Collateral signals come in many types, ers use these to indicate such things as which have been discovered in the close exam- reluctance or demands to speak. ination of spontaneous speech. These include 7. Nonreducedvowels (such as "thee" instead the following: of "thuh" for the word the) and prolonged syllables (Fox Tree & Clark, 1997). Speak- 1. Editing expressions such as I mean, you ers use these to indicate they are suspend- know, that is, no, and sorry (Erman, 1987; ing speech or adding a delay because of Levell, 1983, 1989). Speakers use these to some problem in production. When one point out expressions they wish to amend speaker said, "when you come to look at and why. In "Mallet was uh said something thee . thuh literature," he signaled that he Listening 227

was having problems deciding on litera­ tings. Still, these investigations have estab- ture, which he immediately amended, "I lished many of the processes by which lis- mean you know the actual statements." tening takes place. 8. Preliminary commitments (Clark & Listeners begin with the raw material they Wasow, 1998). Speakers often produce a hear and see-the speaker's vocalizations and word or phrase on its own to commit them- gestures. They recognize that speakers pro- selves to speaking before they are able to duced these in attempts to advance the current proceed fluently. When Reynard says "if joint activity-whether it was diplomacy or uhh. if' he produces the first if to commit gossip, a business transaction, or a card game. himself to the upcoming if-clause that he So, listeners recognize that these signals must cannot yet produce. satisfy two constraints: (a) They must be con- sistent with the raw material heard and seen; In summary, speaking has many origins and (b) they must contribute to the speakers' and constraints. People speak primarily to ad- moves in their current joint activity. Listen- vance their joint activities-from business ex- ing works both from up and from changes to telling stories. They form plans at purpose down. Early on, most investigations many levels-from sections, sentences, and were on the processes that work from the bot- intonation units down to words, suffixes, and tom up, but more and more have revealed pro- phonetic segments. At the same time, people cesses that work from the top down. monitor what they and their interlocutors are doing and saying. They create not only pri- Identifying Words mary signals for their official business, but collateral signals to deal with the on-going Speech doesn't come parsed into words, performance itself. phrases, clauses, and sentences. Most intona- tion units, the main units of speech identifiable LISTENING from prosody, are uninterrupted streams of speech sounds. "I'll have four please" might For every action in speaking, there must be come off "Illhavefourplease," with no no- a corresponding action in listening. (Com- ticeable gaps. Worse yet, the pronunciation pare Newton's third law of motion: "For every of many words and phrases further obscures action, there is an equal and opposite reac- their boundaries. "In boats" is regularly pro- tion.") Just as speaking divides into four lev- nounced "im.boats" (the period marks a sylla- els, listening does also by: ble boundary), "an egg" as "a.negg," "to eat" as "to.weat," and "the apple" as ''the.yapple." 1. Attending to the speakers' vocalizations When speech is informal and quick, "Why and gestures (channel level); don't you eat?" may sound like "Wain.cheat?" 2. Identifying the speakers' signals (signal Listeners must have a remarkable ability to level); discover order in apparent disorder. 3. Understanding what the speakers mean by Evidence suggests that listeners identify those signals (intention level); and words one speech segment at a time. In one study (Marslen-Wilson & Tyler, 1980), people 4. Considering the joint projects proposed (project level). were asked to listen to speech and, when they heard a specific target word, to press a but- Listening has been investigated mostly at the ton as quickly as possible. Some people, for signal and intention levels in artificial set- example, listened for the target word lead in 228 Psycholinguistics

"The church was broken into last night. Some one experiment, listeners sat at a table with thieves stole most of the lead off the roof." candy and other objects on it and were told Listeners identified the target words, which to "Pick up the candy." Their eyes darted averaged 420 ms long, a mean of 50 ms be- toward the candy even before the end of the fore the ends of the words. If itis assumed that word candy. When there was both candy and a pressing the button takes about 200 ms, then candle on the table, they took longer because listeners identified the word about 250 ms be- the candle delayed the uniqueness point for fore the end of a word-less than half way identifying the word candy (Dahan, Swingley, through. How is that possible? Tanenhaus, & Magnuson, 2000; Tanenhaus & According to a model by Marslen-Wilson Spivey-Knowlton, 1996).· and colleagues (e.g., Marslen-Wilson, 1987), If 'Til have four please" is pronounced listeners begin with the first sound of a word "lllhavefourplease," how do listeners know and then use the succeeding sounds to nar- when to start a new word? They don't. In row down the possibilities until they artive at a study by Shillcock (1990), when listeners a unique word. Take trespass. After the first heard, "He carefully placed the bone on the sound "t-," listeners activate in memory the table," they were primed by bone to identify entire cohort of English words that begin with the word rib. This is not surprising. But when "t-." With over 1,000 such words, each one other listeners heard, "He carefully placed gets only a small activation. After "tr-," lis- the trombone on the table," they were just teners reduce that cohort to words that begin as primed by trombone for the word rib. Lis- with "tr-," which may run into the hundreds. teners apparently hear bone in trombone-at By "tresp-," listeners have reduced the cohort least briefly. In working bottom up, listeners to a unique word, trespass, which gets all the initially activate a wide range of extraneous activation. So the "p" in "trespres" is called words. the uniqueness point. It has been shown that Identifying a word requires not only its the earlier the uniqueness point, the earlier lis- phonological shape, or lexeme-such as "tre- teners can identify the word. spres" or "krepten"-but its intended sense, or Are all of these preliminary words acti- lemma. Most words are ambiguous, so listen- vated in memory? The evidence suggests that ers must select from a range of lemmas. An they are. Consider captain. Before the "e" in example of this is the word bug in "He found "krepten," listeners should activate not only several bugs in the corner of his room." With- captain, but also captive (and other words). out knowing more about what the speaker is As a result, they should be primed to iden- trying to say, bug could equally mean "in- tify words related to both captain and cap­ sect" or "hidden microphone." As Swinney tive, for example ship and guard. But just (1979) showed (see also Tanenhaus, Leiman, after the "e" (the uniqueness point for cap­ & Seidenberg, 1979), people who listened to tain), listeners should activate only captain, this sentence were primed by bug to identify which primes ship, but not guard. Indeed, this words related to both of these meanings-say, is precisely what Zwitserlood (1989) found. ant for "insect" and spy for "hidden micro- Research shows that listeners activate entire phone" (compared to neutral sew). Surpris- cohorts of words, which they reduce to unique ingly, however, other listeners were just as words when they get enough evidence (cf. primed by bug to identify both ant and spy Elman, 1989; McClelland & Elman, 1986). (compared to sew) in an utterance that had They can narrow down the options even faster no ambiguity at all: "He found several spi- by taking note of the potential referents. In ders, roaches, and other bugs in the comer of Listening 229 his room." Listeners were primed for both ant chowder go together, and they form a preposi- and spy immediately after bug, but just a few tiona! phrase. English relies heavily on group- syllables later they were primed for ant, but ing for denoting the relations among words, not for spy. These listeners had quickly de­ so listeners should tty hard to identify con- activated the unintended lemma "hidden mi- stituents. crophone." From this and many other inves- 2. Ordering. "Relations among proposi- tigations, it appears that listeners activate all tions tend to be marked by word order" (see common senses of a word and then deactivate Greenberg, 1963). In English, typical sen- those that don't fit. tences are subject + verb + object (as in Listeners, then, have a dual problem: how "I'll have a bowl of clam chowder"), but in to identify the lexemes within the continuous Japanese, they are subject + object + verb. stream of speech, and how to settle on the in- In addition, "word pairs that are alike in func- tended lemmas. What makes it such a problem tion tend to have the same internal ordering" is that listeners activate too many lexemes and (see also Lehmann, 1972, 1973). In English, too many lemmas. They need powerful top- modifiers tend to come before nouns, as in down methods for settling on the right ones. Russian dressing, clam chowder, that dog, and two hamburgers. If the modifiers are , they tend to come after nouns, as in bowl Sentence Structures of clam chowder, dog that I saw, and ham­ Sentences have an orderliness that listeners burgers good enough to eat. In French and can count on as they try to hear words as Spanish, even simple modifiers tend to fol- parts oflarger structures. Although languages low nouns. Therefore, it is wise for listeners of the world differ, they tend to conform to to attend to word order in order to identify a small number of principles about sentence subjects, verbs, objects, and modifiers. structures. It would be odd if listeners did not 3. Case-marking. "Wordsmarkedforcase exploit these principles, and they do. The next denote distinct roles." In English, I and we section describes four such principles. are used for denoting subjects, whereas me and us are used for objects of verbs, objects l. Grouping. As Behaghel noted over a of prepositions, and other functions. Nomina- century ago, "Wbat belongs together men- tive, accusative, and possessive pronouns have tally is placed together syntactically" (see distinct functions. In German, articles, adjec- Venneman, 1973, 1975). Another way to tives, and some nouns also use case-marking. phrase this claim is that words that jointly The man is translated as der Mann, demMann, refer to the same object, event, or process tend and den Mann for the nominative, dative, and to be placed in a single constituent. In En- accusative cases depending on whether the glish, ''I'll have a bowl of clam chowder and a man is, for example, the subject, indirect ob- salad with Russian dressing" divides into con- ject, or directobject of the verb. It is informa- stituents as follows, where each constituent is tion that German listeners rely on. enclosed in a pair of square brackets: 4. Agreement. "Words that agree (in num-

[I'll [have [[a [bowl [of[clamchowder]]l] and ber, gender, etc.) tend to refer to the same ob- [a [salad [with [Russian dressing]]]]]]] ject, event, or process." In French, the three words in le solei! rand ("the round sun") are Mentally, clam and chowder go together (both each masculine, and those in la tune ronde referto the soup) and, indeed, they form a con- (''the round moon") are feminine. French lis- stituent, a noun phrase. Likewise, ofand clam teners can count on agreement to help them 230 Psycholinguistics identify le, solei/, and rand as referring to the comprehending utterances, so this chapter same object. English makes almost no use of addresses only the basic issues. agreement. Grouping and ordering are exploited by Comprehension Processes English listeners, as simple examples demon- strate. Consider "John said he will come How do English listeners identify syntactic yesterday." One reason this sentence sounds relations? According to some proposals, lis- strange is that listeners try to make a con- teners work largely or solely bottom up. Sup- stituent out of he will come yesterday, and pose people read "The reporter saw her friend that makes no sense. Listeners have trouble ..."one word at a time. If they realize th'at saw seeing that yesterday goes with John said be- most often takes concrete objects, they should cause that would create a discontinuous con- infer that "reporter saw friend" is subject + stituent. verb + object. So when the sentence goes on, Or consider: (I) "John figured out that " ... was not succeeding," they should be star- Susan wanted to take the train to New York" tled at was and recover only after a delay, as versus (2) "John figured that Susan wanted was the case in an experiment by Holmes, to take the train to New York ont." In sen- Stowe, and Cupples (1989). Listeners were tence 1 it is easy to create the verb figure out not startled, however, when the sentence be- because figure and out form a constituent. In gan "The reporter saw that her friend ... " Nor sentence 2 it is difficult to see the verb as fig­ were they startled at would in "The candidate Ure out because the verb is discontinuous and doubted his sincerity would be appreciated." because the train to New York out forms an For this utterance, they apparently assumed interpretable constituent. that the direct object of doubt is most often a Finally, consider "The man pitched the ball full clause, such as "that his sincerity would be threw the ball." As we go along, we form appreciated." Evidence like this suggests that a subject-verb-object constituent of the man listeners know about the constructions that pitched the ball, but then we are left with words are most likely to occur in, and they use the fragment threw the ball. The sentence that knowledge in utterances into seems to make no sense. Change it to The man constituents (MacDonald, Pearlmutter, & thrown the ball pitched the ball, and the prob- Seidenberg, 1994). lem disappears. Thrown cannot be the main Listeners also work top down. Consider verb, so we realize that the man thrown the "The burglar blew open the safe with the new ball is a noun phrase and pitched the ball is lock." The phrase the safe with the new lock the main verb and object. It is easy to see why makes sense if there are two safes, one with a parsing "The horse raced past the bam fell" is new lock and one without. It makes less sense so difficult (Bever, 1970). if there is only one safe. In one part of an ex- Languages differ in how they mark syn- periment by Altmann and Steedman (1988), tactic relations. English makes heavy use people read one of the two versions of this of grouping and ordering, whereas German passage: makes greater use of case-marking and less of ordering. Walpiri, a language of Australia, A burglar broke into a bank carrying some dy- makes heavy use of case-marking and almost namite. He planned to blow open a safe. Once none of ordering. Parsing strategies should inside he saw that there was a safe with a new reflect these differences, and evidellce sug- lock and a safe with an old lock [or: a strong­ gests they do. In this volume, Rayner and box with an old lock]. The burglar blew open Clifton review the processes for parsing and the safe with the new lock.

i I i 1._ Listening 231

People read the phrase with the new lock (Erikson &Mattson, 1981; Fillenbaum, 1971, faster in the version with two safes than in the 1974; Reder & Cleeremans, 1990). They sug- version with a safe and a strongbox. But when gest that people do only a partial analysis of with the new lock was replaced by with some many constructions, cutting the process short dynamite, they were faster with one safe than by introducing plausible interpretations. How with two. Readers used their knowledge of the complex must sentences be for people to take situation already described (one vs. two safes) these shortcuts? Probably, no construction is to help them identify which relation was be- too trivial to ignore. ing introduced by with (a modifier for safe or a modifier for blow up). Implicatures Listeners can also exploit their knowledge of the scene around them. In an experiment Speakers ordinarily mean much more than by Tanerthaus, Spivey-Knowlton, Eberhard, they say. When Jane places a phone call to & Sedivy (1995), people sitting at a table that Miss Pink's office and asks the secretary, "Is had apples, towels, and boxes on it were in- Miss Pink in?" she appears to be asking, lit- structed to (among other things) "Put the apple erally, whether or not Miss Pink is in. But she on the towel in the box." As a sentence, "Put expects the secretary, Kate, to recognize that the apple on the towel in the box" is ambigu- the question is a pre-request. In tenninology ous. Is an apple to go on a towel in a box, or introduced by Grice (1975, 1978, 1991), what is an apple on a towel to go in a box? With- Jane says (in Grice's special sense) is a ques- out , people tend to choose the first tion to be answered yes or no. But by saying grouping: Put [the apple] [on the towel in the that, she also implicates that she wants to talk box]. Indeed, when there was one apple on one to Miss Pink. Indeed, Kate first answers the towel, and a second towel, listeners were con- question and then deals with the implicature, fused. Their eyes darted first to the apple on "Well, she's in, but she's engaged at the mo- the towel, then to the second towel, and only ment." In Grice's view, speakers intend their after a delay did they put the apple in the box. addressees to work out these implicatures as But when there were two apples on the table, part of what is meant. one on a towel and one not on a towel, they Listeners must therefore infer what speak- had no trouble at all. Their eyes immediately ers are implicating. Traditionally, these in- settled on the apple on the towel, and they put ferences have been divided into backward it into the box. They used their knowledge of and forward inferences (Clark, 1977a,b; Clark visual layout to help them parse the utterance & Haviland, 1977; see Garrod & Sanford, as intended: Put [the apple on the towel] [in 1994; Sanford & Garrod, 1994; Singer, 1994; the box]. van den Broek, 1994). Although the two types Top down processes appear to be perva- of inferences have been investigated mostly in sive, but no. one knows how pervasive. For reading artificial narratives, the findings prob- years there was a sign in a London hospi- ably extend to listening as well. tal that read, "No head injury is too trivial to ignore," but no one had noticed that it made Backward Inferences no sense (Wason & Reich, 1979). It was taken In Grice's scheme (see also Sperber & to mean that "You should never ignore a head Wilson, 1986), speakers are expected to fol- injury, no matter how trivial" even though low the maxim: ''Be relevant." They are as-· it literally means "There is no head injury sumed to make their current contribution rel- that is so trivial, so small, that it shouldn't evant to the on-going joint activity. Working be ignored." Examples like this are common out how it is relevant leads to implicatures, as 232 Psycholinguistics in the following two artificial fragments of a stand. Still, there is no unified account of discourse: how they are created (Garrod & Sanford, 1994; Sanford & Garrod, 1994; Singer, 1994; (14) I just bought a shirt and tie at Macy's. The shirt was on sale. van den Broek, 1994). Sometimes, they take (15) I just bought a shirt at Macy's. The price was measurable time to create; other times they just right. do not. Most bridging inferences show up in tests of memory of a passage, but some In each sequence, addressees must determine do not. Two points seem clear: Addressees how the second sentence is relevant to the first, base their bridging inferences on the current and draw the inferences needed to establish joint activity or situation-what they ar~ do- that relevance. Recall that definite references ing with their partners at the moment; and the (such as the shirt and the price) require their bridging inferences they draw are the simplest referents to be inferable from current com- inferences needed to establish the speaker's mon ground. So in 14, addressees infer that utterance as the relevant next move in that the shirt in the second sentence refers to the activity. shirt mentioned in the first. The inference is trivial, but essential to establishing what the Forward Inferences speaker means. In 15, the inference is more In Grice's scheme, speakers are also expected complex. Addressees infer that the shirt was to adhere to two other maxims: (a) "Make your bought for a price, which is the referent for contribution as informative as is required (for the price in the second sentence. the purposes of the exchange)" but "no more Inferences needed for establishing rele- informative than is required," and (b) "Be brief vance or coherence have been called bridg­ (avoid unnecessary prolixity)." What follows ing inferences (Clark & Haviland, 1977) and is a breathtaking variety of implicatures. To accommodation (Lewis, 1979). Bridging in- give an idea of their range, we present three ferences take many forms, as the following heuristics that follow from the maxims, as sequences illustrate (see also Clark, 1977a; characterized by Levinson (2000): Mann & Thompson, 1986; Prince, 1981; Singer & Halldorson, 1996; Sperber & Heuristic 1. "What isn't said, isn't." Wilson, 1986): When Ann is asked "How many children do (16) I went for a walk this afternoon. The park was you have?" and she answers, "I have two beautiful. sons," she implicates that these are all of her [Bridge: One place where I walked was a park, the referent of the park.] children. If she had had others, she would have ( 17) Duncan has a black eye. It was Bob who hit him. mentioned them. [Bridge: Duncan has a black eye because someone hit him.] Heuristic 2. "What is simply described (18) Margaret went horseback riding last week. She is stereotypically exemplified." When Charles was sore for three days. [Bridge: Margaret was sore in the way riders says, "The accountant dried her hands," listen- get sore because of the ride.] ers take him as implicating that she dried her (19) They're having a party again next door. I couldn't hands in the ordinary way and not, say, on find a parking place. her dress. What is ordinary, or stereotypical, [Bridge: I believe they're having the party depends on the situation. At the dinner table, because I couldn't find parking.] the accountant might be expected to dry her As these examples show, bridging inferences hands on a napkin, but in the washroom, to are part and parcel of what people under- dry them on a towel. Listening 233

Implicatures based on this heuristic are Depending on the participants' common often called elaborative inferences (Garrod ground, the stereotypical means of transporta- & Sanford, 1994; Sanford & Garrod, 1994; tion could be a car, bus, train, or airplane, and Singer, 1994; van den Broek, 1994), and they the backward inference would be more work. have been widely studied in comprehension Elaborative inferences like this are essential and memory. In one experiment, people who to narratives. heard "The man dropped the delicate glass pitcher on the floor" often misrecognized it Heuristic 3. "What's said in an abnor- later as "The man broke the delicate glass mal way, isn't normal." When Michael says, pitcher on the floor" (Johnson, Bransford, & "Susan stopped the car," he implicates that she Solomon; 1973). In another study, people who stopped it in the stereotypical way-by us- had just read "Steve threw a delicate porcelain ing the foot brake. But when he says, "Susan vase against the wall" were able to name (read caused the car to stop," he selects the wording aloud) the word break faster than people who cause to stop over the expected stop. By doing had just read "Steve went out and purchased so, he implicates that Susan's method was not a delicate porcelain vase" (Murray, Klin, & normal; for example, she may have used the Myers, 1993). On the other hand, when peo- emergency brake. ple were presented with the sentence "The di- Many of the implicatures created by rector and the cameraman were ready to start heuristic 3 have been investigated as instances shooting when suddenly the actress fell from of indirection. Here, again, is a pre-request: the 14th floor," they were primed for the word dead only after a delay (McKoon & Ratcliff, (23) Susan (on telephone) Can you tell me what time you close? 1992). Store manager Yes, we close at nine. Elaborative inferences often anticipate backward inferences that will be needed later, Instead of asking, "What time do you close?'' as in the following sequences: Susan went out of her way to create a pre- request. Why go to all that work? The answer, (20) Keith took his car to London. The car kept according to many, is to be polite (Brown overheating. & Levinson, 1987; Clark & Schunk, 1980; [Bridge: The car mentioned in the first sentence is the referent of the car.] Goffman, 1967; Lakoff, 1973). It is polite (a) (21) Keith drove to London. The car kept overheating. to offer the addressee a way out of the request, [Bridge: Keith drove a car, the referent of and (b) to add to the addressee's self-regard. the car.] Both help the speaker and addressee main- tain face. So speakers set up pre-requests, like In one study (Garrod & Sanford, 1982), it took Susan's, to deal with the greatest obstacles no longer to read "The car kept overheating" to compliance, and addressees infer this in sentence 21 than in 20. Apparently, when (Francik & Clark, 1985; Gibbs, 1986). In this participants read "Keith drove to London," example, Susan pretends that the greatest ob- they inferred the stereotypical vehicle, a car; stacle is the manager's ability to tell her the therefore it was as easy to draw the bridging closing time. It would be odd to pretend that inference as when the car was mentioned ex- it was his happenstance knowledge: "Do you plicitly. But what if drove is replaced by went? happen to know what time you close?'' Ad- dressees often infer the point of pre-sequences (22) Keith went to London. The car kept overheating. [Bridge: Keith drove a car, the referent of without extra time or apparent effort (Gibbs, the car.] 1979, 1983; Gibbs & Mueller, 1988). l

234 Psycholinguistics

In summary,listeners seem to work bottom a specification of the relation between the sig- up. They identify speech sounds and use them nal and the world. The word battery, for exam- to identify words, then phrases, and then entire ple, can mean "'an artillery emplacement," "an intonation units. They use the successive seg- array of objects," or "a device for producing ments of a sound stream to narrow down on the direct current." These are its type meanings. intended word, activating the lemmas, or word Alan used it at the drug store counter with the senses, of all the potential words at any mo- token meaning "a device for producing direct ment. But listeners also work top down. They current." Winking, gazing, nodding, smiling, are normally engaged in a joint activity (e.g., pointing, and others signals have meanings listening to a narrative, answering a ques- too. What are their meanings, and how do they tion, talking about a scene) and that allows acquire those meanings? them to narrow in on intended words more The late-19th-century philosopher Charles quickly, eliminate inappropriate lemmas, and Sanders Peirce offered one influential answer parse utterances into their appropriate parts. in his theory of signs (Buchler, 1940). Accord- Also, they have procedures for inferring what ing to Pierce, signs represent "objects" (phys- speakers mean. Some of these lead to bridging ical things, actions, events, properties) under inferences that establish reference and coher- certain interpretations. A portrait of Napoleon ence with what has come before. Others lead is a sign that represents a particular man un- to elaborative inferences and inferences about der the interpretation "Napoleon Bonaparte." indirection. Signs, in tum, come in three types:

1. Icons. Icons represent their objects by MEANING AND SIGNALS means of a perceptual resemblance to the objects. Napoleon's portrait represents Speakers mean things by what they say. When Napoleon by its perceptual resemblance to Alan asks Beth (at the drug store counter), Napoleon. "Do you uh have size C flashlight batteries," 2. Indexes. Indexesorepresent their objects by he means that she is to say whether the store means of a physical or causal connection has size C flashlight batteries available for to those objects. A road sign represents a sale to him at that moment. This is what is village by pointing to the village. speakers' meaning called (Grice, 1957, 1968, 3. Symbols. Symbols represent their objects 1991). Speakers' meaning is a type of inten- by means of rules. Both dog and chien sig- tion (an intention that speakers intend their nify domesticated canines, one by a rule of addressees to recognize), and it arises from English and the other by a rule of French. what the speakers are trying to accomplish in the current joint activity. Speakers get their What, then, about signals? Speakers addressees to recognize these intentions by make signals by using, creating. or form- speaking, winking, gazing, nodding, smiling, ing signs for their addressees. In spontaneous pointing, and making other gestures. These speech, speakers have three basic methods of actions are signals, or actions by which one signaling: person means something for others. 1. Demonstrating. Demonstrating is signal- ing by means of icons. When Alan shapes Methods of Signaling his hand like a telephone and places it to his The meaning of a signal is very different from ear (forming an icon), he is demonstrating a speaker's meaning. It is not an intention, but the act of telephoning. Meaning and Signals 235

2. Indicating. Indicating is signaling by ing of the word, its lemma. The lexical entry means of indexes. When Alan points at links the two parts. The lexical entry for dog a car (forming an index), he is indicating is a pairing of lexeme and lemma: [/dog/, the car. "domesticated canine"] 3. Describing~as. Describing-as is signaling The notion of mental raises anum- by means of symbols. When Alan gives a ber of issues for psycholinguistics. We con- "thumbs-up" (a symbol) or says "Great" sider four of them: (1) conventions, (2) lexical (another symbol) to a tennis serve, he is items, (3) communal , and (4) sym- describing the serve as excellent. bolic gestures.

Most signals are composite signals, which 1. Conventions are fusions of two or more of these methods. Researchers ordinarily assume that language When Alan points at a car and says, "that car," is conventional-in particular, that words are he is referring to the car with a single signal, conventional. But what is a convention? The but the signal is a composite of indicating the answer is often treated as self-evident, but it car and describing it as a car. is not. The issue is central to the notion of Most work has focused on symbols be- . cause those are what researchers generally The modern analysis of conventions comes think of as "language." Traditional linguis- from David Lewis (1969). As Lewis argues, tics includes the study of , morphol- people, such as Alan and Beth, have to coordi- ogy, syntax, , and , all nate with each other to reach a common goal. of which are primarily symbols. But face-to- They face a coordination problem in reaching face conversation relies on symbols, indexes, that goal. Suppose they want to greet each and icons in both linguistic and nonlinguistic other. Should they hug, kiss, shake hands, methods. or what? The first time they meet, they may solve the problem by agreeing to shake hands. Agreeing to shake hands is a coordination Describing device-a solution to their coordination prob- Tbe prototypical symbols are words ani! the lem. If they meet regularly, they have a sentences created from them. Whenever peo- recurrent coordination problem for which ple select words and create sentences, they they need a general solution. They may come are using symbols, by describing something to mutually expect to shake hands, and shak- as something. How do these symbols work? ing hands becomes a convention. For Lewis For the past thirty years, most accounts (though the wording is ours): of language use have assumed that people A convention is: possess mental lexicons, or dictionaries in (a) a regularity in behavior the he~d. A mental lexicon is an organized ·list of dictionary entries, called lexical en­ (b) that is in common ground in a given tries, to which people refer when producing community and comprehending utterances (see Dell & (c) as a coordination device O'Seaghdha, 1991; Levelt, 1989; Leveltetal., (d) that is partly arbitrary 1991; Levell, Roelofs, &Meyer, 1999). As we (e) for a recurrent coordination problem. discuss earlier in the chapter, each lexical en- try has two parts: (a) the phonological form Shaking hands is (a) a regularity in behavior. It of the word, its lexeme; and (b) the mean- is (b) common ground for Alan and Beth (c) as 236 Psycholinguistics a coordination device (e) for the recurrent co- canine. The word dog is, therefore, conven- ordination problem of greeting each other. It tional. This solution is partly arbitrary, be- is (d) partly arbitrary because, with a different cause if English history had been different, history, they might expect to hug instead. we might be using /wund, chien, or perro in- Most conventions evolve slowly and are stead. The mental lexicon is a system of such learned as part of one's culmre, but, in the conventions organized into lexical items and right circumstances, they can also develop corrununallexicons. quickly. In a study by Garrod and Doherty (1994; also Garrod & Anderson, 1987), pairs 2. Lexicalltems of people sat at separate computer termi- In corrunon parlance, most words have more nals and tried to negotiate their way through than one sense. The word ear has at least three: mazes on their screens. Although they had Sense 1. The visible organ of hearing, as in the same underlying maze (an incomplete ma- "floppy ears"; trix), they were shown different elements of it. To succeed, they had to exchange infor- Sense 2. The sense of hearing, as in "good mation, which made them coordinate on how ear for jazz"; they talked about locations. One pair might Sense 3. The spoke from which com grows, refer to a location as "four lines down and two as in "three ears of corn." boxes over," (using lines and boxes), but an- But how many ''words" do these represent? other might say, "row four column two" (using Let us consider three models. In Model A, rows and columns). In one condition, pairs of there are three distinct words (ear) that just people played with each other multiple times. happen to sound the same. This model treats In another, people were grouped into an in- senses 1 and 2 as unrelated, and that seems formal community, and each played as many wrong. In Model B, there is just one word ear, times as in the first condition, but once with which has three senses. This model also seems every other member of the community. The wrong because it misses the fact that sense 3 isolated pairs developed local agreements for is conceptually unrelated to senses I and 2. referring to location, but each pair tended to In Model C, there are two words or lexical develop a different one. In contrast, the pairs items, one for senses 1 and 2, and a second for in the community began with different local sense 3: agreements, but soon converged on the same solution-typically rows and columns. The ear1: [fir/, "the visible organ of hearing"] convention evolved as a solution to the recur- [fir/, "the sense of hearing"] rent, community-wide coordination problem. earz: [fir/, "the spike on which com Conventions, Lewis (1969) argued, are the grows"] basis for natnral languages. In talking, Alan and Beth have the recurrent coordination In this view, a lexical item is a collection of re- problem of how Alan is to get Beth to see that lated lexical entries. Indeed, most dictionaries he is denoting a domesticated canine. They of English divide ear into just these parts. recognize that they are both members of the Model C reflects a difference between pol­ community of English speakers in which it is ysemy and homonymy. Ear1 is polysemous common ground that dog can be used to de- because it has more than one related lexical note such a beast. They can solve their coor- entry. But ear is also a homonym because it dination problem by Alan using dog and Beth has two unrelated sets of lexical entries, rep- interpreting him as denoting a domesticated resented by ear1 and ear2. It is often easy to Meaning and Signals 237 identify homonyms by examining other lan- to be created from the known lexical entries guages. French, for example, has different for New York, Napoleon, and size 10. words for ear1 and ear2-orielle and ipi­ In the right circumstances, it takes listeners but like English, orielle has the two senses of no longer to interpret novel words than con- English ear1• ventional words. In one study (Gerrig l989b; How do we decide whether or not "visi- see also Gerrig & Bortfeld, 1999; Gerrig & ble organ of hearing" and "sense of hearing" Gibbs, 1988), readers were given brief sto- belong to distinct lexical entries of ear1? The ries that ended with a noun compound like answer isn't simple. Line, for example, has snow-ball or fire-ball. Readers were faster to ,,, five apparently distinct senses: read and understand snow-ball, a compound il they could access quickly, when the story led II Sense 1. A physical mark, as in 'Two parallel up to its conventional meaning ("ball made of lines never meet"; snow") than when itled up to a novel meaning I, I Sense 2. A demarcation, as in "His car was ("dance in honor of a big snowstorm"). For a checked at the state line"; compound like fire-ball, whose conventional .']:! Sense 3. A continuous arrangement, as in meaning could not be accessed as quickly, 'I:;1 "We stood in line for the tickets"; readers were just as fast in reading and under- ·i! Sense 4. A continuous sequence of words, as standing it with the novel meaning ("dance in in "The actress learned her lines"; honor of a famous fire") as with the conven- Sense 5. A sequence of constructs, as m tional one ("ball made of fire"). People appear "What line of work are you in?" able to access conventional meanings at the same time as they create novel meanings, and In a study by Caramazza and Grober (!976), the novel meanings sometimes arrive before people judged sense l to be the most central the conventional ones. sense of line and sense 5 the least central. The line between conventional and inno- From these and other judgments, Caramazza vative is difficult to draw. At the conventional and Grober argued that line has a core mean- end, we have "to fly to Amsterdam," and at ing, "an extension," and the five senses are the innovative end, "to KLM to Amsterdam." derived from it. But do these five senses rep- But look at these exampl~s of the word news­ resent five distinct lexical entries, each with a paper: "The newspaper is on the table" (the different lemma? Or is there just one lexical physical newspaper); ''The newspaper says entry with the lemma "an extension"? it's going to rain today" (an article in today's The issue is one of sense selection versus edition of a newspaper); "I used to work for the sense creation (Clark & Clark, 1977; Clark, newspaper" (the publishing company); "The 1983; Clark & Gerrig, 1983; Rapp & Gerrig, newspaper called me today for an interview" 1999). People invent new senses every day, as (someone who works for the publishing com- these attested examples show: "The initiative . parry); and "I stopped by the newspaper for is aimed at preventing the New Yorking of the my interview" (the office of the newspaper San Francisco SkYline"; "The photographer company). The list begins at the conventional I' asked him to do a Napoleon for the camera"; end, but is the last sense of newspaper con- and "We're looking for a size 10 with a steam ventional, or do we create it on the spot (as iron" (a female roommate who wears size 10 we do for to KLM)? and owns a steam iron). The words New York, Lexical entries are therefore organized into Napoleon, and size 10 do not come with the lexical items (like ear1 and ear2), which needed lexical entries. The novel senses had futther organize themselves-if line is any 238 Psycholinguistics indication. But when people use a word, they they need to establish joint membership in a often treat one of its conventional lexical en- community and use its lexicon. When Alan, tries as a starting point for creating a novel an American, steps off the plane in Tokyo, sense for that occasion-a nonce sense. It he ntight approach Yuko, a stranger, arid ask, is the only way to interpret New Yorking, "Do you speak English?" If she says, "Yes," a Napoleon, size 10, KLM, newspaper, and the two of them can assume joint membership many other such expressions. in the community of English speakers. Still, he cannot go on to "My heart has an infarct" 3. Communal Lexicons without establislting that both are English- In Lewis's scheme, a convention holds only speaking doctors. When people first meet, for a particular community of people. Most they generally spend time establishing com- accounts assume a single community for mon ground for further conversation. That in- the entire English lexicon-the community cludes joint membership in communities of of English speakers. That cannot be right. shared expertise. For example, the words sclerotic and my­ 4. Symbolic Gestures ocardial are in common ground for med- ical doctors, like the words mortmain and Words and constructions are not the only sym- nonfeasance for lawyers. They are common bols oflanguage use. There is also a class of ground only within these communities of gestures called emblems (Ekman & Friesen, shared expertise-medicine and law. If so, 1969; McNeill, 1992). For North Americans, lexical entries are organized, not into a single these include: thumbs-up, thumbs-down, monolithic lexicon, but into many communal greeting wave, farewell wave, thumb and in- lexicons (Clark, 1998). The largest lexicons dex finger in circle ("excellent"), winks, index reflect shared expertise in a language like finger to protruding lips ("be quiet"), crossed English or Japanese. The smallest reflect es- fingers, and shoulder shrugs (see Johnson, oteric types of expertise like contract law, Ekman, & Friesen, 1975). The two most com- lacrosse, or Palo Alto. mon are head-nods ("yes") and head-shakes Almost every community has evolved a ("no"). Most emblems are not used as con- lexicon for its shared expertise, and Lewis' stituents of spoken utterances, but on their account of conventions makes it easy to see own. Most correspond to one-word interjec- why. Conventions arise as solutions to recur­ tions such as yes, no, okay, hello, goodbye, ex­ rent coordination problems. Most of us have cellent, or quiet, or to simple sentences such little need (especially a recurrent need) to as "I'm kidding" or "I don't know." refer to the notion of "tissue death." But doc- Emblems have conventional meanings and tors do, so they have evolved the term infarct. are, in certain other respects, like words. The As a community, they find it a useful term. same gesture (e.g., crossed-fingers) means Most of us, even after being introduced to the radically different things from one commu- term, do not have the expertise or background nity to the next (Morris, Collett, Marsh, & to use it. Doctors, as a community, do, so they O'Shaughnessy, 1979). Many emblems be- find it a usable term. For a word to arise in a long to highly specialized communities. In community, it must be both useful and usable. baseball, an umpire sticking his right thumb It is these twin requirements that lead to the behind his head means "You're out." So em- size and number of communal lexicons. blems must have lexical-like entries that link Communal lexicons are essential to speak- form and lemma, such as [head-nod, "yes"] ing and listening. For two people to talk, they . and [wink, "I'm kidding"], and. that belong to must use the same vocabulary, and to find one, communal lexicons. Sign languages such as Meaning and Signals 239

American Sign Language are complete lan- follow the bearing of his finger. To indicate, guages built on emblem-like gestures. speakers can use any device that directs their The process of describing, therefore, addressee's attention to the referent. works with symbols, or signs associated with objects by rule. The most basic symbols, 1. Parts of the body. Speakers can point with words and emblems, have conventional lex- the finger (Alan's "Have you ever read ical entries, such as [/dog I, domesticated ca- that?"), sweep over an ·area with the arm nine] and [wink, "I'm kidding"]. These are ("All this is yours"), nod at a thing ("She organized into lexical items, which are orga- was standing over there"), touch or tap on nized into communal lexicons. In speaking a thing with 'the hand or foot ("This is the and listening, people must do more than match book [or rug] I want"), tum the head or the correct lemma with the correct lexeme. torso toward a person ("Let us talk"), and They must establish and use joint communal gaze at a person C'I want you and you and lexicons. Often, too, they must create or inter- you to come with me"). In some societies, pret novel words, deriving or inferring nonce speakers conventionally point with pursed meanings from conventional meanings. lips or a protruding upper lip. 2. Voice. Speakers can indicate their locations by the source of their voices ("I'm over Indicating here"), and their identities by the sounds of Indicating is a method of signaling by which their voices (on the telephone: "It's me"). people create indexes for the objects to which Speakers can indicate points in time by , they want to refer. The prototype is point- the timing of their vocalizations (race of- ing with the finger (index, in Latin, means ficial: "Ready ... set ... go"). Most inter- "finger"), which is often called a deictic ges­ pretations of I, here, and now, the so-called ture. In a bookstore, Alan points at a copy of essential indexicals (Perry, 1979), rely on Melville's Moby Dick and asks Beth, "Have this form of indicating. you ever read that?" His pointing (a) speci- 3. Conspicuous events. When Alan and Beth fies a location, and (b) gets Beth to attend to hear an unexpected noise, Alan can ask, a thing at that location. There is an intrinsic "What is that?" Or, when Beth says, "I'd connection between his gesture, the aiming like a bowl. of vichyssoise," he can ask, of his finger, and the thing itself. In Peirce's "How do you spell that?" with confidence scheme for indexes, indicating requires an ad- that Beth will see that he is referring to the ditional step-an interpretation. Every indi- most conspicuous unspellable word in her cation refers to a thing under a particular de­ utterance. scription. For Alan's referent, the description 4, Appendages. People can point with wood- is "something Beth may have read." en or laser pointers, using them as exten- Indicating seems so simple that there is sions of their arms-as appendages. They nothing to explain. But appearances belie can also direct attention by ringing a door- reality. The following sections describe sev- bell, or by telephoning or paging someone. eral complications. Most forms of directing-to are parts of Directing Attention composite signals. When Alan says, "It's me" Pointing with the finger is a technique we refer on the telephone, he refers to himself with his to in this chapter as directing-to. When Alan voice, which is an index to himself, plus the points at a copy of Moby Dick, he is directing conventional word me, a symbol that refers to Beth's attention to the book by getting her to the person indexed by the voice. Alan's "me" .l

240 Psycholinguistics

is a composite indication-plus-description, as orient and place themselves, or simply speak are the other examples of directing-to. up (for indicating/, here, and now). To indi- cate and to understand indicating, people must Placing-for consult mental representations of the space Another technique for indicating things is around them, objects in that space, and things by placing them for others (Clark, in press). physically or causally connected with those When Alan places money on the ticket counter objects. of a cinema for the ticket -seller to take, he is indicating the money as "payment for a Demonstrating ticket he is buying." And when the ticket- In demonstrating an object (a thing, event, seller places the ticket on the counter for him, state, or property), people create an icon that she is indicating the ticket as "the ticket he is resembles it perceptually. A demonstration now buying." People can also indicate them- is really a selective depiction of the object selves by placing themselves for others. When (Clark, 1996; Clark & Gerrig, 1990), and most a waiter places a bowl of vichyssoise in front of are created by two depletive techniques. Beth, he is indicating it as "what she ordered." But when he places himself next to Beth, he 1. Modeling. Alan can denote a telephone by is indicating himself as "a waiter waiting for forming his right hand into the shape of her order." a telephone (making a fist with thumb and pinkie extended). This could also be called Interpreting Indications sculpting. Another form of modeling is Interpreting even the simplest indication is enacting, as when Alan denotes a person complex. When Alan points at the copy of jumping by playing the role himself and Moby Dick and asks, "Have you ever read jumping. that?" he is drawing Beth's attention to a per­ 2. Sketching. Alan denotes a round plate by ceptually conspicuous site. Yet he is refening drawing a circle in the air with his fin- not to the site itself, or to the physical book, ger. This form of sketching could be called but to any printed edition of Moby Dick. His tracing. Another form is delimiting, as reference takes a chain of indexes: (a) his fin- when Alan denotes the length of a fish by ger is an index to the site of the book; (b) the placing the flat palms of his hands the right site is an index to that copy of M oby Dick; distance apart. and (c) that copy is an index to any edition of MobyDick. These techniques are often used in combina- A major challenge is to say how Alan de- tion. Alan can denote a person telephoning by signs his composite signal-"that" plus his forming a telephone with his hand (sculpting) pointing-and how Beth creates the right and placing it to his ear (enacting). Speak- chain of indexes. With the same gesture, ers can use these techniques with their hands, he could have referred to the physical book bodies, faces, and voices. ("That is tom"), the intangible story of Moby Demonstrations with the hands or arms are Dick ("That is such an exciting novel"), called iconic gestures, or illustrators (Ekman Herman Melville ("He was born in 1819 in & Friesen, 1969; Goodwin, 1981; Kendon, New York City"), or even the publisher ("They 1980; McNeill, 1992; Schegloff, 1984). They publish such great novels"). Each requires a have three main stages: (I) preparation; different chain of indexes. (2) stroke, the meaningful portion of the ges- In face-to-face conversation, indicating is ture; and (3) recovery. They can be timed everywhere, as people point, place things, very precisely with speech, and are generally Meaning and Signals 241 associated with intonation units, the apex of the quotation "I'm gonna curl" is the direct the stroke coinciding with an accented syl- object of the verb say, and the sentence lable. Iconic gestures tend to begin about would not be complete without it. 1.0 seconds before the words they go with 2. Composite parts. Many demonstrations and to last about 1.5 seconds beyond them are parts of composite signals. For ex- (Butterworth & Beattie. 1978; Morrel-Samuels ample, when Alan points at a book and & Krauss, 1992). People can also demon- says, "Have you ever read that?" he creates strate with their faces, as they model sympa- a composite signal-a description (that) thetic grimaces, disappointed faces, or think- plus indication (pointing). The same is true ing faces (Bavelas, Black, Lemery, & Mullett, for many iconic gestures. When a woman, 1986; Goodwin & Goodwin, 1986). Fran, was telling a story about the film Most demonstrations with the voice come Some Like it Hot, she extended her arms in the form of quotations (Clark & Gerrig, overhead while saying, "and the girl jumps 1990). In the following example, Kate is up" (Kendon, 1980). She created a com- telling friends about being in the hospital on posite of a demonstration (the gesture) plus an intravenous (I-V) system (Polanyi, 1989): a description ("jumps up"). 3. Independent signals. Some demonstra- (24) I went out of my mind and I just screamed. I said, "Take that out! That's not for me!" ... And I tions are independent of the utterance or shook this 1-V and I said, "I'm on an 1-V, discourse being produced. They are nei- I can't eat. Take it out of here!" ther embedded nor composite parts. 4. Self-talk. Some demonstrations are not In her two quotations Kate does more than performed for the addressees, but for the say the words. She enacts an angry person by speakers themselves. When solving a prob- shouting the words and pretending to shake lem by themselves, people sometimes an I-V. Therefore, although some quotations gesture to help them think about objects, enact merely what someone said, many enact events, and relations in the problem. what someone did. Most quotations are not verbatim, nor are they intended to be (Tannen, There has been much debate about the 1989; Wade & Clark, 1993). They can be cre- communicative role of demonstrations. In one ated even for speechless entities, as the follow- view (Krauss, Morrel-Samuels, & Colasante, ing example shows (Clark & Gerrig, 1990): 1991; Rauscher, Krauss, & Chen, 1996; Rime & Schiaratura, 1991), most iconic ges- (25) The problem is this [the speaker holds up ring tures are self-talk. Speakers are more likely finger] will say, "I'm gonna curl," and then this guy [the pinkie] will say, "Yeah, I'm gonna to gesture when they have difficulty retriev- curl too!" But then it goes ''Aaaaaigh!" ing a word (Morrel-Samuels & Krauss, 1992), and they are more likely to be hindered in de- .Some quotations are all gesture, as in "The scribing scenes when their hands are immo- kid went [rude gesture] and ran away," where bilized (Bilous, 1992; Krauss, 1998; Rime, the gesture is a type of quotation. Schiaratura, Hupet, & Ghysselinckx, 1984). Most demonstrations are performed during This evidence favors a self-directed role for speech, yet they can bear several relations to iconic gestures. the speech: Still, most iconic gestures probably are communicative (Kendon, 1987, 1994). All 1. Embedded parts. Quotations, whether quotations, whether speech, gestures, or a speech or gestures, are embedded parts of combination, are embedded parts of utter- utterances or the discourse. In sentence 25, ances, so they are communicative (see also 242 Psycholinguistics

Kita, 1997). Most iconic gestures carry infor- REPRESENTATIONS OF DISCOURSE mation not carried in the associated words, and listeners register this information as part People carrying out basic joint activities need of what is connnunicated (Engle, 1998, 2000; to represent those activities and update their McNeill, 1992). Speakers use few iconic ges- representations as they go along. When Alan tures when their addressees cannot see them, buys batteries from Beth at the drug store treating most of the gestures as being for their counter, the two of them start with their initial addressees (Cohen & Harrison, 1973). Speak- connnon ground and update the current state ers use as many iconic gestures in retelling a of their activity as they proceed. These two story as they do in telling it for the first time. representations, the initial common ground They do so even though they no longer have and the current state of the activity, have trouble retrieving words (Beattie & Coughlan, been investigated under such headings as sit- 1998; see also 1999; Beattie & Shovelton, uational models, mental models, scripts, and 1999). Additionally, narrators tell better sto- schemas. The trouble is that most investiga- ries when their addressees can react with tions have focused on narratives, especially iconic gestures, which both parties treat as part written ones, so the picture is incomplete of their communication (Bavelas, Coates, & at best. Johuson, 2000). In sunnnary, signals mean what they do by a combination of three methods: describing- Visual and Spatial Representations as, indicating, and demonstrating. Describ- When people communicate, they not only ing -as is an immense memory retrieval describe, but also indicate and demonstrate. process. Speakers and listeners have up to As we just noted, indicating requires speak- I 00,000 lexical entries in their mental lexi- ers and listeners to represent the surrounding cons, which they consult about five times a space and the objects in it, and demonstrating second in the course of a normal utterance. requires them to represent what things look, Indicating, in contrast, is a process of spa­ sound, and feel like. The very act of connnuni- tial . For each indication, speakers cating demands that people create and update and addressees must consult representations the visual and spatial representations of what of the space around them, locate objects in they are discussing. A large body of evidence it, and find connections among them. Demon- shows that they do. strating, finally, is a process of depicting and imagining appearances. With each demon- Spatial Relations stration, speakers and their addressees must call on their knowledge of what things look Bransford, Barclay, and Franks (1972) pro- o.r sound like and imagine a thing from the duced a classic demonstration of spatial rela- features of the demonstration. tions. As part of their experiment, people read Speaking and listening, therefore, cannot either sentence 26a, b, c, or d, and were asked be reduced to words, phrases, and sentences. to remember it. A close look at any face-to-face conversation reveals it to be an intricate mix of describing- (26) a. Three turtles rested on a floating log and a fish swam beneath it. as, indicating, and demonstrating, not only b. Three turtles rested on a floating log and a fish with language (e.g., words, speech timing, and swam beneath them. quotations), but also with gestures (e.g., em- c. Three turtles rested beside a floating log and a fish swam beneath it. blems, pointing or placing, and iconic ges- d. Three turtles rested beside ·a floating log and a tures). People talk with their entire body. fish swam beneath them. Representations of Discourse 243

The scenes described in 26a and 26b are These findings shouldn't be surprising, and alike spatially, for if a fish swam beneath the they are just a sample of a large literature log ("it"), it also swam beneath the turtles on such effects. They remind us that listeners ("them"). However, the scenes described in need visual and spatial imagination for even 26c and 26d are not alike spatially, for if a fish the simplest descriptions. They need to imag- swam beneath the log, it did not necessarily ine the appearance or arrangement of turtles, swim beneath the turtles. Later, in a multiple- logs, tractors, mice, and fences to come to the choice test with the four sentences in random right interpretations. order, people who had seen 26a often chose 26b by mistake, but those who had seen 26c Point of View rarely chose 26d by mistake. Readers must have represented not the sentence per se, but Most stories are told from a narrator's or pro- a visual ·or spatial representation of the scene tagonist's point of view. In Mark Twain's Tom described. Sawyer, Tom Sawyer is the protagonist, and a People consult visual and spatial represen- separate third-person narrator tracks his lo- tations to interpret even single words, such as cation as he moves from place to place. In approach in these three descriptions: Mark Twain's Huckleberry Finn, Huck Finn is both protagonist and first-person narrator, (27) a. I am standing on the porch of a farmhouse so when he moves from place to place, he looking across the yard at a picket fence. A tractor [or: mouse] is just approaching it. describes what he sees from his own perspec- b. I am standing across the street from a post tive. To represent point of view, readers must office with a mailbox in front of it. A man represent Tom's and Huck's immediate sur- crossing the street is just approaching the post roundings and their location in it. We surely office {or: mailbox]. c. I am standing at the entrance to an exhibition do not represent these surroundings as fully hall looking at a slab of marble. A man is just or vividly as we do our own, but we need at approaching it with a camera [or: chisel]. least some representation of that space. In an experiment by Morrow and Clark (1988), Tracking a first-person narrator requires people were given one of the two alternatives following a deictic center-the I, here, and of these and other descriptions and were asked now of the narrator's point of view. This is to estimate the distance of, say, the tractor, especially important for interpreting deictic or mouse, from the picket fence. The follow- expressions. These are expressions whose in- ing table gives the average estimates of those terpretation depends on the speaker's or ad- distances: dressee's current point of view. Examples include: come and go, this and that, here and (27') a. tractor to fence, 39 feet; mouse to fence, 2 feet there, this side and the other side, in front b. man to post office, 28 feet; man to mailbox, 13 feet of and behind, and left of and right of (see c. man with camera to marble slab, 18 feet; man BUhler, 1982; Duchan, Bruder, & Hewitt, with chisel to marble slab, 5 feet 1995; Fillmore, 1975; Levinson, 1996). In Apparently people arrive at a denotation Hemmingway's The Killers, the narrator for approach by considering how near an opens his story this way: object must be to a landmark in order to be (28) The door to Henry's lunchroom opened and two in "interaction with it" for its assumed pur- men came in. pose. These judgments depend on the size of the referent object (as in 27a), the size of the As Fillmore ( 1981) noted, the narrator must landmark (27b ), and the purpose of the person be inside the lunchroom, because he describes approaching (27c). the door as opening by unseen forces and the 244 Psycholinguistics

men as "coming" in, not "going" in. The de- these two sentences about Kathy's move- ictic center is inside the room. Point of view ments, which were followed by a question: is essential to many of the narrator's choices, (31) She walked from the study into the bedroom. and imagining the scene from the narrator's or She didn't find the glasses in the room. protagonist's vantage point is crucial to get- Which room is referred to? ting that point of view right. Abrupt changes in point of view require For different people, the first sentence had dif- abrupt changes in the imagined spatial rep- ferent prepositions (from vs. through vs. past resentation, and these are sometimes diffi- the study, and into vs. to the bedroom) and dif- cult to perform. In a demonstration by Black, ferent verb modalities (walked vs. was walk­ Turner, and Bower (1979), people read sim- ing). All these differences influenced which ple descriptions such as the two following room people inferred to be the referent of the examples: room in the second sentence. The following are the results of two variants (in percent of (29) Bill was sitting in the living room reading the choices by the participants): paper, when John came [or: went] into the living room. (32) She walked from the study into the bedroom The room referred to: the bedroom, 77%; (30) Alan hated to lose at tennis. Alan played a game the study, 21 %; other rooms, 2% oftennis with Liz. After winning, she came (33) She walked past the study to the bedroom [or: went] up and shook his hand. The room referred to: the bedroom 21 %; the study 73%; other rooms, 6% We, as readers, can think of point of view in sentences 29 and 30 by setting up a camera to In sentence 32, most people took Kathy to be view the scenes. For the first clause in 29, we in the bedroom, but in 33, most took her to be would set it up in the living room and leave near the study. To figure out where Kathy was it there when John "comes" in. This is not in 32 or 33, people had to consult their repre- the case when John "goes" in, for the cam- sentation of the model house and, against that, era would need to start out of the living room interpret the combination of walked, from or and then follow John into the room. In sen- past the study, and into or to the bedroom. tence 30, the camera would be near Alan for There are no clear answers as to how they the first two sentences, so it would not need did that. to be moved when Liz "comes" up to him. People also track larger features of the It would need to be moved when she "goes" spatial surroundings (Bower & Morrow, up to him, following Liz when she moved. 1990; Morrow, Bower, & Greenspan, 1989; Changing point of view (as with "went" in 29 Morrow, Greenspan, & Bower, 1987). In a and 30) should be disruptive to understanding, study by Glenberg, Meyer, and Lindem as the study showed. Participants took longer (1987), people were given paragraphs to read, to read the passages with the changed points one sentence at a time. Some read one of the of view, and were also more likely to recall two versions of 34: them incorrectly (see also Bruder, 1995). (34) Warren spent the afternoon shopping at the store. People keep track of changing points of He picked up [or: set down] his bag and went view even without deictic expressions. In an over to look at some scarves. He had been shopping all day. experiment by Morrow ( 1985), people mem- He thought it was getting too heavy to carry. orized the layout of a small model house and then read brief narratives about it, one The pronoun it in the last sentence refers to the sentence at a time. One narrative ended in bag mentioned in the second sentence. When Representations of Discourse 245

the verb in the second sentence is picked up, Then there was a ldtchen, Warren keeps the bag with him as he looks at and lhen bathroom, and then the main room was in the back, living set down, the scarves, but when the verb is he room, I guess. leaves the bag behind. In this stndy, the bag's location was important to the interpretation They would begin at the front door and de- of the pronoun. People read the final sentence scribe a tour that passed through each room a full 0.6 seconds faster when the verb was . precisely once. Apparently, they imagined picked up than when it was set down. The as- someone ("you") taking the tour, for they de- sumption is that they could readily locate the scribed landmarks in relation to the tourist's referent for it when the bag was still with War- instantaneous positions with such deictic ren, but they could not locate the referent as terms as to the left and straight ahead (see readily when Warren did not have the bag. Par- Ehrich & Koster, 1983; Levelt, 1982; Shanon, ticipants must have been consulting a spatial 1984; Ullmer-Ehrich, 1987). Descriptions representation in determining the referent. like these are route descriptions. Only a few Deploying spatial representations in dis- people gave survey descriptions, in which they course is, therefore, complicated. To make described the scene from a bird 's eye's view- these judgments, people need to represent the as a mental map. In these cases, they located protagonist's surroundings and keep track of landmarks with absolute terms such as to the where the protagonist is. To do that, they must north of and next to. Apparently, it was more consult their common ground with the writer, natural to describe apartments with route than including their practical kuowledge of houses, with survey descriptions. department stores, acts of walking, and other How do listeners understand route and common items and events. They must com- survey descriptions? In studies by Taylor bine this with information from the descrip- and Tversky (1992, 1996; see also Perrig & tions, such as the verb (walked), the prepo- Kintsch, 1985), people read either a route or sitional phrases (from the study and into the survey description of a small town and were bedroom), and other items (the bag). The issue then asked inferential questions from either a is how people combine such disparate sources route perspective or a survey perspective. Peo- of information to arrive at their understanding ple were just as fast at answering questions (see Glenberg, Kruley, & Langston, 1994). from either perspective regardless of whether they had read the route or survey description. Menta/Maps Apparently, they created representations of the town that were independent of the type of It is often assumed that people consult men- description they read. These representations tal maps of their homes, neighborhoods, and must be more than simple maps in the head, cities as they travel through them. If so, do for they allow people to jump back and forth they create and consult the maps in using lan- between route and survey perspectives almost guage? In a classic study by Linde and Labov at will. (1975), people were asked, "Could you tell me the layout of your apartment?" Almost all Schemas and Mental Models responded by taking the questioner on an imaginary tour, as in this example: People appear to have special cognitive tools for narrating stories or conversing about (35} You walk in the front door. everyday activities. These tools include sche- There was a narrow hallway. To the left, the first door you came to was a mas and scripts, mental models, and mental tiny bedroom. simulations. 246 Psycholinguistics

Schemas presence and order of everyday events. When l In the early 1900's, developed we go to a restaurant, our "restaurant script" the notion of schema to account for how informs us that we need to order from a menu, people understand and remember stories. A wait for our food, and pay at the end. When schema is a set of cultural preconceptions we hear a description about going to a restau- about causal or other types of relationships- rant, we appeal to the same script. Even if part of communal conunon ground. In the not explicitly told, we assume that the pro- classic experiments by Bartlett (1932), people tagonist ordered food and paid the bill in the were told a Native American folk story, "The proper order (Bower, Black, & Turner, 1979). War of the Ghosts," which included many If we are told that the events occurred in an. elements unfamiliar to Western norms. In unusual order, such as the protagonist paid be- retelling that story, people often distorted it to fore ordering his or her food, we may recall fit their cultural expectations, such as chang- the events in their usual order because that ing "hunting seals" into "fishing," a more fits our "restaurant script." Scripts are part of likely pastime according to their schema. conununal conunon ground, so they vary with Schemas of a different type were pro- the conununity. The restaurant script in North posed for the structure of stories themselves. America is strikingly different from those in According to one account (Rumelhart, 1975), Greece and Japan. stories consist of a setting followed by an Schemas were designed, then, to explain episode; an episode consists of an event plus how people could have a mental representa- a reaction to it; a reaction consists of an inter- tion of a narrative that is more detailed than nal response plus an external response; and so the original. People take the limited input and, on. Listeners are assumed to parse stories into by applying schemas, elaborate on it in appro- these functional sections much as they parse priate ways. sentences into constituents. A rather differ- Mental Models ent account (Labov, 1972) is that narratives of personal experience have six parts: Whereas schemas represent cultnral precon- ceptions, mental models are mental constrnc- 1. An abstract, briefly sununarizing the tions in which people represent specific ob- story; jects, events, and relationships in utterances or narratives (Johnson- Laird, 1983; Garnham 2. An orientation, a stage setting about the & Oakhill, 1996). They are mental instantia- who, when, what, and where of the story; tions of the world being described. People cre- 3. A complicating action; ate mental models based on the discourse, the 4. An evaluation of these actions; situation, and the purposes they have to serve. 5. The result or resolution of the complicat- So, people trying to understand "Three tur- ing action; and tles rested on a floating log and a fish swam 6. A coda, a signal of completion. beneath it" create mental models of ponds, logs, fish, and tnrtles so that they can esti- Narrators and their audiences are assumed to mate where they are in relation to each other. refer to such schemas in producing and under- People trying to interpret approach in "The standing stories. tractor approached the fence" create mental A third class of schemas, known as scripts, models of the scene described in order to was proposed as representations for events judge where tractor and fence must be. Mental (Schank & Abelson, 1977). The idea is that models begin, in effect, with the generic infor- scripts guide people's expectations about the mation represented in schemas (in communal Representations of Discourse 247 common ground), and add visual and spatial street," or "What if he had left two minutes relationships to represent instantiations of a earlier?" (Kahneman & Tversky, 1982). Men- scene or event (in perso.nal common ground). tal simulations require the active participa- Mental models can also represent dynamic tion of the participants, and they introduce a events. If a person is asked how many win- boundary between reality and the simulation dows there are in his or her house, that person (taking the system "off-line" and feeding it is likely to imagine him- or herself walking counterfactual inputs). However, many of the around the house counting the windows-a specifics of mental simulation have yet to be dynamic process (Shepard & Cooper, 1982). tested experimentally. According to Hegarty (1992; Hegarty, Just, & Morrison; 1988), people understand diagrams Fictional Worlds of pulleys in much the same way-through dynamic mental models (see also Gentner & The evidence in this chapter has focused on Stevens, 1983). These seem eminently suited people's reliance upon visual, spatial, sche- for representing the dynamic course of events matic, and scriptal representations for the people consult in telling and understanding actual world-apartment layouts, visits to narratives. restaurants, personal experiences. However, people need something more when the situ- Mental Simulations ations are fictional, and that "sOmething" is Mental simulations were proposed by called joint pretense. People engage in a sim- Kahneman and Tversky (1982) as a type of ple pretense whenever they act as if they were dynamic mental model in which people can doing something they are not actually, really, modify the initial settings of the model and or seriously doing at that moment (Goffman, compare the outcomes. People might simulate 1974). An example is lying. Liars act as if a process for many purposes: (a) to predict its they were actually, really, or seriously claim- outcome, (b) to assess its probability, (c) to as- ing what they appear to be claiming. Fiction, sess counterfactual alternatives ("if only ..."), however, requires a joint pretense, when two and (d) to project the effects of causality. people coordinate on the pretense, mutually When people simulate alternative endings to a aware that they are doing so. story, for example, they tend to make "down- The prototype of joint pretense is the game hill changes" to scenatios, that is they remove of make-believe (Clark, 1996; Walton, 1978, unusual or unexpected aspects of the situa- 1983, 1990). Suppose that Ned and Kenneth, tion. They rarely make "uphill changes," or both 5 years old, are pretending to be lion and changes that introduce unusual aspects, and lion-tamer. To succeed, they must coordinate never make "horizontal" changes, or changes their imaginings. They must simulate the way that alter arbitrary aspects. Mental simulations a lion and lion-tamer would behave toward represent the process of imagining working each other. They must both imagine the back through an event. yard as a circus ring, the back porch as a lion Mental simulations are well suited for cage, and much, much more. In playing their imaginary experiences (see Davies & Stone, game, they are simultaneously engaged in two 1995), and these include emotional experi- [ayers ofjoint action. ences. When people go back over fatal ac- cidents of loved ones, they often experience Layer 1: Ned and Kenneth are playing a I guilt, anger, or regret as they mentally simu- game of make-believe, jointly pre- late alternatives for those accidents-as they tending to be taking the actions at II think "If only she hadn't driven down that layer 2. •

I 248 Psycholinguistics

Layer 2: Ned and Kenneth are a lion and lion- the safety of imagination. Ned and Kenneth, tamer performing in a circus. the two 5-year-olds, play their game because it is exciting to imagine living in the circus The domain of layer I is the real or actual and to simulate experiences they could not world. The domain of layer 2 is a fictional have in the actual world. Reynard's under- world. Both are part of Ned and Kenneth's standing of the joke becomes more exciting current common ground. and more vivid, when he can imagine an ac- All fiction requires two or more layers of tual chemist saying, "Well I've got all kinds, joint action (Bruce, 1981; Clark, 1996; Currie, and all prices, what do you want?" Novels, 1990; Walton, 1978, 1983, 1990). Take the jokes, and short stories are a mixture of telling first lines of a joke told by Sam to Reynard and showing-of diagesis, or description, and (Svartvik & Quirk, 1980): mimesis, or demonstration. As novelist David

(35) let me tell you a story,--- Lodge (1990) noted, "[The] alternation of a girl went into a chemist's shop, and asked for, . authorial description and characters' verbal contraceptive tablets, - - interaction remains the woof and warp of so he said "well I've got . all kinds, and all prices, what do you want," literary narration to this day." she said "well what have you got," Imaginal Props There are three layers to this example. In Novels, jokes, and short stories aren't the layer I, the real or actual world, Sam is an- only venues for fictional language. There are nouncing the story to Reynard, "Let me tell also stage plays, radio plays, operas, op- you a story." In layer 2, a fictional world, a erettas, puppet shows, films, television situ- reporter is telling a friend about a conversa- ation comedies, soap operas, film cartoons, tion between a chemist (a pharmacist) and a comic books, songs, and pantomimes. Many young woman. The quotation in line 3 shows narratives have appeared in several media. the third layer, the world of the chemist speak- Jane Austen's Emma comes as a novel, au- ing to the young women. dio recording, and film, and it could proba- To participate in this joke, Sam and bly be produced as a radio play, comic book, Reynard must engage in joint pretense. When stage play, and opera. These forms are not Sam produces "A girl went into a chemist's all alike. They range in how they engage our shop and asked for contraceptive tablets," he imagination-and in how effectively they do is asking Reynard to join with him in pre- that. tending that an actual reporter is telling an ac- Imaginal props are one device for engag- tual friend about a young woman going into ing the imagination. Imaginal props are de- a chemist's shop. Crucially, the deictic cen- vices that support the imagining of a situation. ter changes with each layer. In layer I, I and They are engineered to get the audience en- you are Sam and Reynard; in layer 2, I and you grossed in a fictional world (Clark & VanDer are the reporter and his friend; and in layer 3, I Wege, 2001), such as the following examples and you are the chemist and the young woman. demonstrate: Reynard cannot interpret me and you in line 1, went in line 2, or I and you in line 3 without 1. Quotation. In sentence 35, Sam quotes the keeping track of these layers. The same goes chemist as saying, "Well, I've got all kinds for many other features as well. and all prices." If he delivers the line well, Joint pretense is valuable because it al- Sam can help Reynard imagine, or experi- lows participants to have vivid experiences in ence, not only the chemist's point of view, Representations of Discourse 249

but also his accent and sympathy. Even the moods in fictional Egypt. The same goes barest quotations add vividness. for background music in films. Excite- 2. Iconic and deictic gestures. In spontaneous ment, suspense, sadness, and other moods stories, speakers often use iconic and deic- would be harder to create without it. tic gestures to depict and point to things Imaginal props are tricks of the fictional in the fictional world (Haviland, 1996; trade. In the hands of the best writers, sto- McNeill, 1992). In an example discussed rytellers, film directors, and actors, they help earlier (Kendon, 1980), Fran tells an anec- engross us in the fictional world. The issue Some Like it Hot. dote from the film At is how. one point she says, "They wheel a big table ·in [sweeping her arm to depict the Experientinl Representations motion], with a big with a big [1.08 sec] When we get engrossed in a story, we often cake on [tracing a horizontal circle to it experience (see Gross & Levenson, depict its shape], and the girl [raises arm 1995). Consider what Walton (1978) calls to depict jumping up], jumps up." The quasi-fear. When we see a horror film, we gestures clarify what she is saying and are afraid of what the monster will do to make the fictional scene that much more the heroine. Our hearts beat faster, our mus- vivid. cles tighten, and our knuckles tum white as 3. Enactments. When a stage actor is the monster approaches her. But do we warn Hamlet, or a film actress is Emma, they her as we would if all this were happening do more than recite their lines. They enact in front of us? This is what makes it quasi- their characters. When we read Emma, we fear and not real fear. Next, consider what work hard to imagine what Emma looks Gerrig (1989a, 1993; see also Gerrig & like-her hair, clothing, and mannerisms. Prentice, 1991 ; Prentice & Gerrig, 1999; Without a background in 19th-century Prentice, Gerrig, & Bailis, 1997) called English style, we may get it wrong. In see- anomalous suspense. Ordinarily, suspense is a ing the film Emma, we are shown what state in which we "lack knowledge about some she looks like, including her hair, clothing, sufficiently important target outcome (p. 79)." mannerisms, and what she does. All we Yet, as Gerrig demonstrated in a series of ex- must imagine is that this particular actress periments, when we read suspense stories, we (say, Gwyneth Paltrow) is in fact Emma. often feel suspense even when we know how 4. Staging. Stage plays, films, operas, and they tum out. Like Walton's quasi-fear, we comic books rely on staging. The pro- compartmentalize our experience as part of duction crew engineers the scenery, scene the story world and separate from the actual changes, timing, close-ups, and other fea- world. tures to help engross the audience in the Most narratives are engineered to elicit right fictional world. Nothing kills imagi- . Novels are classified into genres nation like bad production. largely by the emotions they evoke. Myster- 5. Sound effects. It may seem that the greater ies lead to suspense and fear; adventures to the verisimilitude of the imaginal prop, excitement, fear, and elation; horror stories the better the aid to imagination, but that to horror, loathing, and fear; light romances isn't always true. In Verdi's opera, Aida's to sexual excitement; heavier romances to singing is hardly realistic speech, yet it erotic arousal; satires to amusement; and so helps us create her happy or melancholy on. Films are classified in much the same way. 250 Psycholinguistics

We imagine story worlds as if we were now video recordings, both for laboratory experi- experiencing them before our very eyes. At ments aud for the aualysis of spontaueous con- the same time, we recognize that we are still versation. They can measure reaction times, in the actual world. aualyze aud synthesize speech sounds, and How, then, do people represent fiction? A track brain activity. Second, there have been complete answer must account for at least four breakthroughs in theory. Since Wundt' s time, phenomena (Clark & Vau Der Wege, 2001): linguistic theory has become a highly so- phisticated, if controversial, area of study. 1. Experience. People experience selective Researchers have also developed major theo- features of the narrative world as if they ries for speaking, parsing, , were actual, current experiences. These in- , reading, aud brain acti- clude visual appearauces, spatial relations, vation, to name just a few areas. All of these points of view, movement and processes, theories staud on a foundation of strong em- voices, and emotions. pirical results. 2. Imaginal props. People's imaginings Still, the essence of lauguage use is found appear to be aided by well-engineered in face-to-face talk. It is here that speaking imaginal props. such as direct quotation, and listening arise in their natural, universal gestures, stage sets, sound effects, and states. It is here that researchers can study why background music. speakers say the things they say aud how lis- 3. Participation. Speakers and writers design teners interpret these things-ultimately, as what they say to encourage certain forms a way of coordinating joint activities. It is of imagination, but listeners and readers here that researchers can study people's use must cooperate with them to succeed. of common ground-in everything from iden- 4. Compartmentalization. In participating in tifying words like candy to drawing elabora- narratives, people distinguish their experi- tive inferences. It is here that researchers can ences in the story world from their experi- study how speakers combine description, in- ences in the real world. dication, and demonstration to say what they say. The problem is that too little is known It isn't enough to posit visual or spatial repre- about spontaneous language and how it dif- sentations, schemas, scripts, mental models, fers from reciting, reading aloud, listening to and even mental simulations. It takes layering idealized speech, aud other such forms. Un- and joint pretense to account for these four derstauding lauguage in its natural habitat is phenomena. However, more investigation is a major challenge for the second century of needed to determine how they work. psycholinguistics.

SUMMARY REFERENCES

In the century since Wundt, psycholinguis- Altmann, G., & Steedman, M. (1988). Interaction tics has come a long way. First, there have with context during . been breakthroughs in research methods. In Cognition, 30(3), 191-238. Wundt's time, one could investigate written Arnold, J., Wasow, T., Losongco, A., & Ginstrom, language, or slips of the tongue, as Freud did, R. (2000). Heaviness vs. newness: The effects but not much more. With advances in tech- of complexity and information structure on con- nology, investigators now exploit audio- and stituent ordering. Language, 76, 28-55. References 251

Atkinson, J. M., & Heritage, J. (Eds.). (1984). Bever, T. G. (1970). The cognitive basis for lin- Structures of social action: Studies in conversa­ guistic structures.lnJ. R. Hayes (Ed.), Cognition tion analysis. Cambridge: Cambridge University and the development of language (pp. 279-352). Press. New York: Wiley. Austin, J. L. (1962). How to do things with words. Bilous, F. R. (1992). The role of gestures in Oxford: Oxford University Press. : Gestures enhance lexical ac­ Bach, K., & Harnish, R. M. (1979). Linguistic com­ cess. Unpublished Ph.D. dissertation, Columbia munication and speech acts. Cambridge: MIT University. Press. Black, J. B., TUrner, T. J., & Bower, G. H. (1979). Bartlett, F. C. (1932). Remembering: A study in ex­ Point of view in narrative comprehension. perimental and . Cambridge: Journal ofVerbalLearning and Verbal Behavior, Cambridge University Press. 18, 187-198. Bavelas, J. B., Black, A., Lemery, C. R., & Bock, K., & Levelt, W. J. M. (1994). Language Mullett, J. ( 1986). "!show you how you feel": production: Grarrunatical encoding. In M. A. Motor mimicry as a communicative act. Jour­ Gernsbacher (Ed.), Handbook ofpsycholinguis­ nal of Personality and Social Psychology, 50, tics (pp. 945-984). San Diego: Academic Press. 322-329. Bower, G. H., Black, J. B., & TUrner, T. J. (1979). Cognitive Psychol­ Bavelas, J. B., & Chovil, N. (2000). Visible acts Scripts in memory for text. of meaning. An integrated message model of ogy, 11, 177-220. language in face-to-face dialogue. Journal of Bower, G. H., & Morrow, D. G. (1990). Men- Language and Social Psychology, 19, 163- tal models in narrative comprehension. Science, 194. 247(4938), 44-48. Bavelas, J. B., Chovil, N., Lawrie, D. A., & Bransford, J. D., Barclay, J. R., & Franks, J. J. Wade, A. (1992). Interactive gestures. Discourse (1972). Sentence memory: A constructive vs. Processes, 15, 469-489. interpretive approach. , 3, 193-209. Bavelas, J. B., Coates, L., & Johnson, T. (2000). Listeners as co-narrators. Journal of Personality Brown, P., & Levinson, S. C. (1987). Politeness. & Social Psychology, 79(6). Cambridge: Cambridge University Press. Beattie, G., & Coughlan, J. (1998). Do iconic ges- Bruce, B. ( 1981 ). A social interaction model of tures have a functional role in lexical access? reading. Discourse Processes, 4, 273-311. An experimental study of the effects of repeat- Bruder, G. A. (1995). Psychological evidence that ing a verbal message on gesture production. linguistic devices are used by readers to under- Semiotica, 119(3-4), 221-249. stand spatial deixis in narrative text. In J. F. Beattie, G., & Coughlan, J. (1999). An experimen- Duchan, G. A. Bruder, & L. E. Hewitt (Eds.), tal investigation of the role of iconic gestures in Deixis in narrative: A perspec­ lexical access using the tip-of-the-tongue phe- tive (pp. 243-260). Hillsdale NJ: Er1baum. nomenon. British Journal of Psychology, 90(1), Buchler, J. (Ed.). (1940). Philosophical writings of 35-56. Peirce. London: Routledge and Kegan Paul. Beattie, G., & Shovelton, H. (1999). Mapping Btihler, K. (1934). Sprachtheorie: die Darstel· the range of information contained in the lungsfunktion der Sprache. Jena: Fischer. iconic hand gestures that accompany sponta- Biihler, K. (1982). The deictic field of!anguageand neous speech. Journal of Language & Social deictic words. InR.J.Jarvella& W. Klein(Eds.), Psychology, 18(4), 438-462. Speech, place, and action (pp. 9-30). Chichester, Behaghel, 0. (1909/1910). Beziehungen zwis- England: Wiley. chen Umfang und Reihenfolge von Satzgliedem. Butterworth, B., &Beattie, G. (1978). Gestures and Indogermanische Forschungen, 25, 110---142. silence as indicators of planning in speech. In 252 Psycholinguistics , ' ' R. Campbell & P. T. Smith (Eds.), Recent ad­ Clark, H. H. (1994). Discourse in production. vances in the psychology of language: Fonnal In M. A. Gemsbacher (Ed.), Handbook of and experimental approaches. NATO Confer~ psycholinguistics (pp. 985-1021). San Diego: ence Series ll/:4B (pp. 347-360). New York: Academic Press. Plenum. Clark, H. H. (1996). Using language. Cambridge: Button, G., & Lee, J. R. (Eds.). (1987). Talk and Cambridge University Press. social organisation. Philadelphia: Multilingual Clark, H. H. (1998). Communal lexicons. In Matters. K. Malmkjrer & J. Williams (Eds.), Context in Caramazza, A., & Grober, E. (1976). Polysemy language learning and language understanding and the structure of the subjective lexicon. In (pp. 63-87). Cambridge: Cambridge University C. Rameh (Ed.), Georgetown University round Press. table on language and linguistics (pp. 181-206). Clark, H. H. (in press). Pointing and placing. In Washington, DC: Georgetown University Press. S. Kita (Ed.), Pointing. Where language, culture, Chafe, W. (1979). The flow of thought and the and cognition meet. Hillsdale, NJ: Erlbaum. flow of language. InT. Givon (Ed.), Syntax and Clark, H. H., & Brennan, S. A. (1991). Ground- semantics 12: Discourse and syntax (pp. 159- ing in communication. In L. B. Resnick, J. M. 181). New York: Academic Press. Levine, & S. D. Teasley (Eds.), Perspective Chafe, W. (1980). The deployment of conscious- on socially shared cognition (pp. 127-149). ness in the production of a narrative. In W. Chafe Washington, DC: APA. (Ed.), The pear stories (pp. 9-55). Norwood, NJ: Clark, H. H., & Clark, E. V. (1977). Psychology and Ablex. language: An introduction to psycho linguistics. Chantraine, Y., & Hupet, M. (1994). Efficiency New York: Harcourt Brace Jovanovich. of the addressee's contribution to the estab- Clark, H. H., & Fox Tree, J. E. (2001). Using uh lishment of references: Comparing mono~ and urn in spontaneous speaking, Stanford Uni~ Iogue with dialogues. Cahiers de Psychologie versity. Manuscript. Cognitive/Current Psychology of Cognition, Clark, H. H., & Gerrig, R. J. (1983). Understanding 13(6), 777-796. old words with new meanings. Journal of Verbal Chomsky, N. (1957). . Learning & Verbal Behavior, 22(5), 591--608. 's~Gravenhage: Mouton. Clark, H. H., & Gerrig, R. J. (1990). Quotations as Chomsky, N. (1965). Aspects of the theory of syn­ demonstrations. Language, 66, 764-805. tax. Cambridge: MIT Press. Clark, H. H., & Haviland, S. E. (1977). Compre- Clark, E. V., & Clark, H. H. (1979). When nouns hension and the given-new contract. In R. 0. surface as verbs. Language, 55, 430-477. Freedle (Ed.), Discourse production and com~ Clark, H. H. (1977a). Bridging. In P. N. Johnson- prehension (pp. 1-40). Hillsdale, NJ: Erlbaum. Laird & P. C. Wason (Eds.), Thinking: Clark, H. H., & Marshall, C. R. (1981). Definite in cognitive science (pp. 411-420). Cambridge: reference and mutual knowledge. In A. K. Joshi, Cambridge University Press. B. L. Webber, &I. A. Sag (Eds.), Elements ofdis­ Clark, H. H. (1977b). Inferences in comprehension. course understanding (pp. 10-63). Cambridge: In D. LaBerge & S. J. Samuels (Eds.), Basic Cambridge University Press. processes in reading: Perception and compre~ Clark, H. H., & Schaefer, E. R. (1989). Con- hension. Hillsdale, NJ: Erlbaum. tributing to discourse. Cognitive Science, 13, Clark, H. H. (1979). Responding to indirect speech 259-294. acts. Cognitive psychology, 11, 430-477. Clark, H. H., & Schunk, D. H. (1980). Po- Clark, H. H. (1983). Making sense of nonce sense. lite responses to polite requests. Cognition, 8, In G. B.Floresd'Arcais& R.Jarvella(Eds.), The 111-143. process of language understanding (pp. 297- Clark, H. H., & Van Der Wege, M. (2001). 331). New York: Wiley. Imagination in discourse. In D. Schiffrin & References 253

D. Tannen (Eds.), Handbook of discourse. Wilson (Ed.), Lexical representation and Oxford: Blackwell. process (pp. 227-260). Cambridge: MIT Press. Clark, H. H., & Wasow, T. (1998). Repeating words Engle, R. A. (1998). Not channels but composite in spontaneous speech. Cognitive Psychology, signals: Speech, gesture, diagrams, and object 37(3), 201-242. demonstrations are integrated in multimodal ex- planations. In M.A. Gernsbacher & S. J. Derry Clark, H. H., & Wilkes-Gibbs, D. (1986). Refer- (Eds.), Proceedings of the Twentieth Annual ring as a collaborative process. Cognition, 22, Conference of the Cognitive Science Society. 1-39. Mahwah, NJ: Erlbaum. Cohen, A. A., & Harrison, R. P. (1973).lntention- Engle, R.I. (2000). Toward a theory ofmulti-modal ality in the use of hand illustrators in face-to-face communication: Combining speech, gestures, communication situations. Journal of Personal­ diagrams, and demonstrations in instructional ity and Social Psychology, 28, 276-279. explanations. Unpublished doctoral disserta- Currie, G. (1990). The nature offiction. Cambridge: tion, Stanford University, CA. Cambridge University Press. Erikson, T. A., & Mattson, M. E. (1981). Dahan, D., Swingley, D., Tanenhaus, M. K., & From words to meaning: A semantic illusion. Magnuson, J. S. (2000). Linguistic gender and Journal of Verbal Learning and Verbal Behav­ spoken-word recognition in French. Journal of ior, 20, 540-552. Menwry and Language, 42(4), 465-480. Erman, B. (1987). Pragmatic expressions in Davies, M., & Stone, T. (Eds.). (1995). Mental English: A study of you know, you see, and I simulation. Oxford: Blackwell. mean in face-to-face conversation. Stockholm: Dell, G. S. (1986). A spreading-activation theory of Almqvist & Wiksell International. retrieval in sentence production. Psychological Fillenbaum, S. (1971). On coping with ordered Review, 93(3), 283-321. and unordered conjunctive sentences. Journal of Dell, G. S., Burger, L. K., & Svec, W. R. (1997). , 87, 93-98. and serial order: A func- Fillenbaum, S. (1974). Pragmatic normalization: tional analysis and a model. Psychological Further results for some conjunctive and disjunc- Review, 104(1), 123-147. tive sentences. Journal ofExperimental Psychol­ Dell, G. S., & O'Seaghdha, P. G. (1991). Mediat- ogy, 102, 574-578. ed and convergent lexical in language Fillmore, C. (1975). Santa Cruz lectures on production: A comment on Levelt et al. (1991). deixis. Bloomington, IN: Indiana University , 98(4), 604-614. Linguistics Club. Drew, P., & Heritage, J. (Eds.). (1992). Talk Fillmore, C. (1981 ). Pragmatics and the description at work: Interaction in institutional settings. of discourse. In P. Cole (Ed.), Radical pragmat­ Cambridge: Cambridge University Press. ics (pp. 143-!66). New York: Academic Press. Duchan, J. F., Bruder, G. A., & Hewitt, L. E. Ford, C. E., & Thompson, S. A. (1996). Interac- (1995). Deixis in narrative: A cognitive science tional units in conversation: Syntactic, intona- perspective. Hillsdale, NJ: Erlbaum. tional, and pragmatic resources for the manage- Ehrich, V., & Koster, C. (1983). Discourse or- ment of turns. In E. Ochs, E. A. Schegloff, & ganization and sentence form: The structure S. A. Thompson (Eds.), Interaction and gram­ of room descriptions in Dutch. Discourse mar (pp. 134-184). Cambridge: Cambridge Processes, 6(2), 169-195. University Press. Ekman, P., & Friesen, W. (1969). The repertoire of Fox Tree,J. E., & Clark, H. H. (1997). Pronouncing nonverbal behavior: Categories, origins, usage "the" as "thee" to signal problems in speaking. and coding. Semiotica, I, 49-98. Cognition, 62(2), 151-167. Elman, J. L. ( 1989). Connectionist approaches to Fox Tree, J. E., & Schrock, J. C. (1999). Discourse acoustic/phonetic processing. In W. Marslen- markers in spontaneous speech: Oh what a 254 Psycholinguistics

difference an oh makes. Journal ofMemory and Gerrig, R. J. (1989a). Suspense in the absence of Language. 40(2), 280-295. uncertainty. Journal of Memory and Language, Francik, E. P., & Clark, H. H. (1985). How to make 28(6), 633--<548. requests that overcome obstacles to compliance. Gerrig, R. J. (1989b). The time course of Journal ofMemory and Language, 24, 560--568. sense creation. Memory & Cognition, 17(2), Fromkin, V. A. (1971). The non-anomalous nature 194-207. of anomalous utterances. Language, 47, 27-52. Gerrig, R. J. (1993). Experiencing narrative Fromkin, V. A. (Ed.). (1973). Speech errors as worlds: On the psychological activities of read­ linguistic evidence. The Hague, Netherlands: ing. New Haven: Yale University Press. Mouton. Gerrig, R. J., & Bortfeld, H. (1999). Sense cre- Fussell, S. R., & Krauss, R. M. (1992). Co- ation in and out of discourse contexts. Journal ordination of knowledge in communication: ofMemory and Language, 41(4), 457-468. Effects of speakers' assumptions about what Gerrig, R. J., & Gibbs, R. W. (1988). Beyond others know. Journal of Personality & Social the lexicon: Creativity in language production. Psychology, 62(3), 378-391. Metaphor & Symbolic Activity, 3(1), 1-19. Garnharn, A., & Oakhill, J. V. (1996). The mental Gerrig, R. J., & Prentice, D. A. (1991). The repre- models comprehension. In sentation of fictional information. Psychological Models B. K. Britton & A. C. Graesser (Eds.), Science, 2(5), 336-340. of understanding text (pp. 313-339). Hillsdale, Gibbs, R. W. (1979). Contextual effects in under- NJ: Erlbaum. standing indirect requests. Discourse Processes, Garrett, M. F. (1980). Syntactic processes in 2, 1-10. sentence production. In B. Butterworth (Ed.), Gibbs, R. W. (1983). Do people always process the Speech production (pp. 170-220). New York: literal meanings of indirect requests? Journal of Academic Press. Experimental Psychology: Learning, Memory, Garrod, S., & Anderson, A. (1987). Saying what and Cognition, 9, 524-533. you mean in dialogue: A study in conceptual Gibbs, R. W. (1986). What makes some indirect a~d semantic co-ordination. Cognition, 27(2), Journal of Memory 181-218. speech acts conventional? and Language, 25(2), 181-196. Garrod, S., & Doherty, G. (1994). Conversation, co-ordination and convention: An empirical in- Gibbs, R. W., & Mueller, R. A. (1988). Con- vestigation of how groups establish linguistic versational sequences and preference for indi- conventions. Cognition, 53(3), 181-215. rect speech acts. Discourse Processes, 11(1), 101-116. Garrod, S. C., & Sanford, A. J. (1982). Bridging inferences in the extended domain of reference. Glenberg, A. M., Kru1ey, P., & Langston, W. E. In A. Baddeley & J. Long (Eds.), Attention and (1994). Analogical processes in comprehension: peifonnance IX (pp. 331-346). Hillsdale, NJ: Simulation of a mental model. In M. A. Erlbaum. Gernsbacher (Ed.), Handbook ofpsycholinguis­ tics Garrod, S.C.. & Sanford, A. J. (1994). Resolving (pp. 609-640). San Diego: Academic Press. sentences in a discourse context: How discourse G1enberg, A.M., Meyer, M., & Lindem, K. (1987). representation affects language understanding. Mental models contribute to foregrounding dur- In M.A. Gernsbacher & et al. (Eds.), Handbook ing text comprehension. Journal ofMemory and of psycholinguistics (pp. 675--<598). San Diego: Language, 26(1), 69-83. Academic Press. Goffman, R ( 1967). Interaction ritual: Essays on Gee, J.P. (1986). Units in the production of narra- face-toface behavior. Garden City, NY: Anchor tive discourse. Discourse Processes, 9, 391-422. Books. Gentner, D., & Stevens, A. (Eds.). (1983). Mental Goffman, E. (1974). Frame analysis. New York: models. Hillsdale, NJ: Erlbaum. Harper and Row. References 255

Goodwin, C. (1981). Conversational organization: Hegarty, M. (1992). Mental animation: Infer- Interaction between speakers and hearers. New ring motion from static displays of mechani- York: Academic Press. cal systems. Journal of Experimental Psychol­ Goodwin, C. (1986a). Between and within: Alter- ogy: Learning, Memory, and Cognition, 18(5), native sequential treatments of continuers and 1084-1102. assessments. Human Studies, 9, 205-217. Hegarty, M., Just, M.A., & Morrison, I. R. (1988). Goodwin, C. (1986b). Gestures as a resource for the Mental models of mechanical systems: Indi- organization of mutual orientation. Semiotica, vidual differences in qualitative and quanti- 62, 29-49. tative reasoning. Cognitive Psychology, 20(2), Goodwin, M. H., & Goodwin, C. (1986). Gesture 191-236. and cop~cipation in the activity of searching Holmes, V. M., Stowe, L., & Cupples, L. (1989). for a word. Semiotica, 62, 51-75. Lexical expectations in parsing complement- Gordon, D., & Lakoff, G. (1971). Conversa- verb sentences. Journal of Memory and Lan­ tional postulates, Papers from the seventh guage, 28(6), 668-689. regional meeting of the Chicago Linguistic Hupet, M., Seron, X., & Chantraine, Y. (1991). Society (pp. 63-84). Chicago: Chicago The effects of the codability and discriminabil- Linguistic Society. ity of the referents on the collaborative referring Greenberg, J. H. (1963). Some universals of gram- procedure. British Journal ofPsychology, 82(4), mar with particular reference to the order of 449-462. meaningful elements. In J. H. Greenberg (Ed.), Jefferson, G. (1972). Side sequences. In D. Universals oflanguage (pp. 58-90). Cambridge: Sudnow (Ed.), Studies in social interaction MIT Press. (pp. 294-338). New York: Free Press. Grice, H. P. (1957). Meaning. Philosophical Johnson, H. G., Ekman,P., &Friesen, W. V. (1975). Review, 66, 377-388. Communicative body movements: American Grice, H. P. (1968). Utterer's meaning, sentence- emblems. Semiotica, 15(4), 335-353. meaning, and word-meaning. Foundations of Johnson, M. K., Bransford, J. D., & Solomon, Language,4, 225-242. S. K. (1973). Memory for tacit implications of Grice, H. P. (1975). Logic and conversation. In sentences. Journal ofExperimental Psychology, P. Cole & J. L. Morgan (Eds.), Syntax and 98(1), 203-205. semantics, Vol. 3: Speech acts (pp. 113-128). Johnson-Laird, P. N. (1983). Mental models: New York: Seminar Press. Towards a cognitive science of language, infer­ Grice, H. P. (1978). Some further notes on logic ence, and consciousness. Cambridge: Harvard and conversation. In P. Cole (Ed.), Syntax and University Press. semantics 9: Pragmatics (pp. 113-127). New Kahneman, D., & Tversky, A. (1982). The sim- York: Academic Press. ulation heuristic. In P. Slovic, D. Kahneman, Grice, H. P. (1991). In the way of words. & A. Tversky (Eds.), Judgment under un­ Cambridge: Harvard University Press. certainty: Heuristics and biases (pp. 201- Gross, J. J., & Levenson, R. W. (1995). Emotion 208). Cambridge: Cambridge University elicitation using films. Cognition & Emotion, Press. 9(1), 87-108. Kendon, A. (1967). Some functions of gaze direc- Haviland, J. B. (1996). Projections, transpositions, tion in two~person conversation. Acta Psycholo­ and relativity. InJ. J. Gumperz & S.C. Levinson gia, 16, 22-63. (Eds.), Rethinking (pp. 271- Kendon, A. (1980). Gesticulation and speech: 1\vo 323). Cambridge: Cambridge University Press. aspects of the process of utterance. In M. R. Hawkins, J. A. (1994). A perfornumce theory of Key (Ed.), Relationship of verbal and nonver· order and constituency. Cambridge: Cambridge hal communication (pp. 207-227). Amsterdam: University Press. Mouton de Gruyter. I ~ 256 Psycholinguistics

Kendon, A. (1987). On gesture: It~ complementary Levelt, W. J. M. (1982). Cognitive styles in the relationship with speech. In A. W. Siegman, & use of spatial direction terms. In R. J. Jarvella S. Feldstein (Ed.), Nonverbal behavior and com­ & W. Klein (Eds.), Speech, place, and action: munication (2nd ed., pp. 65-97). Hillsdale, NJ: Studies in deixis and related topics (pp. 251- Erlbaum. 268). Chichester, England: Wiley. Kendon, A. (1994). Do gestures communicate? A Levelt, W. J. M. (1983). Monitoring and self-repair review. Special Issue: Gesture and understand- in speech. Cognition, 14, 41-104. Research on Language ing in social interaction. Levelt, W. J. M. (1989). Speaking. Cambridge: MIT and Social Interaction, 27(3), 175-200. Press. Kita, S. (1997). Two-dimensional semantic Levelt, W. J. M., & Kelter, S. (1982). Surface form Linguistics, analysis of Japanese mimetics. and memory in question answering. Cognitive 35(2[348]), 379-415. Psychology, 14(!), 78-106. Krauss, R. M. (1998). Why do we gesture when Levelt, W. J. M., Roelofs, A., & Meyer, A. S. we speak? Current Directions in Psychological (1999). A theory of lexical access in speech pro- Science, 7(2), 54--{\0. duction. Behavioral & Brain Sciences, 22(1 ), Krauss,R. M., Morrel-Samuels, P., & Colasante, C. 1-75. (1991). Do conversational hand gestures com- Levelt, W.J. M., Schriefers,H., Vorberg, D.,Meyer, municate? Journal of Personality and Social A. S., Pechman, T., & Havinga, J. (1991). The Psychology, 61(5), 743-754. time course of lexical access ill speech produc- Krauss, R. M., & Weinheimer, S. (1964). Changes tion: A study of picture naming. Psychological in reference phrases as a function of frequency of Review, 98(!), 122-142. usage in social interaction: A preliminary study. Levinson, S. C. (1983). Pragmatics. Cambridge: Psychonomic Science, 1(5), 113-114. Cambridge University Press. Krauss, R. M., & Weinheimer, S. (1966). Con- Levinson, S. C. (1996). Frames of reference and current feedback, confirmation, and the encod- Molyneux's question: Crosslinguistic evidence. ing of referents in verbal communication. Jour­ In P. Bloom, M. A. Peterson, L. Nadel, & M. F. nal of Personality & Social Psychology, 4(3), Garrett (Eds.), Language and space (pp. 108- 343-346. 169). Cambridge: MIT Press. Krauss, R. M., & Weinheimer, S. (1967). Effect Levinson, S. C. (2000). Presumptive meanings. of referent similarity and communication mode Cambridge: MIT Press. on verbal encoding. Journal of Verbal Learning and Verbal Behavior, 6(3), 359-363. Lewis, D. K. (1969). Convention: A philosophical study. Labov, W. (1972). The transformation of expe- Cambridge: Harvard University Press. rience in narrative syntax. In W. Labov (Ed.), Lewis, D. K. (1979). Scorekeeping in a l.an- Language in the inner city: Studies in the Black guage game. Journal of Philosophical Logic; 8, English vemacular(pp. 354-396). Philadelphia: 339-359. University of Pennsylvania Press. Linde, C., & Labov, W. (1975). Spatial networks Lakoff, R. (1973). The logic of politeness; or, as a site for the study of language and thought. minding your p's and q's. Papers from the Language, 5/(4), 924-939. ninth regional meeting of the Chicago Linguis­ Lodge, D. (1990). Narration with words. In tics Society (pp. 292-305). Chicago: Chicago H. Barlow, C. Blakemore, & M. Weston-Smith Linguistics Society. (Eds.), Images and understanding (pp. 141- Lehmann, W. P. (1972). On converging theories in 153). Cambridge: Cambridge University Press. linguistics. Language, 48, 266-275. MacDonald, M. C., Pearlmutter, N. J., & Lehmann, W. P. (1973). A structural principle of Seidenberg, M. S. (1994). The lexical nature language and its implications. Language, 49, of syntactic ambiguity resolution. Psychologi­ 47-66. cal Review, 101(4), 676-703. References 257

Mann, W. C., & Thompson, S. A. (1986). Paek, T. S. Y. (2000). Fonnal and computational Relational propositions in discourse. Discourse methods for achieving mutual understanding Processes, 9, 57-90. in conversations between and comput­ Marslen-Wilson, W., & Tyler, L. K. (1980). The ers. Unpublished doctoral dissertation, Stanford temporal stmcture of spoken language under- University, CA. standing. Cognition, 8(1), 1-71. Perrig, W., & Kintsch, W. (1985). Propositional Marslen-Wilson, W. D. (1987). Functional par- and situational representations of text. Journal allelism in spoken word-recognition. Special of Memory and Language, 24(5), 503-518. Issue: Spoken word recognition. Cognition. Perry, J. (1979). The problem of the essential in- 25(1-2), 71-102. dexical. Nous, 13(1), 3-21. McClelland, J. L., & Elman, J. L. (1986). The Polanyi, L. (1989). Telling the American story. TRACE model of speech perception. Cognitive Cambridge: MIT Press. Psychology, 18(1), 1--136. Prentice, D. A., & Gerrig, R. J. (1999). Explor- McKoon, G., & Ratcliff, R. (1992). Inference ing the boundary between fiction and reality. during reading. Psychological Review, 99(3), In S. Chaiken, Y. Trope & et al. (Eds.), Dual­ 440-466. process theories in social psychology (pp. 529- McNeill, D. (1992). Hand and mind. Chicago: 546). New York: Guilford Press. University of Chicago Press. Prentice, D. A., Gerrig, R. J., & Bailis, D. S. Merritt, M. (1976). On questions following ques- (1997). What readers bring to the processing of tions (in service encounters). Language in fictional texts. Psychonomic Bulletin & Review, Society, 5, 315-357. 4(3), 416-420. Morrel-Samuels, P., & Krauss, R. M. (1992). Word Prince, E. F. (1981). Towards a taxonomy of given- familiarity predicts temporal asynchrony of hand new information. In P. Cole (Ed.), Radical prag­ gestures and speech. Journal of Experimental matics (pp. 223-256). New York: Academic Psyclwlogy: Learning, Memory, and Cognition, Press. 18(3), 615-622. Rapp, D. N., & Gerrig, R. J. (1999). Eponymous Morris, D., Collet~ P., Marsh, P., & verb phrases and ambiguity resolution. Memory O'Shaughnessy; M. (1979). Gestures: Their & Cognition, 27(4), 612-618. origins and distribution. New York: Stein & Rauscher, F. H., Krauss, R. M., & Chen, Y. (1996). Day. Gesture, speech, and lexical access: The role of Morrow, D. G. (1985). Prepositions and verb aspect lexical movements in speech production. Psy­ in narrative understanding. Journal of Memory chological Science, 7(4), 226-231. and Language, 24, 390-404. Reder, L. M., & Cleeremans, A. (1990). The role Morrow, D. G., Bower, G. H., & Greenspan, S. L. of partial matches in comprehension: The Moses (1989). Updating situation models during nar- illusion revisited. In A. C. Graesser & G. H. rative comprehension. Journal of Memory and Bower (Eds.), The psychology of learning and Language, 28(3), 292-312. : Inferences and text comprehension Morrow, D. G., & Clark, H. H. (1988). Interpret- (pp. 233-258). San Diego: Academic Press. ing words in spatial descriptions. Language and Rime, B., & Schiaraoua, L. (1991). Gesture and Cognitive Processes, 3, 275-291. speech. In R. S. Feldman, & B. Rime (Eds.), Morrow, D. G., Greenspan, S. E., & Bower, G. H. Studies in emotion & social interaction (pp. 239- (1987). Accessibility and situation models in 281). New York: Cambridge University Press. narrative comprehension. Journal of Memory Rime, B., Schiaratura, L., Hupet, M., & and Language, 26, 165-187. Ghysselinckx, A. (1984). Effects of relative im- Murray, J.D., Klin, C. M., & Myers, J. L. (1993). mobilization on the speaker's nonverbal behav- Forward inferences in narrative text. Journal of ior and on the dial~gue imagery level. Motiva­ Memory and Language, 32(4), 464-473. tion and Emotion, 8, 311-325. 258 Psycholinguistics

Rumelhart, D. E. ( 1975). Notes on schemas for sto- zation of repair in conversation. Language, 53, ries. In D. G. Bobrow & A. M. Collins (Eds.), 361-382. Representation and understanding: Studies in Schegloff, E. A., & Sacks, H. (1973). Opening up cognitive science (pp. 211-236). New York: closings. Semiotica, 8, 289-327. Academic Press. Schiffrin, D. (1987). Discourse markers. Sacks, H., Schegloff, E. A., &Jefferson, G. (1974). Cambridge: Cambridge University Press. A simplest systematic~ for the organization Schober, M. F., & Clark, H. H. (1989). Understand- of tum-taking in conversation. Language, 50, ing by addressees and overhearers. Cognitive 696-735. Psychology, 21(2), 211-232. Sanford, A. J., & Garrod, S. C. (1994). Selec- Schourup, L. C. (1982). Common discourse par· tive processing in text understanding. In M. A. tides in English conversation. New York: Gemsbacher (Ed.), Handbook ofpsycholinguis­ Garland. tics (pp. 699-719). San Diego: Academic Press. Searle, J. R. (1969). Speech Acts. Cambridge: Schank, R. C., & Abelson, R. P. (1977). Scripts, Cambridge University Press. plans, goals, and understanding: An inquiry Searle, J. R. (1975a). Indirect speech acts. In into-human knowledge str!-(ctures. Hillsdale, NJ: P. Cole & J. L. Morgan (Eds.), Syntax and Erlbaum. semantics: Vol. 3. Speech acts (pp. 59-82). New Schegloff, E. A. (1972). Notes on a conversational York: Seminar Press. practice: Formulating place. In D. Sudnow (Ed.), Searle, J. R. (l975b). A taxonomy ofillocutionary Studies in social interaction (pp. 75-119). New acts. In K. Gunderson (Ed.), Minnesota studies York: Free Press. in the (pp. 334-369). Schegloff, E. A. (1980). Preliminaries to prelimi- Minneapolis: University of Minnesota Press. naries: "Can I ask you a question?" Sociological Discourse Inquiry, 50, 104-152. Sharron, B. (1984). Room descriptions. Processes, 7, 225-255. Schegloff, E. A. (1982). Discourse as an inter- Shepard, R N., & Cooper, L.A. (Eds.). (1982). actional achievement: Some uses of ''uh huh" Mental images and their transfonnations. and other things that come between sentences. In D. Tannen (Ed.), Analyzing discourse: Text Cambridge: MIT Press. and talk. Georgetown University Roundtable on Shillcock, R. ( 1990). Lexical hypotheses in contin- Languages and Linguistics, 1981 (pp. 71-93). uous speech. In G. T. M. Altmann (Ed.), Cog­ Washington, DC: Georgetown University Press. nitive models of speech processing: Psycholin­ guistic and computational perspectives Schegloff, E. A. (1984). On some gestures' relation (pp. 24- to talk. In J. M. Atkinson & J. Heritage (Eds.), 49). Cambridge: MIT Press. Structures of social action: Studies in con­ Singer, M. (1994). Discourse inference processes. versation analysis (pp. 262-296). Cambridge: In M. A. Gemsbacher (Ed.), Handbook of Cambridge University Press. psycholinguistics (pp. 479-5!5). San Diego: Schegloff, E. A. (1987). Recycled tum beginnings: Academic Press. A precise repair mechanism in conversation's Singer, M., & Halldorson, M. (1996). Construct- turn-taking organization. In G. Button & J. R. E. ing and validating motive bridging inferences. Lee (Eds.), Talk and social organization (pp. 70-- Cognitive Psychology, 30(1), l-38. 85). Clevedon, UK: Multilingual Matters. Slobin, D. I. (1996). From 'Thought and Lan- Schegloff, E. A. (1988). Presequences and indi- guage" to "Thinking for Speaking." In J. J. rection: Applying speech act theory to ordi- Gumperz & S. C. Levinson (Eds.), Rethink­ nary conversation. Journal ofPragmatics, 12(1 ), ing linguistic relativity (pp. 70-96). Cambridge: 55-62. Cambridge University Press. Schegloff, E. A., Jefferson, G., & Sacks, H. (1977). Sperber, D., & Wilson, D. (1986). Relevance. The preference for self-correction in the organi- Cambridge: Harvard University Press. References 259

Stalnaker, R. C. (1978). Assertion.ln P. Cole (Ed.), Vin den Broek, P. (1994). Comprehension and Syntax and semantics 9: Pragmatics (pp. 315- memory of narrative texts: Inferences and co- 332). New York: Academic Press. herence. In M.A. Gernsbacher (Ed.), Handbook Svartvik, J., & Quirk, R. (Eds.). (1980).A corpus of of psycholinguistics (pp. 539-588). San Diego: English conversation. Lund, Sweden: Gleerup. Academic Press. Swinney, D. A. (1979). Lexical access during sen- Venneman, T. (1973). Explanation in syntax. In tence comprehension: (Re )consideration of con- J. Kimball (Ed.), Syntax and semantics (Vol. 2, text effects. Journal of Verbal Learning and pp. 1-50). New York: Seminar Press. Verbal Behavior, 18, 645-660. Venneman, T. (1975). An explanation of drift. In Talmy, L. (2000). Toward a cognitive semantics. C. N. Li (Ed.), Word order and word order Cambridge: MIT Press. change (pp. 269-305). Austin: University of Texas Press. Tanenhaus, M. K., Leiman, J. M., & Seidenberg, M. S. (1979). Evidence for multiple stages in Wade, E., & Clark, H. H. (1993). Reproduction and the processing of ambiguous words in syntactic demonstration in quotations. Journal ofMemory contexts. Journal of Verbal Learning and Verbal and Language, 32(6), 805-819. Behavior, 18(4), 427-440. Walton, K. L. (1978). Fearing fictions. Journal of Tanenhaus, M. K., & Spivey-Knowlton, M. J. Philosophy, 75, 5-27. (1996). Eye-tracking. Language & Cognitive Walton, K. L. (1983). Fiction, fiction-making, and Processes, 11(6), 583-588. styles of fictionality. Philosophy and Literature, Tanenhaus, M. K., Spivey-Knowlton, M. J., 8, 78-88. Eberhard, K. M., & Sedivy, J. C. (1995). Inte- Walton, K. L. (1990). Mimesis as make-believe: gration of visual and linguistic infonnation in On the foundations of the representational arts. spoken language comprehension. Science, 268, Cambridge: Harvard University Press. 1632-1634. Wason, P. C., & Reich, S. S. (1979). A verbal Tannen, D. (1989). Talking voices: Repetition, illusion. Quarterly Journal ofExperimental Psy­ dialogue and imagery in conversational dis­ chology, 31, 591-597. course. Cambridge: Cambridge University Wasow, T. (1997). Remarks on grammatical Press. weight. Language Variation and Change, 9, Taylor, H. A., & Tversky, B. (1992). Spatial mental 81-105. models derived from survey and route descrip- Wilkes-Gibbs, D., & Clark, H. H. (1992). Co- tions. Journal of Memory and Language, 31(2), ordinating beliefs in conversation. Journal of 261-292. Memory and Language, 31(2), 183-194. Taylor, H. A., & Tversky, B. (1996). Perspective Wundt, W. M. (1900). VOlkerpsychologie. Eine in spatial descriptions. Journal of Memory and Untersuchung der Entwicklungsgesetze von Language, 35(3), 371-391. Sprache, Mythus und Sitte. Leipzig: Englemann. Ullmer-Ehrich, V. (1982). The structure of liv- Y ngve, V. H. (April 1970). On getting a word ing space descriptions. In R. J. Jarvella & W. in edgewise. Paper presented at the sixth Klein (Eds.), Speech, place, and action: Stud­ regional meeting of the Chicago Linguistic ies in deixis and related topics (pp. 219-249). Society, Chicago, IL. Chichester, England: Wiley. Zwitserlood, P. (1989). The locus of the effects Underhill, R. (1988). Like is, like, focus. American of sentential-semantic context in spoken-word Speech 63(3), 234-246. processing. Cognition, 32(1), 25-64.