Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Conversational initiations and responses in kindergarten by a blind child and his sighted identical twin Huyghebaert, Danine Andrea 1995

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1995-0167.pdf [ 4.62MB ]
Metadata
JSON: 831-1.0098947.json
JSON-LD: 831-1.0098947-ld.json
RDF/XML (Pretty): 831-1.0098947-rdf.xml
RDF/JSON: 831-1.0098947-rdf.json
Turtle: 831-1.0098947-turtle.txt
N-Triples: 831-1.0098947-rdf-ntriples.txt
Original Record: 831-1.0098947-source.json
Full Text
831-1.0098947-fulltext.txt
Citation
831-1.0098947.ris

Full Text

C O N V E R S A T I O N A L INITIATIONS A N D RESPONSES IN K I N D E R G A R T E N B Y A BLIND CHILD A N D HIS SIGHTED IDENTICAL TWIN by D A N I N E A N D R E A H U Y G H E B A E R T B . A . Hons., University of Saskatchewan, 1992 A THESIS SUBMITTED IN PARTIAL F U L F I L M E N T OF T H E REQUIREMENTS FOR T H E D E G R E E OF M A S T E R OF SCIENCE in T H E F A C U L T Y OF G R A D U A T E STUDIES The School of Audiology and Speech Sciences We accept this thesis as conforming to the required standard T H E UNIVERSITY OF BRITISH C O L U M B I A April 1995 ® Danine A . Huyghebaert, 1995 In p resen t ing this thesis in partial f u l f i lmen t of t h e r e q u i r e m e n t s f o r an advanced d e g r e e at the Univers i ty o f Brit ish C o l u m b i a , I agree that t h e Library shall make it f reely available f o r re ference and s tudy. I fu r ther agree that pe rmiss ion f o r ex tens ive c o p y i n g of th is thesis f o r scholar ly p u r p o s e s may b e g r a n t e d b y t h e head o f m y d e p a r t m e n t or by his o r her representat ives. It is u n d e r s t o o d that c o p y i n g o r pub l i ca t i on o f this thesis fo r f inancial gain shall n o t be a l l o w e d w i t h o u t m y w r i t t e n permiss ion . The Univers i ty o f Brit ish C o l u m b i a Vancouver , Canada D e p a r t m e n t DE-6 (2/88) 11 ABSTRACT The current study compares the conversational turns of identical twin boys, one of whom is blind. Previous research regarding the conversational interactions of children who are blind suggests an emergence of conversational initiation and maintenance strategies different from those of sighted children. This literature focuses on child-adult dyads; there is little information about the conversational interactions between blind children and their peers. Comparison of B's conversational skills with those of his sighted identical twin provides a unique research opportunity, as individual differences among children on such measures are so great that it is difficult to determine what differences should be attributed to the lack of vision per se. The case of comparing identical twins is as well controlled a study as is possible. The twins' conversational turns and utterances were transcribed from audio-videotapes of the twins interacting with peers in their kindergarten classroom and then coded for discourse categories relevant to initiations and responses. Discourse categories were analysed for each child at three different time periods over the year: (a) ages 4 years 11 months (4; 11) and 5;0, (b) 5;3 and 5;4, and (c) 5;6. Results indicate that although the number of interactions each twin attempted to initiate was similar, B devoted proportionally almost twice as many turns to initiation than did G. This difference seems to be related to the findings that 1) less than half of B's initiations I l l gained a response; 2) B persisted for a greater number of turns than G after an initial bid failed, due to repeated failed attempts; and 3) B used a large proportion of nonspecific and nonroutinized attention-getting devices, as well as simply relying on the addressee's proximity in the nonverbal domain. The results are consistent with reports that blind children have difficulty initiating conversational interaction (Dunlea, 1989; Kekelis & Andersen, 1984; Mulford, 1983; Rowland, 1983). Due to the single-subject design of this study, additional research is needed to advance us further toward being able to definitively answer the question of why this difficulty occurs and perhaps offer explanations of how it may be resolved. iv T A B L E OF CONTENTS ABSTRACT ii T A B L E OF CONTENTS iv LIST OF TABLES vi A C K N O W L E D G E M E N T viii CHAPTER ONE: INTRODUCTION A N D REVIEW OF T H E LITERATURE Introduction „ 1 Conversational Turn-Taking: A Review of the Literature 3 The Sequential Structure of Adult-Adult Conversation 3 Conversational Maxims 3 Conversational turn-Taking 4 The Relation Between Turns 6 The Development of Conversational Competence by Children Who are Sighted 10 Protoconversations 10 The Development of Initiations 13 The Development of Responses 15 The Organization of Conversational Turns 15 Factors Influencing Conversation 16 The Development of Conversational Competence by Children who are Blind 18 The Pre verbal Period 19 Nonverbal Attention-Getting Strategies 22 Verbal Attention-Getting Strategies 23 The Influence of Conversational Participants 26 Summary and Statement of the Problem 27 CHAPTER II: M E T H O D Overview of the Study 30 Subjects 30 Procedures for Data Collection and Analysis 31 Turns 33 Explication of the Coding Taxonomy and Considerations for Use 34 Addressee 34 Interactional Tier: Discontinuous Coding Categories 36 Interactional Tier: Continuous Coding Categories 41 Success in Gaining Listener Response 44 Verbal Utterance Characteristics 47 Nonverbal Utterance Characteristics 55 V CHAPTER III: RESULTS Question 1: Do the blind twin and the sighted twin differ in the extent to which they initiate conversational interaction? 61 Question 2a: Do the blind twin and the sighted twin differ in their devices for initiating interaction? 63 Question 2b: Do the blind twin and the sighted twin differ in their topic change devices? 68 Question 2c: Do the blind twin and the sighted twin differ in their devices in attention-getting turns? 71 Question 2d: Do the blind twin and the sighted twin differ in their devices for establishing a turn? 73 Question 3a: Do the blind twin and the sighted twin differ in their success in obtaining a response to conversational initiations? 77 Question 3b: Do the blind twin and the sighted twin differ in their success in obtaining a response to conversational openers? 80 Question 4a: Do the blind twin and the sighted twin differ in the degree to which they persist when an initiation fails? 83 Question 4b: Do the blind twin and the sighted twin differ in the degree to which they persist when an opener fails? 84 Question 5a: Do the blind twin and the sighted twin differ in the strategies they use when an initiation fails? 87 Question 5b: Do the blind twin and the sighted twin differ in the strategies they use when opener fails? 87 Question 6: Do the blind twin and the sighted twin differ in the extent to which they respond to previous utterances when a response is obliged? 89 Question 7: Do the blind twin and the sighted twin differ in the extent to which they respond to previous utterances when a response is not obliged? 89 Summary of Results 90 CHAPTER IV: DISCUSSION Data Interpretation Considerations 93 Initiation of Conversational Interaction 94 Verbal and Nonverbal Characteristics of Conversational Beginnings 97 Success in Gaining a Response 102 Persistence 104 Strategies used when an Initiation or an Opener Failed 105 Responses 106 Social Conversational Patterns of B and G 107 Challenges of the Classroom and Peer Conversation 110 Educational Implications 112 Implications for Future Research 113 Conclusion 115 REFERENCES 117 vi LIST OF TABLES Table Page 2.1 Number of utterances and turns for each twin for each time period 32 3.1 Total number of discontinuous turns 60 3.2 Total number of continuous turns 61 3.3 Number of turns (and %) of total turns coded as initiation bids 62 3.4 Verbal devices used in initiation turns 64 3.5 Number (and %) of verbal device types for initiation turns 66 3.6 Nonverbal devices used in initiation turns 66 3.7 Number of nonverbal devices used for initiation turns 67 3.8 Number (and %) of verbal devices used in conjunction with contextual support or nonverbal devices for initiations 67 3.9 Number (and %) of opening turns for each twin at each time period 68 3.10 Nonverbal devices used in openers 69 3.11 Verbal devices used in openers 70 3.12 Number (and %) of verbal devices used in conjunction with contextual support or nonverbal devices for openers 71 3.13 Verbal devices used in attention bids 72 3.14 Nonverbal devices used in attention bids 73 3.15 Verbal devices used in turn bids 74 3.16 Nonverbal devices used in turn bids 75 3.17 Number (and %) of verbal devices used in all discontinuous turns 76 V l l Table _ Page 3.18 Number (and %) of nonverbal devices used in all discontinuous turns 77 3.19 Responses to initial initiation turns 78 3.20 Responses to reTinitiation turns 79 3.21 Number (and %) of initiations responded to on a first or subsequent try 79 3.22 Responses to initial opening bids 80 3.23 Responses to openers after modification 82 3.24 Number (and %) of openers responded to on a first or subsequent try 82 3.25 Number (and %) of discontinuous turns responded to 83 3.26 Number of $mod, $elb, $rep, $att turns following $ini turns that (a) received no response, (b) received a noncontingent response 85 3.27 Number of $mod, $elb, $rep and $att turns following $opn turns that (a) received no response, (b) received a noncontingent response 86 3.28 Strategies used for an initial initiation after no response, a noncontingent or negative response or nonverbal attention 87 3.29 Number (and %) of strategies used in a subsequent attempt when an opener received no response, a noncontingent or negative response or nonverbal attention 88 3.30 Number (and %) of strategies used in subsequent initiations and openers after receiving no response, a noncontingent or negative response, or nonverbal attention 88 3.31 Number (and %) of responses in obligatory contexts actually given 89 3.32 Number (and %) of nonobliged responses 90 A C K N O W L E D G E M E N T I would like to express my sincere gratitude to my thesis supervisor, Dr. Carolyn Johnson, for her discourse expertise and constant support. Your tireless reading and re-reading of the manuscript and words of encouragement were much appreciated. Thanks also to Dr. Judith Johnston and Dr. Barbara Bernhardt for their contributions to this thesis. I am grateful to my peers in the program for their friendship and encouragement. A special thanks is owed to Kim who always provided computer troubleshooting services with a smile. Another thanks for computer aid is owed to John Nicol. I am especially appreciative for the loving support of my parents, Delvyn and Dolores Huyghebaert. Words can not express how much strength and courage you provided with your constant reminders to believe in myself and in my capabilities. Finally, I would like to thank my fiance, Wally Sawchuk, who's patience, understanding and love are truly limitless. You helped me maintain perspective and in doing so, kept my life focused. 1 CHAPTER I INTRODUCTION AND REVIEW OF T H E LITERATURE Introduction For sighted individuals, eye contact, direction of gaze and gesture do much conversational work, such as securing, maintaining and monitoring a listener's attention, marking the initiation and organization of conversational turns, and establishing and maintaining the topic under discussion. Much of the research regarding the conversational interactions of children who are blind suggests a pattern of turn-taking and an emergence of conversational initiation and maintenance strategies different from those of sighted children. This literature focuses on child-adult dyads; there is very little information about the conversational interactions between blind children and their peers, despite the fact that conversational organization is influenced by the age of the conversational participants. The goal of this study is to contribute to the literature regarding the conversational competence of children who are blind by describing a blind child's conversational discourse with his peers in a kindergarten classroom as it compares with that of his sighted identical twin. Of particular interest are comparisons of frequency of conversational initiation and degree of conversational maintenance, as well as the strategies used to initiate interactions and introduce new conversational content. The identical twin situation allows a unique opportunity to make such comparisons. As identical twins, B and G share physical, cognitive, environmental and some personality characteristics and as such, are as equivalent as is naturally possible. In addition, B and G are almost always in close proximity to each other, consequently sharing social interactions and language input to a large degree. Thus, they differ only in visual ability; B has no functional vision and G has functional vision. Due to the twin condition, this thesis is better able to credit differences to the ability to see than studies that compared children from different families, in which variables such as parenting styles, number of siblings and tempermental aspects of conversational style are difficult to control. Chapter I provides a brief review of relevant current research regarding what adults know about conversational initiation and maintenance, as well as the organizational principles underlying conversation, and more generally about the dynamics of conversation. Then consideration is given to how children, both sighted and blind, develop such conversational competencies, providing the motivation for the present study. The chapter ends with a statement of the research questions on which the study is based. Chapter II describes how the data was collected and analyzed. The twins, B and G, are introduced, and the discourse coding categories are defined and justified. Chapter III presents the results, and in Chapter IV the results are discussed in relation to each other and to recent literature on the conversational competence of children who are blind. 3 Conversational Turn-Taking: A Review of the Literature The Sequential Structure of Adult-Adult Conversation As adults, our knowledge about conversation is complex. Adults' communicative competence involves knowledge of: (i) communication intentions, including how to respond to communication intentions and how to express such communication intentions as, for example, coaxing, comforting and questioning, (ii) the situational conditions regulating speech acts (see Searle, 1975), (iii) general cooperative principles of conversation (Grice, 1975), and (iv) turn-taking (Sacks, Schegloff & Jefferson, 1974). The present study focuses on the sequential structure of conversation, more specifically, turn-taking, and how children develop their knowledge of how to initiate, maintain and terminate conversations, how to take and pass a turn, and how to introduce new topics. Conversational Maxims Communicative interaction occurs smoothly. To achieve such precision the participants must be guided by a set of underlying principles. Grice (1975) proposed the following four conversational maxims that cooperative speakers adhere to: (i) make your conversational turn no more and no less informative than is required, (ii) be truthful, (iii) be relevant, and (iv) speak clearly. But conversational competence must involve more than just knowledge of these maxims. A large part of the order and progression of conversation is due to the precisely managed alternation of turns. 4 Conversational Turn-Taking Obviously, speaker change does occur, but why? Ervin-Tripp (1979) attributes the occurrence of turns to the inherent interest of humans in the speech of others and to politeness factors, whereas Sacks et al. (1974) posit that turn-taking is more an economy-based organization. Turns are valued and, as such, must be distributed in a rule-bound manner. Ervin-Tripp (1979) reviews the general characteristics of adult-adult conversational turn-taking. The timing of turns is very precise, such that gaps between speakers are brief (i.e. 0.40 seconds in phone conversations, Brady, 1968, as cited in Ervin-Tripp, 1979) and, generally, only one speaker talks at a time. When speaker overlap occurs, it does so only in transition-relevant places (i.e., at the completion of a sentential, clausal, phrasal or lexical unit) and is very brief (Sacks et al., 1974). In addition, adults typically overlap only information that is not central to the conversation (i.e., is predictable or redundant, Ervin-Tripp, 1979). On the rare occasion that important information is overlapped, the overlap is repaired. In the following example, angle brackets indicate overlapped speech. Desk: What is your last name < Loraine > . Caller: <Dinnis>. Desk: What? Caller: Dinnis. (Sacks et al., 1974, p.702) Sacks, Schegloff and Jefferson (1974) posit that there must be some basic aspect underlying all conversations. They propose that this is the turn-taking system, which they describe as consisting of two components and a set of rules. The first component of the system is the turn-constructional component, which dictates that unit types are sentential, clausal, phrasal, and lexical. In contrast, Goffman (1976) argues that the turns of natural 5 conversation are composed of more varied components than just verbal units. A turn may be a verbal unit, a gesture, or a nonlinguistic action. For example, the request "shut the door" may elicit a verbal response (e.g. "okay") and the nonlinguistic action of closing the door, or it may elicit only the nonlinguistic action. Goffman considers this nonlinguistic action, whether accompanied by a verbal responses or not, to be a turn. If nonlinguistic actions function as turns just as verbal units do, they also influence the organization of conversation. The second component proposed by Sacks et al. (1974) is the turn-allocation component, which dictates the use of two techniques. The next turn can be allocated by (1) the current speaker or (2) self-selection, according to the application of the following rules. Rule 1 contains three parts and applies initially at the first transition-relevance place of any turn. Rule 1(a) says that if the current speaker selects the next speaker in the current turn, then the current speaker must stop speaking, and the next speaker must speak next, transition occurring at the first transition-relevance place after next-speaker selection. Rule 1(b) says that if the current speaker does not select the next speaker, then any party may self-select, first speaker gaining rights to the next turn. Rule 1(c) says that if the current speaker has not selected the next speaker, and no other party self-selects under option (b), then the current speaker may (but does not need to) continue. Rule 2 applies at all subsequent transition-relevance places and says that when rule 1(c) has been applied by the current speaker, then at the next transition-relevance place Rules 1 (a) through (c) apply, and recursively at the next transition-relevance place, until speaker change is effected. Sacks et al. (1974) maintain that gaps and overlaps are minimized because transfer of turns can only occur at transition-relevance places. In addition, turns are allocated to only a 6 single individual. However, as Goffman (1976) points out, there are many examples of common conversational occurrences, such as asides and parenthetical remarks, in which conversational participants speak at places other than a transition-relevance place without having gained the floor. Sacks et al. (1974) argue that the turn-taking system is highly regular and precise due to local management. That is,- the system operates on a turn-by-turn basis, organizing just the transition from current speaker to next speaker using the turn-constructional component, turn-allocation component, and the ordered and cyclic rules. Furthermore, Sacks et al. present their turn-taking organizational system as basic (i.e., common to all conversation) and thus as context-free, but also as capable of being context-sensitive. In contrast, other researchers (Goffman, 1976; Camaioni, 1979) argue that the turn-taking organization of natural conversation is much more flexible than Sacks et al.'s (1974) framework implies, and that consideration of context is critical to description of the organization of conversation. Specifically, Camaioni (1979) asserts that the structure of turn-taking is influenced by the type of interaction (determined by comparing participants' ages, quantity of world knowledge, and sociocultural backgrounds) and the social situation (e.g., a teaching situation versus a play situation). The Relation between Turns According to Schegloff and Sacks (1973), the adjacency pair is the fundamental unit of conversational organization. The basic or core adjacency pair is defined as consisting of adjacent parts, with each part produced by a different speaker. The parts are typed, such that a particular first part requires a particular second part. For example, the following is a greeting-greeting adjacency pair: A: Hi . B: Hello. There are many other possible adjacency pairs as well. Another example is the question-answer adjacency pair: A: What time is it? B: It's almost two o'clock. The notion of adjacency pair is further modified such that larger stretches of text may be organized around a core adjacency pair (Sacks et al., 1974; Schegloff & Sacks, 1973). Schegloff and Sacks (1973) propose presequences and insertion sequences. Presequences can occur before part one of the core adjacency pair and are meant to be understood as preliminary. A preinvitation is an example of a presequence: A: I wonder if you've got any plans Friday night? (preinvitation) B: Nothing special. A: Well, I wonder if you'd like to come for supper? (part one of the adjacency pair) B: Okay, (part two of the adjacency pair) Insertion sequences occur after part one of an adjacency pair, when there is a strong expectation that part two will follow. A clarification sequence is an example of an insertion sequence: A: Did you see Tory last night? (Question) B: Who? (Clarification request) A: Tory, that guy in biology class? (Clarification) B: Oh yeah, he was looking good. (Answer) 8 The adequacy of the adjacency pair as a structural unit for conversation has been challenged (Goffman, 1976; McTear, 1985). Goffman (1976) argues for a larger frame of reference in the analysis of conversation than Sacks et al.'s (1974) system, which operates on a turn-by-turn basis and hinges of the notion of the adjacency pair as a unit of conversation, as just described. Goffman (1976) demonstrates the need for a large frame of reference by showing that "second-pair parts" may have a reference that reaches much farther back than the immediately preceding turn. For example, take the following telephone conversation: A: Hello C: Is this the Y? A: You have the wrong number C: Is this KI five, double four, double o? A: Double four, double six C: Oh, I'm sorry A: Good-bye (hangs up) (Goffman, 1976, p. 285). In this example, Goffman (1976) posits that the apology "oh, I'm sorry" has as its reference the fact that "C" unnecessarily made "A" answer the phone. In addition to the problem of chaining between a first pair part and second pair part, researchers (Goffman, 1976; McTear, 1985) have argued that, in natural conversation, the distinction between a first-pair part and a second-pair part is often artificial. The following example nicely illustrates this point: George: Did you want an ice lolly or not? Zee: What kind have they got? George: How about orange? Zee: Don't they have Bazookas? George: Well here's twenty pence, you ask him. (Brown & Yule, 1983, p.230). 9 In this example, it is difficult to determine which interrogative moves are questions and which are answers. Similarly, McTear (1985) asserts that the adjacency pair is an unsatisfactory descriptive unit because it does not account for relations between all utterances in a conversation and is not defined adequately to distinguish between possible and impossible responses. McTear (1985) proposes a system of analyzing conversation based on a distinction between initiations and responses. Initiations are defined broadly such that they function to break continuity with the preceding discourse and predict a response, but can vary in the extent to which they are prospective. Initiations are marked prosodically by high pitch relative to the preceding utterances and by the use of response-eliciting devices such as vocatives and other attention-getting cues. Prerequisites for successful initiation of conversation include the abilities to gain and maintain attention, take into account the speaker's knowledge when formulating an initiation, and take the necessary steps to reinitiate if the original initiation is unsuccessful in gaining a desirable response (McTear, 1985). Reinitiations are defined as any utterance which attempts to elicit a response following no response or an unsatisfactory response. Reinitiations may involve whole or partial repetition of the initial initiation, with or without attention-getting devices, or rephrasing of the original initiation. Conversational competence also includes the ability to respond to another's initiations and to make one's responses contingent. McTear (1985) defines a response as an utterance which is predicted by a preceding utterance. Possible responses are distinguished from impossible responses such that the former support discourse expectation of a preceding initiation and the latter do not. McTear's (1985) conversational initiations and responses provide a framework for the coding scheme used in 10 this thesis. For the purpose of this thesis research, initiating moves will include turns that function to 1) initiate interaction, 2) change the topic of the interaction, 3) attempt to gain attention, and 4) attempt to gain a turn within multiparty interaction. In this thesis responses that support discourse expectation are classified as contingent, negative and responses from nonaddressee. Responses that do not support discourse expectation include noncontingent responses and null responses. In summary, conversational competence critically includes knowledge of the underlying principles that guide conversational participants, the general principles of turn-taking, and the ability to initiate and respond to conversational moves. The Development of Conversational Competence by Children who are Sighted Protoconversations When the infant is only two months of age, caregiver-infant exchanges already display a characteristic pattern (Trevarthen, 1977). Trevarthen describes the pattern of interaction in Euro-American middle-class culture as follows: The infant displays "prespeech" movements of the mouth, which are accompanied by "gesticulation," characteristic movements of the infant's hands and arms, and eye contact. When the infant looks away and stills, the mother responds by smiling and talking to the infant. These caregiver-infant exchanges are precisely timed, characterized by predominantly alternate moves and only brief periods of overlap (Trevarthen, 1977). The basis of the complex rules of turn-taking is already being learned. At this early stage, it is the caregiver who is shaping the exchange by way of responding to the actions and vocalizations of the infant as communicative. The infant's vocalizations and 11 actions are merely reflexive; that is, they are not used as intentional attention-getting devices. It is the caregiver who is responsible for capturing and maintaining the infant's attention through visual stimulation, and for sustaining the interaction (Bruner, 1977). In the typical family, caregivers work very hard to make their comments interesting to their infants, and they are very good at determining topics that interest their infants. Collis and Schaffer (1975) demonstrated that mothers follow their infants' eye gaze to determine objects of interest and then comment and label these referents (Collis & Schaffer, 1975). At the end of the first year, in the period between nine and twelve months, a new form of communication emerges with the development of the notions of reciprocity and intentionality (Bruner, 1977; Schaffer, 1977; Trevarthen, 1977). The beginning of reciprocity is marked with the development of giving of objects (Bruner, 1977). Interaction between the infant and caregiver takes on the two-sided nature characteristic of true communication. The infant can now take both an initiating and a response role, thus enabling him/her to maintain communicative interaction. During this same time infants develop the related notion of intentionality, the realization by children that their actions can be used to purposely influence the actions of others in order to achieve their own goals (Schaffer, 1977). At approximately nine months of age the infant uses eye gaze, crying, vocalization and smiling to intentionally gain attention to self (Bates, Camaioni & Volterra, 1975; note that Schieffelin, 1975, gives the age of these developments as one year). However, there is much more to intentionality than being able to direct attention to self. Sugarman (1983) asserts that it is not until ten to twelve months that the child is able to coordinate goal-oriented behaviour involving external objects and attention-getting, thus indicating intention to communicate. For 12 example, a reaching or pointing gesture towards a desired object in conjunction with direct solicitation of adult aid (through eye contact or vocalization) would constitute an intention to communicate. Schaffer (1977) also deems as essential for the development of dialogue the cognitive developments of differentiation of self and other, object permanence and representation skills. Differentiation of self and other is often interpreted as established when the child is able to correctly use the pronouns "I" and "you." Sighted children begin to use these pronouns between the ages of 12 to 24 months, with full discrimination not occurring until 30 to 36 months. Piagetian theory places the development of object permanence at 10 months of age and the development of representational skills, evidenced by the occurrence of deferred imitation and make-believe play, at 18 to 24 months (Gleitman, 1986). [Note that there is an extensive literature on object permanence and representation skills and on the limitations of Piagetian theory, which is beyond the context of this discussion]. Some researchers (Bruner, 1977; Snow, 1977; Trevarthen,1977) assert that the caregiver, by attributing communicative significance to the vocalization and actions of the infant, is actually shaping the child into a communicative being. Snow (1977) describes the development of conversation through the caregiver's selective responsiveness to certain infant actions and vocalizations. When the Euro-American middle-class infant is three months of age, the caregiver responds to burps and yawns as communicative, meaningful turns. Later, at seven months, the caregiver is no longer responsive to reflexive burps and yawns, but does respond to the infant's babble. Thus, over time the caregiver raises expectations as to what will be regarded as a contribution by the infant. However, Sugarman (1983) argues against 13 the implied notion that, through their responsiveness to preverbal vocalization, caregivers are "teaching" infants to communicate. She states as evidence Schieffelin's (1976) report on Kaluli mothers, who do not respond to their infants' preverbal vocalization as communicative, but whose infants do learn language. Bruner (1977) emphasizes the context of mother and infant carrying out joint actions with objects for the learning of communicative skills. Through repetition, simple object-oriented games such as give-and-take and games involving anticipation become highly familiar and conventionalized. Bruner (1977) asserts that these familiar frameworks provide a safe context for the child to take a next step and continue to make successive advances. For example, at first the infant takes only a passive role in the give-and-take game; she can only "take" the object from the caregiver. Later, the infant takes the role as initiator, as she can now "give" the toy to the caregiver. As the infant becomes more confident with her role as "giver" she is able to sustain the interaction for longer periods of time. Soon she is comfortable enough with her role that she can introduce variations. Now the infant begins to combine vocalization and gesture. Still later she is ready to take the next step and insert words into these routinized interactions (Bruner, 1977). The Development of Initiations The use of attention-getting devices is essential when trying to begin a conversational interaction with another, when introducing new discourse, or when reintroducing previous content into the discourse (Keenan & Schieffelin, 1976). There are both nonverbal and verbal attention-getting devices. As mentioned earlier, eye gaze and crying are the earliest 14 developing attention-getting devices, with intentional use emerging in the child between nine months and one year (Bates et al., 1975; Schieffelin, 1975). Gaze direction and eye contact serve not only to gain attention, but also to differentiate between addressees, and to check the attention status of a listener (McTear, 1985). Other nonverbal attention-getting devices are smiling (used in conjunction with eye-gaze), touching, orientation towards a listener, and movement towards a listener (Keenan & Schieffelin, 1976). Vocal attention-getting devices consist of crying, vocatives, expressives (e.g. "uh oh," "ouch," "wow"), expressions such as "you know what?" and "guess what," and general addresses such as "hey" used in conjunction with eye-contact (Keenan & Schieffelin, 1976). Vocatives are especially useful, functioning not only to gain general attention, but to specify "who should attend"; as such, they are especially useful in multiparty conversations. Vocatives also function to check the status of a listener's attention and to maintain attention (Keenan & Schieffelin, 1976). To be able to initiate a conversational exchange, a child also has to be able to direct a listener's attention to a referent entity (McTear, 1985). Pointing and showing are the most frequently utilized attention-directing strategies (McTear, 1985; Wellman & Lempers, 1977). With regard to the development of attention-directing strategies, Wellman and Lempers (1977) investigated ten children between 2;2 and 3;0 in the naturalistic context of a play group. The children were observed to use a wide variety of attention-directing strategies, including pointing, showing, displaying, bringing a referent entity to the addressee so as to establish his/her attention to the object, and leading the addressee to the referent entity. The children were successful in directing another's attention eighty percent of the time. 15 The Development of Responses Another important aspect of conversational competence is the ability to maintain conversation. By two years of age children are able to respond to another's questions and to other adjacency pair first-parts; however, their contingency skills are poor (Ervin-Tripp, 1979). In fact, Bloom, Rocissano & Hood (1976) reported that seventy percent of two-year-olds' responses were adjacent, that is, simply occurred immediately after an adult response. However, the majority of the two-year-olds' responses were noncontingent (sixty percent) and another good proportion were repetitive responses (fifteen percent). Greater response contingency (forty-five percent) and few repetitive responses (less than five percent) were observed in the three-year-old children. McTear (1985), in a review of the literature, found a general trend towards greater response contingency with children's increasing age. Fey (1986) states that the ability to maintain and extend conversation when not required to do so demonstrates greater understanding in the child of her responsibility to be contingent and keep the conversation going than does merely answering a question to maintain conversation. The Organization of Conversational Turns Important changes in the organization of conversational turns occurs between the ages of two and four years of age (Camaioni, 1979; Ervin-Tripp, 1979). At two years of age, turn-taking is regulated purely by the children's interest in what their conversational partners have to say (Ervin-Tripp, 1979). Among peers, conversational turns are mainly simultaneous; that is, "two actions or behaviours overlap in time" (p.329) and revolve around a shared-attention 16 focus or shared actions (Camaioni, 1979). Camaioni offers an example of one child (BI), who pretends to talk on the telephone, while another child (B2) watches. As soon as BI hangs the telephone up, B2 picks up the receiver and begins to talk on the telephone (p. 329). Thus, the conversations of two-year-olds consist of many nonverbal imitations where language plays a minor role in achieving and sustaining the interaction (Camaioni, 1979). Interaction becomes more language-focused as the child gets older, and at around four years of age contingent conversational turns are present in greater numbers in conversations between children (Camaioni, 1979). Verbal strategies are used in conjunction with nonverbal strategies or alone to initiate and sustain interaction (Camaioni, 1979). Camaioni offers the following illustrative example: BI: (approaching the telephone) Then I'll telephone the police. No, I'll phone. B2: (going up to BI and trying to snatch the telephone) BI: Oh no! but there are two telephones, (looks and smiles at B2). B2: (smiles) (1979,p.330) Factors Influencing Conversation Research comparing child-child and adult-child conversations provides evidence that child-child conversations have a different structure than adult-child conversations and, thus, also display different conversational rules (Camaioni, 1979; Ervin-Tripp, 1979). Adult-child conversations demonstrate few overlaps or lengthy gaps, due to adult control of conversational turns. In contrast, child-child conversations, in general, are characterized by longer gaps and more overlap. The degree of appropriateness of overlap timing increases with age, and at about age 4;6, children start to repair overlapped information (Ervin-Tripp, 1979). 17 Wellman & Lempers (1977) observed preschool children in a play group and found that eighty percent of communicative interactions were adult-directed as opposed to peer-directed. In addition, child speakers were observed to be more sensitive to listener needs when speaking to an adult as opposed to a peer. Camaioni (1979) reports on the nature of the conversations of children two and four years old with their mothers. It seems that three types of conversational units recur. One recurring pattern is that of child action followed by adult comment. Similarly, the child may vocalize, upon which the adult comments. Camaioni posits that, although both of these conversational units appear to be initiated by the child, it is predominantly the adult who initiates. Camaioni reasons that the majority of child actions or verbalizations are not meant to elicit conversational interaction. Instead, it is the adult who actively tries to lure the child into conversational interaction by commenting on the child's action or verbalization. McTear (1985) reports that peer conversational interaction is characterized by more reciprocity than conversational interaction between a child and an adult. Negotiation for access to a game or for a turn in a conversational interaction is often necessary in peer interaction. The child is no longer automatically allowed to be a participant in interaction or encouraged to be part of a conversation, as in exchanges with adults. Peer interaction also differs from child-adult interaction in that peer interaction often involves fantasy play, in which there exist fewer constraints on allowable conversational initiations and responses. Peer interaction also involves more conversational breakdown due to communicative immaturity, and thus necessitates more reinitiations. 18 The number of conversational participants also influences conversational interactions. Ervin-Tripp (1979) states that three-or-more party conversations are more difficult than dyads for several reasons. First of all, three-party conversations involve more complex monitoring skills in order to keep track of what participants say. In conversations involving several participants, one must wait longer for a turn, creating the possibility that a young child may forget what it was that she wanted to say. In addition, more speakers result in greater competition for a turn, so one must work harder to achieve a turn and overlap is more likely. Thus, in a three-party conversation, children younger than 4;6 displayed difficulty in gaining attention, and consequently succeeded in interruptions only 12% of the time. Children between the ages of 4;6 and 6;0 were successful in gaining a turn 27% of the time (Ervin-Tripp, 1979). Thus, the ability to successfully gain a turn in a three-party conversation seems to develop with increasing age. The above research regarding the development of conversational initiations and responses in sighted children provides a basic framework for defining and coding conversational exchanges by blind children and their peers. Because eye contact, eye gaze and gesture are often invoked, it also raised interesting questions about blind children's conversational development. The Development of Conversational Competence by Children who are Blind Blind children are widely reported to have difficulty gaining listener attention and initiating conversational interaction (Dunlea, 1989; Kekelis & Andersen, 1984; Mulford, 1983; Rowland, 1983). In addition, although Urwin reports an "unusually early mastery of 19 basic conversation-maintaining procedures" (1978:p.l49), Rowland (1983) states that by the end of her study only one of the children at the age of 1;9 was able to maintain conversational interaction. The other two children, aged 1;5 and 1; 10, had to be prompted for each conversational turn. The Preverbal Period The above review of the development of conversational competence in children who are sighted suggests several prerequisites to the development of successful conversational skills. However, studies of children who are-blind (Mulford, 1983; Rowland, 1983; and Urwin, 1978, 1983) provide evidence that what have been labelled prerequisites may be merely precursors to conversational development in the sighted population. These studies suggest an alternate route to development in blind children, one that perhaps proceeds at a different pace. Urwin (1978, 1983) investigated the early dialogue of two children, one child defined as totally blind and the other child as possessing limited vision in one eye, from the preverbal stage until the age of 1;8. She also reports data from a third subject, described as totally blind, who was observed from the preverbal stage to the age of 1;10, with additional sessions at 2;0 and 2;3. Note that in the following discussion the age equivalents of developmental milestones by sighted children have been placed in square brackets behind Urwin's (1978, 1983) findings for purposes of comparison. With regard to the development of dialogue, Urwin (1978) reports the emergence of object permanence at 1;6 [compare 10 months for sighted], shown by the occurrence of 20 pretend play and search behaviour. The child with limited vision in one eye is reported to have used words to demand objects at 1;6, and showed evidence of reaching to demand objects at 1;8, both evidence of the emergence of intentionality [compare 1;0]. All three children are reported to show evidence of the development of the concept of reciprocity by 1;8 [compare 1;0]. The verbal role-play of one of the blind children, age 1;6, is taken to be evidence of the ability of this child to differentiate self and other [compare l;0-2;0/ 2;6-3;0 for full discrimination]. Andersen, Dunlea, & Kekelis (1984) also found evidence of pronoun distinction in the verbal role-play of the children they studied; however, evidence of pronoun confusion in these children's conversational language suggested that differentiation of self and other had not emerged. Andersen et al. also submit that the pronoun confusion supports the contention that the children were not able to take another person's perspective and, thus, reciprocity could not be said to be achieved. With regard to representational skills, Urwin argues that, although her totally blind subject did not request objects until the age of 1;9, "semantic analysis of her language at this time provided examples of each of the two-term semantic relations which Brown (1973) has argued are related to, and depend on, the emergence of representation" (p. 151) [compare 1;6 -2;0]. Urwin (1983) asserts that naming or requesting distant objects is a more complex development for children who are blind than for sighted children, who are cued by the sight of objects in the environment. She argues that a parallel development in sighted children would be naming or requesting an object hidden from view. 21 The preverbal interactions of mothers and their blind children have been described as largely uncommunicative (Adelson, 1983; Rowland, 1983). Rowland (1983), who studied the vocal and verbal behaviours of three blind infants and their mothers, reports that the mothers of blind infants rarely responded to their infants' vocalizations, responding more readily to infant gesture. On the other hand, Urwin (1983) argues that, despite no opportunity for mutual eye-gaze, preverbal conversational interactions do take place, involving a different structure and different nature than those of sighted infants and their caretakers. For example, early preverbal conversations between blind children and their caretakers often involve tactile forms of interaction, such as touching and tickling. Rowland (1983) implies that, because blind children's vocalizations do not receive consistent response in the preverbal period, turn-taking behaviour may fail to be adequately shaped. Similarly, Kekelis and Andersen (1984) studied the communication styles of four blind children (two totally blind and two with residual vision) and two sighted children and their families. The families of the sighted children used a high proportion of descriptive language and often expanded their children's utterances. In contrast, the families of the blind children displayed a frequent usage of requests for action and repetition, as well as a high proportion of adult-initiated topics, which tended to inhibit the blind children's responses. Kekelis and Andersen propose that such family communicative styles develop due to limitations imposed by the blindness. They conclude that "the limited focus of their conversations denies young blind children opportunities to develop discourse skills necessary for satisfying interactions" (p. 63). 22 Continued unsuccessful or negative experiences with initiating conversational interactions may result in blind children refraining from conversational interaction (Mills, 1993) and, thus, social interaction. In such a case, the child would miss out on valuable language input and conversational interaction practise. Nonverbal Attention-Getting Strategies Children who are blind possess a limited repertoire of nonverbal strategies for gaining listener attention and initiating conversational interaction (Mulford, 1983; Parke, Shallcross & Anderson, 1980; Rowland, 1983). Nonverbal strategies such as eye-contact, eye-gaze and pointing are not available for children who are blind. Mulford (1983) reports that her three congenitally blind subjects observed at preschool/kindergarten and at home between the ages of five and six years of age, utilized touching/holding and proximity to gain listener attention. Although tactile strategies and proximity provide blind children with information regarding the physical presence of a listener, attention is not guaranteed (Mulford, 1983). Some blind children develop inappropriate initiation devices such as pinching, which are likely to be neither well received nor successful (Mulford, 1983). With regard to nonverbal conversational0strategies of older children, Parke, Shallcross and Anderson (1980) compared the nonverbal communicative behaviours of 30 legally blind and 30 sighted students between the ages of 5;8 and 15;10. The students who were blind used shorter duration head nods and used these significantly less often than sighted students. Since head nod is used to regulate conversational turns, Mills (1993) posits that this less frequent use of head nod by blind children may suggest a different pattern of turn-taking by the two 23 groups. Blind students also spend more time smiling than sighted students (Parke et al., 1980). Smiling displays a readiness to communicate and functions as an invitation for conversation (Trevarthen, 1977). Extended smiling, functioning as an invitation for others to initiate conversational interaction, may represent a good compensatory strategy for those who have difficulty initiating interaction. Verbal Attention-Getting Strategies Due to a limited repertoire of nonverbal attention-getting strategies, children who are blind necessarily rely heavily on verbal attention-getting strategies such as questions, requests for labels, vocatives, expressive particles, repetition, greetings, and routines (Dunlea, 1989; Mulford, 1983). Blind children rely heavily on questions to gain listener attention, initiate and maintain conversational interaction (Andersen, 1984; Andersen & Kekelis, 1982; Mulford, 1983). Andersen (1984) reports that blind children are often not sensitive to listener's needs, resulting in questions which appear to the listener to be "out of the blue." The following example of "out-of-the-blue" questions is from Alex, a five-year-old child who is blind: Adult: And you know it, don't you? You know your address already? Alex: (feels block) Is it something to eat? Adult: Is it something to eat? What? That? (looks at block). Alex: Yeah. Adult: No, it's just a plain old block. Are you hungry at all? Alex: Do you have a Feliciano record? (Andersen & Kekelis, 1982, p.24) 24 As is evident in the above example, the use of such self-centered topics often leads to communication breakdown (Andersen & Kekelis, 1982). Blind children also use requests for labels to gain listener attention and to initiate conversational interaction (Andersen & Kekelis, 1982; Kekelis & Andersen, 1984). Andersen & Kekelis (1982) report that an over-reliance on labelling as a strategy for conversational initiation increases the probability of communication breakdown. They provide the following example from Teddy (2;1): Adult: Teddy's gonna have some coffee cake. Teddy: Is this a table? (feeling table) Adult: Yes it's a table. You know it's a table. Teddy: Hi. Adult: Hi. Teddy: Is it the table? Adult: Yes, it's the table. • (P-22) As is demonstrated in the above example, some blind children use greetings inappropriately in the middle of a conversation. Such greetings often function to switch topic or addressee within discourse (Andersen, 1984). Vocatives, or calling a potential listener's name, are reported to be frequently used by blind children to initiate conversational interaction and to assess listener attention (Mulford, 1983; Urwin, 1983). Vocatives are an especially effective means of gaining the attention of a specific listener. However, when relying primarily on auditory cues and without the benefit of visual cues, blind children do face the problems of mistaken identity and mistaken proximity when they use vocatives. Thus, calling out a wrong name is a distinct possibility, especially when a blind child addresses an unfamiliar listener. Young listeners are especially 25 challenging to identify due to their high pitched voices, regardless of sex (Mulford, 1983). Blind children also tend to call the name of children who are absent, cued by a familiar context in which the person named usually participates (Mulford, 1983) or cued by voices from a distance (Urwin, 1983). Although some researchers view blind children's use of repetition as meaningless (Nagera & Colonna, 1965, as cited in Prizant, 1984), other investigators define repetition as a verbal strategy with communicative value (Prizant, 1984; Prizant & Booziotis, n.d.; Rowland, 1983; Urwin, 1978; Wills, 1979, as cited in Prizant, 1984). Prizant & Booziotis state that blind children use immediate repetition both communicatively, to initiate and maintain conversational interaction, and as a strategy to facilitate information processing. Delayed repetition is associated with routines and memorized chunks of language (Prizant, 1984; Urwin, 1978). Urwin (1983) reports that parents of blind children often use routines for such purposes as gaining attention, initiating and maintaining interactions and teaching conversational rules. Blockberger & Johnson (1985) investigated the use of routines by one blind child and his sighted identical twin between the ages of 1;0 and 1; 10. Results revealed a differential developmental trend with regard to the use of routines by the blind twin and the sighted twin. Over the observed eight months, as the blind twin gained a larger routine repertoire, he relied more heavily on routines for maintaining interactions. In contrast, over time the sighted twin participated in routines more rarely, joining only in the routines he initiated himself. Similarly, Urwin (1978) describes the declining usage of routines in favour of interactions revolving around object-centered activities in the social interactions of a child with limited 26 vision in one eye. The dependence on routines persisted in the interactions of the totally blind child and his caretakers. The blind child in both studies initiated few routines, although participating in many. Blockberger and Johnson (1985) note that routines that were self-initiated by the blind twin were largely unsuccessful due to difficulty in determining listener attention. Dunlea (1982) noted that the use of attention-getting strategies is highly adaptive for blind children, because it allows them to increase their participation in, and access to, their environment. The Influence of Conversational Participants Andersen (1984) reports that there are differences in the conversational interactions between blind children and adults and those of blind children and their siblings. The difference seems to be mainly one of accommodation. That is, adults tend not only to accept the blind child's instances of pragmatically inappropriate language but also to work very hard to interpret what it was that the child meant to say. In contrast, siblings are less accommodating; they do not accept instances of pragmatically inappropriate speech. In fact, siblings often challenge blind children to re-evaluate and revise their incorrect statements. Andersen and Kekelis (1982) hypothesize that this pressure from siblings may impact on blind children by forcing them to "stretch their communicative skills" (p.24). Mulford also postulates that blind children's peers would also display this less accommodating style of interaction: "Successful interaction with blind children requires some special skills in assessing their foci of attention and in interpreting their utterances from the perspective of a speaker without vision" (Mulford, 1983, p. 106). 27 Summary and Statement of the Problem In summary, for sighted individuals, eye contact, eye gaze and gesture do much conversational work such as securing, maintaining, and monitoring the listener's attention, assessing interest and other emotions, establishing and maintaining the topic under discussion and initiating and organizing conversational turns. Much of the research on the conversational interactions of children who are blind suggests a different pattern of turn-taking and the possible emergence of conversation-initiation and conversation-maintenance strategies different from those found for sighted children. This literature has focused on the conversational dyadic interactions between blind children and adults. There has been little investigation of conversational interaction between blind children and their peers, despite the fact that conversational organization is impacted by the status of the conversational participants. The bulk of conversational interaction among peers occurs in the classroom setting. Initiating and maintaining precise turn-taking in three- or more than three-party conversations is especially difficult for children because of the high level monitoring skills that are involved (Camaioni, 1979, Ervin-Tripp, 1979, Sacks et al., 1974). The difficulty is increased for blind children, who must monitor conversational moves without the aid of vision. In addition, there have been many studies that have looked at the factors involved in the development of conversational competence in blind and sighted children separately. However, there is no way to directly compare this information. The twin condition presents a unique opportunity to make such comparisons. 28 With this in mind, it is the purpose of this study to respond to the need for research of the conversational discourse of blind children in the classroom environment, with emphasis on conversational initiations and responses. In studying the classroom. conversational discourse of one young blind boy and his sighted identical twin brother the main underlying question is: Do the blind twin and the sighted twin differ in their initiation of conversational interactions with their peers in the classroom environment? In order to respond to this question a number of specific questions were posed: 1 . Do the blind twin and the sighted twin differ in the extent to which they initiate conversational interaction? 2. Do the blind twin and the sighted twin differ in their strategies: (a) for initiating interaction? (b) for opening bids? (c) for gaining attention within the interaction? (d) for establishing a turn? 3 . Do the blind twin and the sighted twin differ in their success in obtaining a response to conversational initiations, openers, or bids for turns? 4. Do the blind twin and the sighted twin differ in the degree to which they persist when an initiation or opener fails? 5. Do the blind twin and the sighted twin differ in the strategies they use when an initiation or an opener fails? 29 In addition, the following question is asked: Do the blind twin and the sighted twin differ in their ability to maintain conversational interaction with their peers in the classroom environment? In order to respond to this question the following specific questions were posed: 6. Do the blind twin and the sighted twin differ in the extent to which they respond to previous utterances when a response is obliged? 7. Do the blind twin and the sighted twin differ in the extent to which they respond to previous utterances when a response is not obliged? r 5 30 CHAPTER II M E T H O D Overview of Study This is a study of identical twins, one of whom is blind. The twins' conversational turns and utterances in peer conversations in Kindergarten were coded for discourse categories and analysed for the purpose of answering the questions posed in Chapter I. Subjects The subjects of this study are B and G, identical twin boys. The twins are part of a larger, longitudinal study of language acquisition initiated by Blockberger and Johnson (1985), when the twins were one year old. The twins were born at twenty-seven weeks gestation, approximately three months prematurely, and suffered several medical complications, including hypocalcemia, pneumothorax, hyaline membrane disease, and retinopathy of prematurity (ROP). Retinopathy of prematurity refers to a fibrous growth of blood vessels behind the lens of the eye and is described in terms of several stages of progression. G, the first born twin, was diagnosed with stage III ROP. Later, at the age of 2;6, he was diagnosed with limited vision (15 centimeters) in one eye, with good distance vision in the other eye. B is the second born twin. B suffered stage IV ROP with bilateral retinal detachments, resulting in total blindness. 31 Other than the visual impairment, the twins' neurological development and medical history are normal. The twins uttered their first words at approximately 1;7 and other linguistic milestones have occurred within normal limits. B and G are the only children of a North American, middle-class family. Their father is employed with a rental firm for construction equipment. Their mother had a career as a clerk until the birth of B and G, at which time she decided to stay home full time with the children. In addition to supportive and loving parents, B and G have frequent contact with members of their extended family as well as with friends of the family. Thus, B and G have grown up in a supportive and stimulating environment. Procedures for Data Collection and Analysis The data were obtained from a library of video- and audio-recordings and transcripts of B and G. Data collection on the twins began with the Blockberger and Johnson (1985) study and further data were collected by Johnson for the purpose of continued longitudinal study of the twins' language acquisition. Videorecordings were taped using a Panasonic WV-3240 videorecorder. Audiotapes were recorded using a Marantz PMD430 stereo taperecorder. Samson SR-2/ST-2/ECM 144 remote microphones were used to capture the twins' speech. A research assistant transcribed each tape as it was collected. Johnson then transcribed twenty percent of the corpus to establish reliability. For the purposes of this study, only taped samples of the twins' language in their Kindergarten classroom were of interest. A collection of transcripts was selected to span the 32 available corpus of data. Three time periods were selected for coding. Period One consists of the earliest tapes of the twins in their classroom environment, sampling their language at 4; 11 and 5;0. Period Two consists of two transcripts, sampling the twins' language at 5;3 and 5;4. Finally, a transcript, sampling the twins' language at 5;6 was selected as Period Three. For each of the transcripts, all interactions between each twin and one or more peers that occurred during that day in class were coded. Thus, unstructured free play and semi-structured activities involving peer involvement (e.g. peer conversation during art activities) were coded. Structured activities primarily involving interaction between a twin and an adult were not coded. However, if an adult intervened during the primarily peer interaction, the twins' initiations and responses to the adults were coded. The decision to code these conversational moves was made to preserve the continuity and naturalness of the conversation. Table 2.1 shows the number of coded utterances and turns for each of the twins for each time period. Table 2.1. Number of utterances and turns for each twin for each time period Discourse Period One Period Two Period Three Total Unit B G B G B G B G Utterances 129 243 235 137 152 106 516 486 Turns 112 231 221 127 144 103 477 461 33 Turns Transcripts were coded for interactional status at the level of the turn. For the purposes of this study a turn was defined according to Sacks, Schegloff and Jefferson (1974) and Goffman (1976). Thus, a turn was defined as consisting of at least one turn-constructional unit. A turn-constructional unit may be a verbal unit (i.e., sentence, clause, phrase, or lexical unit) or a nonverbal unit (i.e., gesture or nonlinguistic action). The boundaries of a speaker's turn were also defined. The beginning of a particular "Speaker A's" turn was marked according to the moment speaker A began to talk. The end of speaker A's turn was delimited by (a) a pause of at least half a second, (b) another speaker starting to talk, (c) a turn-terminating move (e.g. a question), or (d) nomination (by Speaker A) of another speaker (Sacks et al., 1974, rule la; see Chapter I for explication). Completion of talk usually occurs at a transition-relevance place (Sacks et al., 1974). However, a turn may also be incomplete, such that there exists trailing off of speech before the end of a turn unit. In addition, transcripts were coded at the utterance level for addressee and for verbal and nonverbal characteristics. For the purposes of this study an utterance is defined as typically coinciding with syntactic-phrasal structures, which are sometimes marked by extended pausal boundaries, a single intonation unit, and utterance-final intonation (see Brown & Yule, 1983). Inter-rater reliability scores were initially 90% for the addressee tier, 70% for the interactional tier, 72% for the success tier, 85% for the verbal characteristics tier, and 87% for the nonverbal characteristics tier. Code category definitions were refined until two coders could independently code at a rate approximating 100%. 34 Explication of the Coding Taxonomy and Considerations for Use The following coding taxonomy was motivated by the questions posed in Chapter I and previous research in the area of conversational competence. It consists of five tiers: (1) Addressee (2) Interactional Status (3) Success in Gaining Listener Response (4) Verbal Utterance Characteristics (5) Nonverbal Utterance Characteristics. Addressee (%add) The addressee tier indicates who the child intended as receiver of his message. The essential distinctions here were whether an addressee was selected at all, and whether a selected addressee was a child or an adult. Although peer interaction was the focus of the study, it was inevitable in the classroom that, at times, adults were involved or intervened in what were primarily peer interaction. Child-child conversations, as compared to adult-child interactions, are characterized by longer gaps between turns and more turn overlaps (Ervin-Tripp, 1979), a lesser extent of recipient design (Wellman & Lempers, 1977), more reciprocity, fewer constraints on allowable conversational initiations and responses, and a greater amount of communication breakdown (McTear, 1985). Thus, consideration of addressee became a relevant issue when defining and coding the twins' conversational moves. The Addressee tier was coded for the utterance and included the following categories: 35 X Y Z : Name of addressee NO A: No Addressee A M B : Ambiguous X Y Z Name of Addressee Coding criteria and contextual clues: When speech is directed to a specific person or persons, X Y Z is replaced by the first three letters in the participant code. For example, Jacqueline asserts, "you can borrow my pencil," and G responds "maybe I'll use Ian's." G's response was coded on the addressee tier with "JAC" . NOA No Addressee Coding criteria and contextual clues: When speech is directed to a general addressee or to no specific addressee when an interaction has not already been established. For example, B is walking around the classroom, hears voices and says, "hey." This utterance was coded on the addressee tier as NOA, as it was directed to the general group of children and not any child in particular. A M B Ambiguous Coding criteria and contextual clues: When it is uncertain to whom speech is directed within an interaction. For example, B has been talking with Natalie and occasionally with Nicky who has been mocking B. B says, "I'm in the house" which was coded as ambiguous because it is unclear whom B is addressing. 36 Interactional Tier (%int): Discontinuous Coding Categories Each turn-initial utterance was coded for the interactional function of the turn. The interactional function tier was divided according to Keenan & Schieffelin's (1976) notion of discontinuous and continuous discourse. Discontinuous codes represent turns that are unrelated to the content and/or presupposition of the prior turn. This definition captured the different types of beginnings that this study was interested in, compatible with McTear's notion of 'initiations'. Discontinuous coding categories were as follows: $ini Bid for interaction $trn Bid for turn $att Bid for attention $opn Opener $rep Exact repetition of a discontinuous turn $mod Repetition with modification $elb Elaboration $sel Noninteractional turns $ini Bid for interaction Coding criteria and contextual clues: When there has not been prior verbal interaction or there is not a group activity that carries a heavy expectation that there will be interaction; also when a conversational interaction is re-established after an interruption. For example, B and Natalie were engaged in a conversational interaction that was interrupted by activity regarding a fallen art object and by an interruption from B's aide. The aide leaves and B and Natalie remain sitting beside one another. B's turn, "I don't suck my thumb anymore," was coded as a bid for interaction. 37 $trn Bid for a turn "Bid for a turn" was a necessary code due to the conversational data involving mainly three-, or more, party conversations. In such conversational situations there can be great competition for conversational turns and a child must frequently work hard to achieve a turn (Ervin-Tripp, 1979). Coding criteria and contextual clues: When a turn is empty with regard to content and functions only to gain the floor or when a child must exert considerable effort to gain a turn (e.g., multiparty conversation in which a turn is not guaranteed). For example, B and G are sitting with a research assistant. G is telling the research assistant about a story book. G dominates the conversation, although B occasionally asserts his contribution. B's utterance, "and [/] and the owls an [/] an [/] an the owls" was coded as a bid for a turn because B needed to exert great effort to gain this turn. $att Bid for attention Having initiated a conversational interaction does not guarantee the undivided attention of your listeners for the whole of the interaction. Attempts to gain the attention of a participant are often necessary, especially in the context of a kindergarten classroom, where the children are easily distracted by the numerous activities that are going on at the same time. Bids for attention were defined as conversational moves that (1) occurred only within established interactions, (2) functioned to gain another's attention when that attention could not reasonably be expected to be secured, and (3) were marked by one or more attention-getting devices. 38 Coding criteria and contextual clues: A bid for attention is a turn which functions solely to gain attention to self in an established interaction. A bid for attention involves the use of an attention getting device (see description of verbal utterance characteristics below for a list of attention-getting devices). For example, B says, "I don't suck my thumb anymore. N querys, "You don't suck your thumb anymore?", and B responds, "No, not anymore. I'm a big boy now". There is a long pause and B calls out, "Natalie". This vocative was coded as a bid for attention. $opn Opener Included in conversational competence is the ability to initiate and extend a topic. Fey (1986) includes the frequent use of topic initiations and topic extensions as part of his profile of an "active conversationalist." Sometimes a topic is extended beyond an allowable limit and is thus labelled a tangential comment. Kekelis and Andersen (1984) report that the blind children they observed during the period between nine months and 3;0 tended to make frequent "out-of-the-blue," or tangential, comments. Due to accommodation, adult listeners often make sense of such off-topic comments. In contrast, siblings are less accomodating and, as such, conversational breakdown is more likely to occur (Kekelis & Andersen, 1984). Note that what is deemed an allowable topic extention depends on contextual factors. For example, in pretend play scenerios the limits of allowable topic comments are extended beyond that of everyday conversation. The opener code was included as a type of initiation, one which introduces a new topic within an established interaction. Tangential comments are also included under the 'opener' classification. 39 Coding criteria and contextual clues: A turn that is topically discontinuous with a preceding turn, or that re-introduces a topic that was previously on the floor. An opener is not a bid for interaction, a bid for a turn, a bid for attention, or an elaboration (i.e. does not function as an attempt to elicit a response upon a failed attempt). For example, B and K are playing with their teddy bears in the play hospital. K has been talking about her bear and continues, "I made him a doctor. Wanna feel him?" B responds, "okay". K adds, "Here's the knee thing. I put that in his coat that I put on. Here's the mask". B says, "cradle the baby". B's turn, "cradle the baby" was coded as an opening bid. $rep Exact repetition of a discontinuous turn Strategies used when a beginning failed were also coded. Exact repetitions, modified repetitions and elaborations were coded when an initial initiation, an opener, a bid for attention or a bid for a turn failed to elicit the desired response. McTear (1985) reports that one of the main strategies for reinitiations includes repetitions. Coding criteria and contextual clues: An exact repetition of a discontinuous turn occurs when a child repeats using the same words and same intonation and loudness level. For example, B is telling N about a pretend dog and asks, "who's that doggie back here?" N does not answer and B repeats "who's that doggie back here?" Since B's repetition utilized the same words as his previous turn plus the same intonation and loudness level, this turn was coded as an exact repetition. 40 $mod Repetition with modification McTear (1985) states that revisions (i.e., repetitions with modification) are another main strategy used by children to reinitiate. In addition to reinitiations, the modified repetition code was also used to mark alterations of openers, bids for attention, and bids for a turn. Coding criteria and contextual clues: A repetition with modification occurs when a discontinuous turn is not successful in gaining a response, so an alteration in presentation is utilized. Modified repetitions include changes of intonation and loudness. For instance, if, in the above example, B had repeated "who's that doggie back here?" in a louder voice, it would have been coded as a repetition with modification. The turn would also have been coded as a modified repetition if B had repeated, "who is that doggie?" $elb Elaboration/continuation This code was used to mark a rephrasing or elaboration, other than a repetition, of a discontinuous turn. Coding criteria and contextual clues: An elaboration was defined as the continuation of an initiation or opener. Continuation must occur after an opportunity of another to speak to be coded as an elaboration. For example, B uses the opener, "the doggie's not comin here today." There occurs no response and after a pause, B continues with "where's the doggie gone?" This turn was coded as an elaboration because B allowed sufficient time for a response but, because there was no uptake, he continued to expand his topic. 41 For the purposes of this study, opportunity for another to speak was defined according to Sacks, Schegloff and Jefferson's (1974) rule 1. In keeping with these rules, opportunity for another speaker to take the floor may occur whether or not there has been an actual interturn gap. A new turn occurs when Rule 2 applies. $sel Noninteraction turn This code functions to keep track of the amount of self-addressed speech in the twins' discourse. Coding criteria and contextual clues: A noninteraction turn occurs when a turn is self-addressed and carries no apparent communicative intent. Noninteractional turns are often said in a soft voice and serve as self-instructions or description of a task. For example, G cuts paper while working on an art project and mutters, "cutting circles." This was coded as a noninteractional turn because it was said with a soft voice and served to describe the task at hand. Interactional Tier (%int): Continuous Coding Categories Keenan and Schieffelin's (1979) notion of continuous discourse includes conversational turns that repeat the content of the previous turn or add new information that elaborates the previous turn. Thus, the notion of continuous discourse nicely describes the following different ways of maintaining conversation: $res Response $cnt Continued $req Required 42 $act Nonverbal action response A final category that was sometimes necessary is: $unc Uncodable $res Response A response represents a turn with continuation of content that is not obliged. Fey (1986) states that the ability to maintain and extend conversation when not required to do so demonstrates that a child understands his responsibility to be contingent and keep the conversation going. Being able to use a nonobliged response represents a more complex skill than does merely answering a question to maintain conversation. Coding criteria and contextual clues: A response constitutes a turn which represents a continuation of content with respect to the previous speaker's utterance. A turn may be semantically linked to the previous utterance, and/or repeat the content of the previous utterance and/or function as an acknowledgement of the previous utterance. For example, N and B are talking of Halloween and N says, "I like cats but I'm not being a cat." B says, "I'm gonna be a cat." This was coded as a response because N did not oblige a response from B. Sent Continued Coding criteria and contextual clues: The code continued is defined as a subsequent utterance within a turn, or a subsequent utterance across turns which does not take into account or recognize another speaker's turn. Thus, any utterance marked with the continued 43 code alone, would be a within-turn utterance. For example, a research assistant asked G, "do you have zippers in this pack too?" G gave the required response, "not [/] not lots" and added, "only one." The utterance, "only one" was coded as continued because it occurred within a turn. An utterance that does not take into account another speaker's turn and happens also to count as a separate turn (i.e., by virtue of the other speaker taking a turn) is doubly coded with the relevant interactional status code in addition to the continued code. For example, G initiated an interaction with the vocative, "Mrs. Cullen." Mrs. Cullen answered, "Hi G", and G continued, "Can I play in the house?" G's latter turn was coded as $ini$cnt. $req Required response A required response represents continuation of content that is obliged by the previous utterance, and thus applies mainly to answers to questions. Being able to answer another's question is a conversational maintenance strategy that develops by the time a child is two years old (Ervin-Tripp, 1979). Past research has indicated that blind children often receive adult input consisting of a high proportion of questions. It was interesting to consider if the blind twin would have a higher proportion of required responses in comparison to the sighted twin in peer conversation. Coding criteria and contextual clues: When a turn represents a continuation of content with regards to the previous utterance for which the turn is obliged. For example, N says, "G, where is my pencil?" and G responds, "what pencil?" G's turn was coded as a required response because he responded to N's question, in this case with another question. 44 $act Nonverbal action response Speech is not the only mode by which one may respond to another in conversational interactions. Nonverbal actions may also constitute a response. Coding criteria and contextual clues: Nonverbal action responses include the use of bodily actions in response to another's previous utterance. Nonverbal action responses include laughter and actions in response to a command. For example, G is commanded by C to "come here." G's action of walking over to C was coded as a nonverbal action response. $unc Uncodable Coding criteria and contextual clues: Uncodable utterances included whole or partially unintelligible utterances and utterances for which coding was not possible because the twin was not in the camera's view. For example, the turn, "I xxx make" where xxx represents an unintelligible word, and not enough of the utterance was intelligible to allow coding, was given a code of $unc. Success in Gaining Listener Response (%suc) The Success Tier was primarily motivated by Kekelis and Andersen's (1984) report of poor contingency of response in the speech of the blind children they studied, partly due to the prevalence of self-absorbed topics in these children's speech. While adult accommodation tends to prevent break down due to such tangential remarks, peers may not have the awareness or means to accommodate to such remarks. 45 Success is coded on the turn-initial utterance of discontinuous coding categories, but refers to the whole turn. The Success Tier included the following levels of success: $gre Successful in gaining a response Coding criteria and contextual clues: When a speaker's utterance receives a contingent response. A response is considered contingent if it relates to the speaker's utterance in form (e.g. repetition) or in content. Nonverbal action responses and clarification requests were also considered responses. For example, B's (4; 11) turn, "I want a doggie to come here," was coded $gre because it was followed by Natalie's contingent response, "then you're the dog." $gat Successful in gaining attention Coding criteria and contextual clues: When a speaker's utterance gains only a nonverbal attend. For example, a child makes the invitation, "Do you want to play?". A code on the success tier of successful in gaining attention would be appropriate if another child responded with eye gaze towards the speaker, unaccompanied by a verbal response or a physical action suggesting the intent to go play. $gre $gat $ncr $neg $rna $nor $unc Successful in gaining a response Successful in gaining attention Noncontingent response Negative response Response from a non-addressee Failure in gaining a response Uncodable 46 $ncr Noncontingent response Coding criteria and contextual clues: When the speaker's turn receives a response but the response is noncontingent. A noncontingent response corresponds neither to the form nor the content of the speaker's utterance. For example, B (5;3) and D play with a space station. B tries to place a spaceman in his spaceship and says to Davy, "Maybe he couldn't fit in there?" Davy responds, "We're up in the moon." B's turn was coded on the success tier as $ncr because Davy's response was judged to be noncontingent. $neg Negative response Coding criteria and contextual clues: When the speaker's turn receives a nondesirable response. Negative responses are a subtype of contingent (i.e., successful) responses. For example, N answers, "no way" to B's (4; 11) request, "pretend I was a baby." B's turn was coded $neg on the success tier. $rna Response from a nonaddressee Coding criteria and contextual clues: When the speaker's turn receives a response from someone other than the named addressee or the implied addressee (i.e. the participant of an ongoing conversational interaction). For example, B (4; 11) says, "Ah Natalie, a doggie's not in here." B is responded to by Nicky, as opposed to Natalie, so B's turn is coded $rna. 47 $nor Failure in gaining a response Coding criteria and contextual clues: "When a speaker's turn does not succeed in gaining a listener response when one is expected. For example, B and D are both playing with the space station toys but they are playing separately. B (5;3) says, "Hey you put a guy in there please." D does not respond, so B's turn is coded $nor. $unc Uncodable Coding criteria and contextual clues: When interaction ends due to the video or audiotape running out or due to change of camera focus. Verbal Utterance Characteristics (%vst) Many verbal utterance characteristics were chosen for coding, on the basis of being identified as predominant characteristics by prior research. The list is by no means all-inclusive. The verbal utterance characteristics chosen for coding included both attention-getting strategies and conversation /topic-maintaining devices. The attention-getting strategies were primarily motivated by a list compiled by Keenan & Schieffelin (1979) and Ochs, Schieffelin & Piatt (1979) and included the following: $nam Selects addressee by name $gen Selects addressee by general term $exp Expressive particle $que Question $grt Greeting $dei Deictic $pro Prosodic device 48 $loc Locating device $dir Directive $rou Routine $pre Pretend play $lab Request for label $oat Other attention-getting device The following maintaining devices were also coded: $coh Cohesion $agr Agreement/Disagreement $ran Repetition of another's utterance $row Repetition of one's own utterance $oma Other maintaining devices $nov No verbal strategy $vun Uncodable Verbal strategies were coded on each utterance. Thus, all utterances included in a turn were coded for verbal utterance characteristics. Any single utterance could be coded for more than one verbal utterance characteristic. A consequence is that a given turn might be counted as having a given characteristic even if not all the utterances in the turn have that characteristic. Investigation of the twin's usage of attention-getting strategies was motivated by reports that some blind children have difficulty gaining the attention of others (Blockberger & Johnson, 1985; Kekelis & Andersen, 1984) and by reports of differential frequency (Dunlea, 1989) and pattern of usage between blind and sighted children (Andersen & Kekelis, 1982; Blockberger & Johnson, 1985; Kekelis & Andersen, 1984; McGinnis, 1981; Mulford, 1983; Prizant, 1984, Urwin, 1978, 1983). 49 $nam Selects addressee by name Vocatives function to gain listener attention, initiate conversational interaction, and, especially in the case of individuals who are blind, assess listener attention (Urwin, 1983). Vocatives are one of the most frequently used attention-getting devices by children between the ages of 5;0 and 6;0 who are blind (Mulford, 1983). Coding criteria and contextual clues: When the speaker utilizes an addressee's name or title in an utterance which functions to attract listener attention, initiate conversational interaction, a turn, or a topic, or assess listener attention. For example, B (4; 11) initiated interaction with the vocative, "Cheryl," which was given the $nam code. $gen Selects addressee by general term Coding criteria and contextual clues: When the speaker utilizes a general term to select addressees. For example, "You guys, you need to help clean up." $exp Expressive particle Coding criteria and contextual clues: When the speaker utilizes conventional emotional reaction particles such as hey, uh oh, oh dear, ouch, woopsey, wow, and you know for the purpose of gaining another's attention or drawing attention to the self or an object/activity. For example, several children are playing in the teddy bear hospital. B has just entered the hospital and begins to play. There is a long pause and B (5;6) calls out, "hey I need a new mask!" 50 $que Question Blind children are reported to rely heavily on questions to initiate conversational interaction (Andersen, 1984; Andersen & Kekelis, 1982; Mulford, 1983) and use significantly more questions to change conversational content (McGinnis, 1981). Coding criteria and contextual clues: When the speaker requests information or clarification, for example, "what?" $grt Greeting All children use greetings to initiate conversational interaction; however, some blind children also use greetings in other ways. Some blind children use greetings to initiate new content or to get the attention of a new addressee in the middle of a conversation (Andersen, 1984; Andersen & Kekelis, 1982). Coding criteria and contextual clues: When the speaker utilizes a politeness routine that functions to welcome or acknowledge another, either at the beginning or within a conversational interaction. This category includes use of a greeting to initiate new content. For example, B has requested a game of pretend play from N , who has declined. B begins the pretend play and receives no response from N. In a "silly" high-pitched voice B (4; 11) says, "hi!" 51 $dei Deictic expression Deictics are an important device for referring to an entity. Many blind children have difficulty with deictics (Andersen & Kekelis, 1982; Mulford, 1983), which are used infrequently by blind as compared to sighted children between the ages of 3;5 and 5;0 (McGinnis, 1981). Difficulties with referential language may result in communication breakdown, especially when listeners are young and relatively unskilled in interpretation of the language of the blind. Coding criteria and contextual clues: When a speaker utilizes a spatial deictic term to indicate a referent to a listener, specifically, use of the demonstratives this, that, here, or there, for the purpose of drawing attention to a specific object or location. For example, B and A are putting blocks away and B (5;3) says, "I'll put those up there. " $pro Prosodic device Coding criteria and contextual clues: When the speaker raises pitch, increases loudness (e.g. shout) or decreases loudness (e.g. whisper) from a normal speaking voice. Nonfinal intonation contour (i.e. no pitch fall) was also coded as a prosodic device. Contrastive stress was not coded as a prosodic device, because its function is primarily semantic. $loc Locating device Coding criteria and contextual clues: When a speaker utilizes such devices as look at X, see X, or watch this to indicate a referent to a listener. For example, N and A play with blocks. B (5;3) plays nearby and says, "Lookit, I got a motorcycle." 52 $dir Directive Coding criteria and contextual clues: When a speaker instructs another to perform an action, even if the request is indirect. For example, C and B are in the teddy bear clinic. C greets B and then moves away. B (5;6) requests, "You have to give my bear some medicine cause he's sick." $rou Routine Routines are used to gain attention and to initiate and maintain conversational interaction. Blind children rely on routines to do conversational work to a greater extent than sighted children (Blockberger & Johnson, .1985; Urwin, 1978, 1983). Coding criteria and contextual clues: When a speaker uses a ritualized interaction for the purpose of attracting attention. For example, use of a knock knock joke to initiate interaction. $pre Pretend play Coding criteria and contextual clues: When a speaker is engaged in fantasy play. This category includes the use of animal or baby sounds. For example, N directs B to sit beside her and instructs, "you be the dog." B (4; 11) responds, "ruff." Slab Request for Label Blind children are reported to use an abundance of requests for labels to gain attention and initiate interaction. Requests are also used by children who are blind as a means of 53 gaining control over their surroundings (Andersen & Kekelis, 1982; Kekelis & Andersen, 1984). Note that this is a subtype of $que. Coding criteria and contextual clues: When a speaker requests the name of an object. For example, B and N are sitting beside one another. B reaches down, grabs N's ankle, and N responds with, "You're hanging [//] holding onto my ankles silly." B (4; 11) requests a label with, "Is this you toe?" $oat Other attention-getting devices The 'other' attention-getting device code was devised due to the awareness that the above list of attention-getting devices is not exhaustive. The 'other' code allows for the possibility of detection of additional attention-getting devices. Coding criteria and contextual clues: Includes any device that is used for the purpose of attracting attention and is absent from the above list. Maintaining devices serve to support the conversational interaction and/or preserve a topic. $coh Cohesion Coding criteria and contextual clues: Includes cohesive ties such as reference, conjunction, ellipsis and substitution, and lexical devices (e.g. repetition of a word, synonymy). For example, N instructed B "You're the doggie." B's (4; 11) response of, "You are N" was coded on the verbal characteristic tier for cohesion because it utilizes the cohesive ties of ellipsis and repetition. 54 $agr Agreement Coding criteria and contextual clues: When the child asserts that a previous utterance is true/untrue. Also affirmation, negation, or acknowledgement of another's assertion. For example, B's (4;11) response, "I'm not" to N's statement "you're the doggie" was coded for agreement. $ran Repetition of another Repetition is a maintenance device that is relied upon heavily by children who are blind (Prizant, 1984; Rowland, 1983; Urwin, 1983). Coding criteria and contextual clues: Included whole or partial repetition of another speaker' s utterance. $row Repetition of own Coding criteria and contextual clues: Includes whole or partial repetition of one's previous utterance; may be repetition of an immediately preceding utterance or repetition may be separated by several turns. Repetition may occur after not receiving a response or after gaining a response. For example, B (5;6) asserted and then repeated, "hey, I need a new mask!" 55 Soma Other maintaining devices The 'other' maintaining device code was devised due to an awareness that the above list of maintaining devices is not all-inclusive. The 'other' code allows for the possibility of detection of additional maintaining devices. Coding criteria and contextual clues: Includes any noticeable device, other than those listed above, used by the Speaker to support conversational interaction and/or to preserve the topic on the floor. $nov No verbal strategy Coding criteria and contextual clues: When an utterance is apparently lacking of any of the above verbal utterance characteristics. $vun Uncodable Coding criteria and contextual clues: When a whole or partial utterance is unintelligible to the extent that a verbal utterance characteristic is not identifiable. Nonverbal Utterance Characteristics (%nsf) Blind children are reported to possess and use a limited repertoire of nonverbal strategies for gaining listener attention and initiating conversation (Mulford, 1983; Rowland, 1983). The list of nonverbal utterance characteristics is by no means exhaustive and was mainly modelled on nonverbal attention-getting strategies listed by Keenan & Schieffelin (1976). Nonverbal utterance characteristics included the following categories: 56 $ctx $tou $ges $smi $eye $ast $non $nau Context Touch Gesture Smile Eye-contact Physical action No nonverbal strategy Uncodable $ctx Context Coding criteria and contextual clues: Includes utterances spoken in proximity to a potential Hearer, utterances said when a Speaker has bodily orientation towards a potential Hearer, and utterances spoken when a Speaker is making physical advance towards a potential Hearer. The speaker can reasonably assume his utterance is supported by such contextual features. $tou Touch Coding criteria and contextual clues: When the speaker comes into contact with or brings part of his body, especially a hand, in contact with a potential Hearer for the apparent purpose of gaining attention to the self. For example, N and A are playing with blocks and cars. B picks up a truck and leaves his spaceship toys to play nearer N and A. B moves towards the boys and reaches out his arm, touching N, while saying, "I ran out of gas." 57 $ges Intentional body gesture A differential pattern of use of gesture has been reported for children who are blind and children who are sighted. McGinnis (1983) reports that blind children do not use gesture alone in a communicative manner, although all sighted children use gesture alone to communicate messages. Parke, Shallcross & Anderson (1980) report differential use of head nod in blind and sighted children. Coding criteria and contextual clues: Includes showing, giving or pointing gestures. $smi Facial gesture Parke et al. (1980) report that blind children spend more time smiling than do sighted children. Coding criteria and contextual clues: Includes smiles and other facial gestures accompanying or used in lieu of a verbal message. $eye Eye contact Coding criteria and contextual clues: When eye contact is attempted but not necessarily established. Attempted eye contact is suggested by head orientation towards the intended target. Often a reorientation of eye gaze coincides with a new utterance. For example, G uttered, "Tyler", as he looked toward Tyler. 58 $ast Physical action Coding criteria and contextual clues: When the speaker uses bodily movements in apparent attempt to draw attention to self. For example, in the teddy bear hospital, B (5;6) played near a group of children who had their bears in a baby stroller. B made several unsuccessful hints that his bear needed a bed. He tried again to gain the other children's attention by jumping up and down while saying, "oh boy we're in a hospital." $non No nonverbal strategy Coding criteria and contextual clues: When an utterance is apparently lacking any of the above nonverbal utterance characteristics. $nau Uncodable Coding criteria and contextual clues: Used to mark instances where view of the child was obstructed so that it could not be determined whether eye-contact was established or if there was use of facial or body gesture/action. As such, the "uncodable" code does not always mean the absence of a nonverbal utterance characteristic. 59 CHAPTER III RESULTS To address the research questions posed, the discourse turns produced by each boy at each of the three time periods were analyzed according to the categories defined in Chapter II. These results are described below. Although raw numbers are given in tables, due to differences in the number of turns produced by each twin in each time period, proportions are also quoted to allow comparison. Table 3.1 and Table 3.2 summarize the total number of discontinuous and continuous turns produced by each twin at each of the three time periods. These turns represent the data used to answer the research questions. B produced almost an equal proportion of discontinuous and continuous turns. Thus, 194 (or 40.9%) of B's 477 total turns were coded as discontinuous, whereas 41.9% were coded as continuous. B's remaining turns were coded as self-directed (11.5%) or uncodable (5.7%). In contrast, G produced a larger proportion of continuous turns (59.2%) as compared to discontinuous turns (33.0%). In addition, 4.3% of G's turns were coded as self-directed and 3.5% could not be coded. 60 Table 3.1. Total number of discontinuous turns Code Period One Period Two Period Three Total B G B G B G B G $att 6 7 4 — 1 1 11 8 $ini 6 17 15 10 20 3 41 30 $ini$att — — — — 5 — 5 — $ini$cnt — 1 — — 1 — 1 1 $ini$elb 2 4 6 2 12 — 20 6 $ini$mod — 4 8 2 7 2 15 8 $ini$rep — 1 — — — 1 — 2 $opn 19 31 23 10 10 11 52 52 $opn$att 2 — 1 — — 1 3 1 $opn$cnt — 1 — 1 — — — 2 $opn$elb 10 4 10 3 1 3 21 10 $opn$mod 1 5 6 3 2 1 9 9 $opn$rep 1 — 1 — 1 1 3 1 $trn 2 10 8 2 3 10 13 22 Total 49 85 82 33 63 34 194 (40.7) 152 (33.0) 61 Table 3.2. Total number of continuous turns Code Period One Period Two Period Three Total B G B G B G B G $req 13 17 23 28 16 8 52 53 $res 28 77 56 47 40 42 124 166 $res$cnt 4 10 12 7 5 16 21 33 $act 2 15 1 6 — -- 3 21 Total 47 119 92 88 61 66 200 273 (41.9) (59.2) Question 1: Do the blind twin and the sighted twin differ in the extent to which they initiate conversational interaction? There are two ways to approach this question with the conversational data: (1) the number of interactions each twin attempted to initiate, and (2) the proportions of each twins' turns devoted to attempting to initiate interaction. In the first instance, the relevant turns are those coded as initial initiation bids ($ini), because further consecutive initiation turns ($ini$cnt and $rep, $mod, $elb and $att when they followed $ini without an intervening turn type by the same speaker) were part of the same general bid. In the second instance, all of the above turn types contributed to the total number (or proportion) of turns devoted to initiating interaction. 62 Table 3.3. Number of turns (and %) of total turns coded as initiation bids Turn Period One Period Two Period Three Total Type B G B G B G B G $ini 6 17 15 10 20 3 41(8.6) 30 (6.5) $rep — 1 — — — 1 — 2 (0.4) $mod — 4 8 2 7 2 15(3.2) 8(1.7) $elb 2 4 6 2 12 — 20(4.2) 6(1.3) $att — — — — 5 — 5(1.1) — $inicnt — 1 — — 1 — 1(0.2) 1 (0.2) Total 8 27 29 14 45 6 82(17.2) 47(10.2) Turns 112 231 221 127 144 103 477 461 In general, both twins initiated interaction relatively infrequently (see Table 3.3). Forty-one of B's turns and thirty of G's turns were coded as initial initiation bids, 8.6% and 6.5%of their 477 and 461 total turns, respectively. Exact repetitions, modifications, elaborations, and new turns that were continuations of initiations ($ini $cnt) were not included in this calculation, since they indicate persistence rather than frequency of attempting to establish interaction. In period one, B initiated fewer interactions than G, (6 as compared to 17, respectively). In period two, the twins initiated a roughly equal proportion of the time. Of B's turns, 6.8% were new attempts to initiate, compared with 7.9% of G's. The largest difference between the twins was evidenced in period three, where 20 of B's turns were new attempts to initiate as compared to 3 of G's turns, proportionally almost five times as many. 63 When turns indicative of persistence are considered, B devoted a greater proportion of turns to initiation then G; compare 17.2% and 10.2%. This difference is due almost entirely to period three, during which 31.3% of B's turns were attempts to initiate interaction, in contrast to G's 5.8%. Question 2a: Do the blind twin and the sighted twin differ in their devices for initiating interaction? Verbal and nonverbal devices were coded for each utterance belonging to any initiating turn so, in addition to initial initiation bids, devices for relevant within-turn ($cnt) utterances were also counted. Most of the twins' initiation turns consisted of single utterances, with B having a total of two within-turn initiation utterances. Table 3.4 demonstrates that each twin used a range of verbal devices in initiating turns. B used routines, whereas this device was entirely absent from G's initiations. On the other hand, G used greetings, which B did not use in initial initiation bids. B used pretend play and prosodic devices in initiations more frequently than G. G used questions for initiation more frequently than did B. Table 3.5 shows the number of verbal device types for initiation turns. In general, B was more likely than G to use no verbal device (although this difference is small) or several verbal devices together within each initiation turn. B had turns that combined three devices in all time periods, whereas G's initiation turns included a maximum of two. Table 3.4. Verbal devices used in initiations turns 64 Verbal Devices Period One Period Two Period Three Total B G B G B G B G Directive — 2 — — 5 3 5 5 Deictic — 4 5 1 1 — 6 5 Expressive 1 3 — 1 2 — 3 4 General addressee — 1 1 1 — — 1 2 Greeting — 2 — — — — — 2 Locating device — — — 1 1 — 1 1 Addressee name 4 5 — 1 7 2 11 9 Pretend play 1 1 5 — 14 — 20 1 Prosodic device 3 2 4 1 10 -- 17 3 Question 1 8 4 4 2 7 12 Repetition of own — 1 1 . — 1 — 2 1 Routine -- — 1 — 1 — 2 — Other devices — . -- — — — 1 — 1 No verbal device 1 2 5 — — — 6 2 # Initiation Turns 6 17 15 10 20 3 41 30 65 Each twin also used a range of nonverbal devices for initiation (see Table 3.6); however, B relied on context alone to a higher degree than did G; compare 24 out of 41 initiations (58.5%) for B and 14 out of 30 initiations (46.7%) for G. B also initiated using no nonverbal devices 19.5% of the time whereas G always used a device or could rely on context in initiations. As would be expected, B did not use eye contact or gesture for initiation. G used eye contact in 20.0% of his initiations and gesture in 6.7% of initiating turns. Many of B's turns did not have nonverbal devices, whereas G either used a nonverbal device or relied on context in initiating turns, when they could be coded (see Table 3.7). B relied on context in a large proportion of initiating turns in both periods two and three. Note that almost half of G's turns could not be coded for specific nonverbal devices, whereas B had only one completely uncodable turn. In period three, none of G's initiation turns could be coded, while all but one of B's 20 turns could (although two could not be coded for types beyond context). The majority of both twins' initiation turns utilized a verbal device in addition to being supported by context (see Table 3.8). B used a verbal device without a nonverbal device in 9 (or 22.0%) of initiations. In one case, B used a verbal device where the nonverbal device was uncodable (+/-nst). In 7 (23.3%) of G's initiations that included a verbal marker, nonverbal devices could not be coded; there is at least a possibility that these turns included nonverbal markers. 66 Table 3.5. Number (and %) of verbal device types for initiation turns # Verbal Period One Period Two Period Three Total Devices B G B G B G B G No device 1 2 5 2 1 — 7(17.1) 4(13.3) 1 device type — 6 6 5 6 2 12(29.3) 13(43.3) 2 device types 4 9 3 3 9 1 16(39.0) 13(43.3) 3 device types 1 — 1 — 4 — 6(14.6) --# $ini Turns 6 17 15 10 20 3 41 30 Table 3.6. Nonverbal devices used in initiation turns Nonverbal Devices Period One B G Period Two B G Period Three B G Total B G Context alone 1 13 9 1 14 — 24 14 Ctx+uncodable — — 1 3 1 — 2 3 Ctx+strategy® 2 4 4 4 3 — 9 8 Eye contact — 3 — 3 — — — 6 Gesture — 1 — 1 — — — 2 Physical action — — 2 — 4 -- 6 — Smile 2 1 1 1 — — 3 2 Touch — — 2 — -- — 2 — No device 3 — 1 — 4 8 — Uncodable — 10 — 2 2 3 2 15 # $ini Turns 6 17 15 10 20 3 41 30 *Turn supported by context and possibly by additional (uncodable) strategy/strategies @Turn supported by context and at least one of eye contact, gesture, physical action, smile, touch. This category double codes strategies listed singly. 67 Table 3.7. Number of nonverbal devices used for initiation turns Nonverbal Period One Period Two Period Three Total Device B G B G B G B G Uncodable — 2 — 2 1 3 1 7 No device 3 — 1 5 — 9 Device, no ctx — — 1 1 1 — 2 1 Context alone 1 5 8 1 10 — 19 6 Ctx + uncodable — 7 1 3 1 — 2 10 Ctx + 1 device 2 2 3 3 2 — 7 5 Ctx + 2 Devices — 1 1 — — 1 1 Total # $ini Turns 6 17 15 10 20 3 41 30 Table 3.8. Number (and %) of verbal devices used in conjunction with contextual support or nonverbal devices for initiations Device Period One Period Two Period Three Total Status B G B G B G B G -vst+ctx 1 2 4 1 1 6(14.6) 3 (10.0) +vst-nst 3 — 1 — 5 9(22.0) — +vst+ctx 2 13 9 6 12 23(56.1) 19(63.3) -vst+nst — — 1 1 — 1 (2.4) 1 (3.3) +vst+nst — — — — 1 1 (2.4) — +vst+/-nst — 2 — 2 1 3 1 (2.4) 7(23.3) Total$ini 6 17 15 10 20 3 41 30 Note: +/- indicates the nonverbal device was uncodable, and as such, there was a possibility that a nonverbal device did exist. 68 Table 3.9 Number (and %) of opening turns for each twin at each time period Code Period One Period Two Period Three Total B G B G B G B G $opn 19 31 23 10 10 11 52(10.9) 52(11.3) $opncnt — 1 — 1 — — — 2 (0.4) $rep 1 — 1 — - 1 1 3(0.6) 1 (0.2) $mod 1 5 6 3 2 1 9(1.9) 9 (2.0) $elb 10 4 10 3 1 3 21(4.4) 10 (2.2) $att 2 — 1 — — 1 3((0.6) 1 (0.2) Total 33 41 41 17 14 17 88(18.4) 75(16.3) Turns 112 231 221 127 144 103 477 461 Question 2b: Do the blind twin and the sighted twin differ in their topic change devices? Table 3.9 shows that B and G had a comparable proportion of opening turns; that is, 10.9% of B's 477 turns and 11.3% of G's 461 turns were coded as openers. Both twins used a variety of nonverbal markers of topic openers. Only B used the nonverbal strategy of physical action (see Table 3.10). B was also more likely to use touch (9.6%) than G (3.8%). G used eye contact (25.0%) and gesture (15.4%) more than any other single marker. Table 3.11 shows that B and G both used a wide variety of verbal devices to change topic. G was more likely to change topic without the use of a specific verbal marker, doing so in 28.8% of openers, while B neglected to use a verbal marker in opening turns only 17.3% of the time. Deictics, locating devices and routines were used to roughly the same degree by both twins. B used name of addressee (9.6%), pretend play (15.4%), directives 69 (26.9%), prosodic devices (32.7%), and questions (40.4%) to a much greater extent than G (1.9%, 3.8%, 5.8%, 15.3%, 25.0% respectively). In addition, B used repetition of his own utterance, a strategy that was absent from G's opening utterances. B used one or more verbal strategies supported by context in a greater proportion of opening turns than did G; compare 75.0% with 59.6% of 52 opening turns,-respectively (see Table 3.12). G had a greater proportion of opening turns in which context was used for support without an identifiable verbal strategy; compare 26.9% for G and 17.3% for B. Table 3.10. Nonverbal devices used in openers Nonverbal Period One Period Two Period Three Total Devices B G B G B G B G Context alone 13 10 18 3 3 4 34 17 Ctx + uncodable 1 10 5 1 3 5 9 16 Ctx + strategy 5 6 1 7 2 3 8 16 Eye contact — 2 — 8 — 3 — 13 Gesture — 5 — 3 — — — 8 Physical action 3 — 1 — 1 — 5 — Smile 3 — — 2 — 2 3 4 Touch 3 2 — — 2 — 5 2 No device 2 1 -- — 2 — 4 1 Uncodable — 7 — — 1 1 1 8 Total # openers: 19 31 23 10 10 11 52 52 70 Table 3.11. Verbal devices in openers Verbal devices Period One Period Two Period Three Total B G B G B G B G Directive 2 3 7 — 5 — 14 3 Deictic 1 5 6 1 — 1 7 7 Expressive — 5 — — 1 — 1 5 General addr. — 1 3 — — — 3 1 Label — 2 — — — — — 2 Locating — — 3 2 — 1 3 3 Addressee, nam 4 1 1 — — — 5 1 Pretend play 1 2 4 — 3 — 8 2 Prosodic 4 4 6 1 7 3 17 8 Question 13 10 6 1 2 2 21 13 Repetition, own — — 1 » — 1 — 2 — Routine 1 1 — -- — 1 1 2 Other devices — — 1 1 1 7 2 8 No device. 6 9 3 5 — 1 9 15 Total # openers 19 31 23 10 10 11 52 52 71 Table 3.12. Number (and %) of verbal devices used in conjunction with contextual support or nonverbal devices for openers Device Value Period One Period Two Period Three Total B G B G B G B G -vst -non — — — — — — -vst +ctx 4 7 5 6 1 9(17.3) 14(26.9) +vst -nst 1 1 -- 2 3 (5.8) 1 (1.9) +vst +/-nst — 2 — — — 2 (3.8) -vst +/-nst — 3 — — — 3 (5.8) +vst -l-ctx 13 17 18 4 8 10 39(75.0) 31(59.6) -vst +nst — — — — — — +vst +nst 1 1 — — 1 (1.9) 1 (1-9) # of Openers: 19 31 23 10 10 11 52 52 Question 2c: Do the blind twin and the sighted twin differ in their devices in attention-getting turns? Attention bids occurred infrequently in both twins' conversations. B used attention bids twice as frequently (4.0% of turns) as did G (2.0% of turns). All but one of the twins' attention bids was an initial attention bid. Both B and G utilized deictics, expressive particles, locating devices and prosodic devices as verbal utterance devices to gain attention (see Table 3.13). B also used the name of an addressee and pretend play as verbal devices in attention-getting turns, neither of which was used by G. G used repetition of his own utterance, a verbal device that was not used by B. 72 Table 3.13. Verbal devices used in attention bids Verbal device Period One Period Two Period Three Total B G B G B G B G Deictic — — — 1 1 — 1 1 Expressive 1 4 — — 2 1 3 5 Locating device — 1 4 2 1 1 5 4 Addressee name 7 — 1 — 2 — • 10 — Pretend play — — 1 — — — 1 — Prosodic device 4 2 2 — 2 2 8 4 Repetition, own — — — 2 — — — 2 # Attn. Bids 8 7 5 0 6 2 19 9 B neglected to use a nonverbal device in attention-getting turns in 31.6% of his 19 attention bids (see Table 3.14). For the turns that could be coded, G relied on context and the nonverbal devices of eye contact, physical action and gesture. B used the nonverbal devices of gesture, physical action, smiling, and touch. 73 Table 3.14. Nonverbal devices used in attention bids Devices Period One Period Two Period Three Total B G B G B G B G Context alone 3 3 1 — 2 — 6 4 Ctx+ uncodable — 1 -- — 1 — 1 2 Ctx+ strategy 1 — 2 — 1 2 4 2 Eye contact — — — — — 1 — 1 Gesture — — 1 — — — 1 1 Physical action — — 1 — 1 1 2 1 Smile 1 — — — — — 1 — Touch 3 — — — — — 3 — No device 2 — 2 — 1 — 6 — Uncodable — 3 1 — 1 — 2 3 # Attn. Bids 8 7 5 0 6 2 19 9 Question 2d: Do the blind twin and the sighted twin differ in their devices for establishing a turn? Turn bids were infrequent in both of the twins' conversation. Turn bids consisted 2.7% of B's 477 turns, 4.8% of G's 461 turns. All turn bids were initial turn bids. Both twins used a variety of verbal devices for establishing turns: agreement, cohesion, deictics, expressive particles and prosodic devices (see Table 3.15). Differences existed in the proportions of some of these devices, but the total numbers are small. B used deictics to establish a turn four times (30.8%of turn bids), while G used a deictic only once. G used 74 considerably more expressives, 13 (59.1% of turn bids), than did B, who used only one. B used name of addressee and repetition of self as verbal device for establishing a turn; G used neither of these. B used a wider selection of nonverbal devices for establishing a turn than did G (see Table 3.16). While G only relied on context, B used the nonverbal devices of gesture and physical action. It is important to remember that these counts are on a very small number of turns, and that a number of G's turns could not be coded because he was off-camera. , Table 3.15. Verbal devices used in turn bids Verbal Period One Period Two Period Three Total Devices B G B G B G B G Agreement 1 1 — — 1 1 2 Cohesion 1 3 — — — 3 1 Deictic — 2 — 2 1 4 1 Expressive 5 1 2 — 6 1 13 General addres . 1 — — — — — 1 Name addressee — 3 — — — 3 — Prosodic device — 4 1 — 2 4 3 Repetition, own — 1 — — — 1 — No verbal strat. 2 3 — — 1 — 3 3 Uncodable — — — — 1 — 1 # Turn Bids 2 10 8 2 3 10 13 22 75 Table 3.16. Nonverbal devices used in turn bids Nonverbal Period One Period Two Period Three Total Devices B G B G B G B G Context alone 2 3 5 — 2 6 9 9 Ctx+ uncodable 7 1 2 — 3 1 12 Ctx+ strategy — 2 — — — 2 — Gesture — 2 — — — 2 — Physical action — — — 1 — 1 — Uncodable — — — 1 1 1 1 # of Turn Bids 2 10 8 2 3 10 13 22 Table 3.17 shows the total verbal device usage for all discontinuous turns. B used directives (14.4%), an addressee's name (23.2%), pretend play (23.2%) and prosodic devices (36.8%) in a greater proportion of discontinuous turns than did G (7.1%, 8.0%, 2.7%, and 15.9%, respectively). G used a greater proportion of expressives (23.9%) than did B (6.4%). G alone used greetings and labels in discontinuous turns. G did not use a verbal device in 18.0% of his 113 discontinuous turns, compared with 14.4% of 125 discontinuous turns for B. The twins used deictics and questions in similar proportions. Table 3.18 displays the nonverbal device total for all discontinuous turns. B used physical action (11.2%) and touch (8.0%) in a larger proportion of discontinuous turns than did G (0% and 1.8%, respectively). B was also more likely to rely on context alone to support his turn than was G; compare 58.4% of 125 discontinuous turns for B and 38.1% of 113 for G. B did not use a nonverbal device in 13.6% of discontinuous turns, compared with 0.9% for G. G used gesture (8.8%) in a larger proportion of discontinuous turns than did B (2.4%). Note that a large proportion of G's discontinuous turns (23.9%) were uncodable, as compared to 4.8% of B's. Table 3.17. Number (and %) of verbal devices in all discontinuous turns Code Period One Period Two Period Three Total B G B G B G B G $coh — 1 3 — — — 3 (2.4) 1 (0.9) $agr — 1 1 — — 1 1 (0.8) 2 (1.8) $dir 2 5 7 — 9 3 18(14.4) 8 (7.1) $dei 1 9 13 3 4 2 18(14.4) 14(12.4) $exp 2 17 1 3 5 7 8 (6.4) 27(23.9) $gen — 3 4 1 — — 4 (3.2) 4 (3.5) $grt — 2 -- — — — — 2 (1,8) $lab — 2 — — — — — 2 (1.8) $loc — 1 7 5 2 2 9 (7.2) 8 (7.1) $nam 15 6 5 1 9 2 29(23.2) 9 (8.0) $pre 2 3 10 — 17 — 29(23.2) 3 (2.7) $pro 11 8 16 3 19 7 46(36.8) 18(15.9) $que 14 18 10 5. 4 2 28(22.4) 25(22.1) $row -- 1 3 2 2 — 5 (4.0) 3 (2.7) $rou 1 2 — — 1 1 2 (1.6) 3 (2.7) $oat — — 1 1 1 8 2 (1.6) 9 (8.0) $nov 9 14 8 5 1 1 18(14.4) 20(18.0) $unc — — — — — 1 — 1 (0.9) Turns 35 65 51 22 39 26 125 113 77 Table 3.18. Number (and %) of nonverbal devices in all discontinuous turns Nonverbal Period One Period Two Period Three Total Devices B G B G B G B G $eye — 5 — 11 — 4 — 20(17.7) $ges — 6 3 4 " — — 3 (2.4) 10 (8.8) $ast 3 — 4 — 7 — 14(11.2) — $smi 6 1 1 3 — 2 7 (5.6) 6 (5.3) $tou 6 2 2 — 2 — 10 (8.0) 2 (1.8) $ctx alone 19 29 33 4 21 10 73(58.4) 43(38.1) $ctx +$unc 1 18 7 6 5 8 13(10.4) 32(28.3) $non 7 1 3 — 7 — 17(13.6) 1 (0.9) $unc — 20 1 2 5 5 6 (4.8) 27(23.9) Disc. Turns 35 65 51 22 39 26 125 113 Question 3a: Do the blind twin and the sighted twin differ in their success in obtaining a response to conversational initiations? Success or nonsuccess after one turn were the data used to answer this question, i.e., only turns coded $ini, whether or not they were later elaborated, repeated, or repeated with modification. G was more successful than B in gaining a response to his conversational initiations; compare 14 (46.7%) and 15 (36.6%), respectively. Note, in addition to contingent responses, noncontingent and negative responses, responses by someone not addressed, and nonverbal attends are included as indicative of success. Initial initiations that were later elaborated, repeated, followed by attention bids, or any combination of these strategies are indicated in Table 3.19 with an asterisk. 78 Table 3.19. Responses to initial initiation turns Response Period One Period Two Period Three Total Type B G B G B G B G Contingent 2 6 5 3 2 1 9(22.0) 10(33.3) Attention 1 — — 1 1 — 2( 4.9) 1 (3.3) Noncontingent — *1 1 — — — 1( 2.4) *1 (3.3) Negative *1 — — — — — *1(2.4) — Non-addressee — 1 — — 2 1 2 (4.9) 2(6.7) No response 2 4 5 2 6 — 13(31.7) 6(20.0) — *5 . *4 *4 *9 *1 13(31.7) 10(33.3) Total (no response) 2 9 9 6 15 1 26(63.4) 16(53.3) Total (response) 4 8 6 4 5 2 15(36.6) 14(46.7) Total $ini: 6 17 15 10 20 3 41 30 * indicates an initial turn that is later modified Subsequent tries following initiation bids occurred not only after initial bids failed to get any response, but also following negative and noncontingent responses. Success ratings following such subsequent attempts to initiate are found in Table 3.20. Whereas G was successful in all 11 re-inititions, B failed to gain a response after all 4 re-initiation attempts in Period Two and more than half, 55.6%, of re-initiation attempts in Period Three. Table 3.21 illustrates the number of initiations responded to, regardless of whether on a first or a subsequent try. B was successful in gaining a response (i.e., contingent, noncontingent, negative, or a response from a person not addressed) or a nonverbal attend in 46.3% of his 41 initiation bids, which represents, proportionally, a litle more than half the success rate of G (80.0%). 79 Table 3.20. Responses to re-initiation turns ($rep, $mod, $elb, $cnt, $att) Response Period One Period Two Period Three Total Type B G B G B G B G Contingent — 5 — 1 • 3 1 3(21.4) 7 (63.6) Attention — — — 1 — — — 1 (9.1) Noncontingent — 1 — 1 1 — 1(7.1) 2 (18.2) Negative — — — 1 ~ — — 1 (9.1) Non-addressee 1 — — — — — 1(7.1) — No response — — 4 — 5 — 9(64.3) — Total(success) 1 6 — 4 4 1 5(35.7) 11(100.0) # re-initiations: 1 6 4 4 9 1 14 11 Table 3.21. Number (and %) of initiations responded to on a first or subsequent try Response Period One Period Two Period Three Total Type B G B G B G B G Contingent 2 11 5 4 5 2 12 (29.3) 17 (56.7) Attention 1 — — 2 1 — 2 (4.9) 2 (6.7) Nonconting. — 1 1 1 1 — 2 (4.9) 2 (6.7) Negative — — — 1 — — — 1 (3.3) Nonaddressee 1 1 — — . 2 3 3 (7.3) 2 (6.7) No Response 2 4 9 2 11 — 22 (53.7) 6 (20.0) Total (success) 4 13 6 9 9 3 19 (46.3) 24(80.0) #ini 6 17 15 10 20 3 41 30 Note: Each counted unit represents one occasion on which ini was attempted, regardless of whether it took place over one or several contiguous turns. Response type was coded for each final turn in such a sequence. 80 Question 3b: Do the blind twin and the sighted twin differ in their success in obtaining a response to conversational openers? Success of an opener after one turn is shown in Table 3.22. Thus, only turns coded $opn, whether or not they were later repeated, elaborated or followed by a bid for attention, are included. Proportions for the success categories ranged from 42.1% to 78.3% for B and 45.5% to 60.0% forG. Table 3.22. Responses to initial opening bids Response Period One Period Two Period Three Total Types B G B G B G B G Contingent 4 9 15 4 2 4 21(40.0) 17(32.7) Attention *2 2 — 1 — 1 *2 (3.8) 4 (7.7) Noncontingent — 4 2 3 — 5 (9.6) 4(7.7) *2 *1 *1 *1 (1.9) *3(5.8) Negative *1 -- *1 — — — *2 (3.8) — Nonaddressee 1 ~ — — — — 1 (1-9) — No response 4 8 1 1 2 1 7(13.5) 10(19.2) *6 *6 *4 *3 *2 *4 *12(23.1) *13(25.0) Total 10 14 5 4 4 5 19(36.5) 23(44.2) (no response) Total(response) 8 17 18 6 6 5 32(61.5) 28(53.8) Uncodable 1 — — — — 1 1 (1.9) 1(1.9) Total $opn 19 31 23 10, 10 11 52 52 indicates an initial opener that was later modified 81 Note that initial openers that were re-attempted after failure to gain a response, a negative response, a noncontingent response, or a bid for attention are indicated in Table 3.22 with an asterisk. Success ratings for these subsequent opening attempts are presented in Table 3.23. In Period One the majority of both B and G's modified openers received a response. In Period Two none of B's five modified openers gained a response, but three out of four of G's openers gained a response (see Table 3.23). In Period Three, B gained a response to two of his three openers. G received a response to two of his openers and failed to get a response to the other two. Table 3.24 illustrates the number of openers responded to, regardless of whether on the first or a subsequent try. Although which twin had the greater proportion of openers responded to varies according to the time period, the total proportions are similar for B and G; 73.1% of B's 52 openers and 69.2 % of G's 52 openers were successful. Note that success includes the response types of contingent, noncontingent, non addressee, negative, and nonverbal attention. Table 3.25 looks at the number of discontinuous turns, most globally, (i.e., $ini and $opn) that gained a response or nonverbal attend. When the success rates of initiations and openers were added together, G was more successful than B in gaining a response or nonverbal attend. Compare 73.2% of G's 82 discontinuous turns and 60.6% of B's 94 discontinuous turns. 82 Table 3.23. Responses to openers after modification Response Period One Period Two Period Three Total Type B G B G B G B G Uncodable — 1 — — — — — 1.(6.3) No Response 1 1 5 1 1 2 7(41.2) 4(25.0) Contingent 6 5 — 3 1 2 7(41.2) 10(62.5) Noncontingent 2 — — — 1 — 3(17.6) — Nonaddressee — 1 — — — — — 1 (6.3) Total (success) 8 6 — 3 . 2 2 10(58.8) 11(68.8) # "modified" 9 8 5 4 3 4 17 16 Total # of openers 19 31 23 10 10 11 52 52 Table 3.24. Number (and %) of openers responded to on a first or subsequent try Response Period One Period Two Period Three Total Type B G B G B G B G Contingent 10 14 15 7 3 6 28(53.8) 27(51.9) Attention — 2 — 1 — 1 — 4 (7.7) Noncontingent 3 4 2 — 4 — 9(17.3) 4 (7.7) Negative — — — — — — — — Nonaddressee 1 1 — — — — 1 (1-9) 1 (1-9) No Response 5 9 6 2 3 14 14(26.9) 14(26.9) Uncodable 1 1 — — — 1 1 (1.9) 2 (3.8) Total 14 21 17 8 7 7 38(73.1) 36(69.2) (success) # $opn 20 31 23 .10 10 53 52 52 Note: Each counted unit represent one occasion on which an opening bid was attempted, regardless of whether it took place over one or several contiguous turns. Response type was coded for each final turn in such a sequence. „ 83 Table 3.25. Number (and %) of discontinuous turns responded to Response Period One Period Two" Period Three Total Type B G B G B G B G Contingent 12 25 20 11 8 8 40(42.6) 44(53.7) Attention 1 2 — 3 1 1 2 (2.1) 6 (7.3) Noncontingent 3 5 3 1 5 — 11(11.7) 6 (7.3) Negative — — — 1 — — — 1 (1-2) Nonaddressee 2 2 — — 2 1 4 (4.3) 3 (3.7) No Response 7 13 15 4 14 3 36(38.3) 20(24.4) Uncodable 1 1 — — — 1 1 ( L D 2 (2.4) Total (success) 18 34 23 16 16 10 57(60.0) 60(73.1) Response Types 26 48 38 20 30 14 94 82 Question 4a: Do the blind twin and the sighted twin differ in the degree to which they persist when an initiation fails? G was more likely than B to persist when an initiation failed. Whereas G persisted for 10 of his 16 (62.5%) failed initiations, B persisted in 13 of his 26 (50.0%) failed initiations (see Table 3.26). When B did persist, he necessarily did so over a greater number of turns than did G. Thus, B persisted for as many as seven turns in Period Three, whereas G never needed more than three subsequent attempts in order to gain a response. The following example demonstrates B's persistence. B (5;3) is playing near N and A. B has made several prior attempts to join their interaction. Notice that B continues has bid after getting a negative response from A. 84 B: B B B B A B B B B I'll [/] I'll park it < into the garage > [ > ]. $ini Situation: B puts his truck on the track that N and A are building. A and N do not notice B as he has placed his truck in the middle of the track and each boy is busily building at opposite ends of the track. I'm in the garage. $mod Well that's where I park my car # in the garage. $mod You know where I did park my car? $mod eh <my> [>] +/. $unc <Bryan> [<] go in your own room. Get my motorbike. $elb Mine is gonna park outside. $mod <I'm> [//] I parked it outside. $mod I parked it outside. $mod (B receives no response from either N or A.) Both twins received one noncontingent response after an initial initiation bid. B made no subsequent attempt to gain another response, whereas G did attempt to gain a contingent response. On receipt of his one negative response to an inital initiation bid, B persisted in his attempt to gain a more desirable response. Question 4b: Do the blind twin and the sighted twin differ in the degree to which they persist when an opener fails? There was a slight difference in the degree to which B and G persisted when an opener failed. B persisted in 12 of the 19 (63.2%) opening turns that received no response (see Table 3.27). G persisted in 13 of the 23 (56.5%) opening turns that received no response. Similar to data on initiations, B and G differed regarding how many turns each needed to persist. B persisted for up to nine turns in Period Two. In contrast, G persisted for up to three turns in Period Two, but usually persisted for one or two turns. Both twins were less persistent if they gained a noncontingent response; if they did persist, they did so for only one turn. 85 Most globally, adding together the data for initiations and openers, G persisted to a slightly greater degree than did B. Thus, G persisted for 23 out of 39 (59.0%) failed discontinuous turns, whereas B persisted in 25 out of 45 (55.6%) failed discontinuous turns. Table 3.26. Number of $mod, $elb, $rep, $att turns following $ini turns that (a) received no response # Turns Period One Period Two Period Three Total Persisted B G B G B G B G #ini+nor 2 9 9 6 15 1 26 16 #nor + 0 turns 2 4 5 2 6 — 13 6 #nor+ 1 turn ~ 3 1 4 3 — 5 8 #nor+ 2 turns -- 1 1 — 2 — 3 1 #nor+ 3 turns — 1 — 1 1 1 2 #nor+ 4 turns — — 1 — 2 — 3 --#nor+ 6 turns — — — 1 — 1 — #nor+ 7 turns — — 1 — — — 1 — Total persisted: — 5 4 4 9 1 13 10 %: — (55.6) (44.4) (66.7) (60.0) (100) (50.0) (62.5) (b) received a noncontingent response # Turns Period One Period Two Period Three Total Persisted B G B G B G B G #ini+ ncr 1 1 — 1 1 #nor+ 0 turns 1 — 1 — #nor+ 1 turn 1 -- — — 1 86 Table 3.27. Number of $mod, $elb, $rep and $att turns following $opn turns that a) recieved no response # Turns Period One Period Two Period Three Total Persisted B G B G B G B G #opn+ nor 10 14 5 4 4 5 19 23 #opn+ 0 turns 4 8 1 1 2 1 7 10 #opn+ 1 turn 5 5 1 1 1 2 7 8 #opn+ 2 turns 1 1 1 1 2 2 4 #opn+3 turns — — 1 — — — 1 #opn+ 5 turns -- 1 — — — 1 — #opn+ 6 turns 1 — — — — — 1 — #opn+ 9 turns — 1 — — — 1 — Total persisted 6 6 4 3 2 4 12 13 (60.0) (42.8) (80.0) (75.0) (50.0) (80.0) (63.2) (56.5) b) received a noncontingent response # Turns Period One Period Two Period Three Total Persisted B G B G B G B G #opn+ ncr 6 2 1 4 — 6 7 #opn+ 0 turns 4 2 — 3 — 5 4 #opn+ 1 turn 2 ~ 1 1 — 1 3 87 Table 3.28. Strategies used in an initial initiation after no response, a noncontingent or negative response or nonverbal attention Strategies Period One Period Two Period Three Total B G B G B G B G Exact repetition — 1 — — — 1 — 2(12.5) Modified rep. — 4 8 2 7 2 15(37.5) 8(50.0) Elaboration 2 4 6 2 12 — 20(50.0) 6(37.5) Attention bid — — — — 5 — 5(12.5) — # ini followed up 1 6 4 4 9 1 14 11 # Reinitiations 2 9 14 4 24 3 40 16 Question 5a: Do the blind & the sighted twin differ in the strategies they, use when initiation fails? Table 3.28 shows that both twins utilized modified repetition and elaboration as re-initiation strategies. However, B used elaboration to a greater extent than G; compare 50.0% of 40 reinitiation turns and 37.5% of 16 reinitiation turns, respectively. G was more likely to use repetition ($rep, $mod) than B; compare 62.5% and 37.5%, respectively. Only G utilized exact repetition, and only B utilized attention bids. 5b: Do the blind and the sighted twin differ in the strategies they use when an opener fails? Both B and G utilized all coded types of strategies in the instance of a failed opener (see Table 3.29). There were some differences regarding their proportions of usage of strategies within the different time periods; however, the number of turns is so small that it is meaningless to compare types within each period. But, overall, B used elaboration most frequently, 21 of 36 modified openers (58.3%). G, however, used elaboration and repetition ($rep and $mod) to an equal extent. 88 Table 3.30 shows total strategy use for subsequent initiations and subsequent openers. Again, B used elaboration most frequently and G used elaboration and repetition to an approximately equal extent. Table 3.29. Number (and %) of strategies used in a subsequent attempt when an opener received no response, a noncontingent or negative response or nonverbal attention Strategies Period One Period Two Period Three Total B G B G B G B G Exact repetition 1 — 1 — 1 1 3 (8.3) 1 (4.8) Modified repetition 1 5 6 3 2 1 9(25.0) 9(42.9) Elaboration 10 4 10 3 1 3 21(58.3) 10(47.6) Attention bid 2 — 1 — -- 1 3 (8.3) 1 (4.8) # "modified" 9 8 5 4 3 4 17 16 openers Total # strategies 14 9 18 6 4 6 36 21 Table 3.30. Number (and %) of strategies used in subsequent initiations and openers after receiving no response, a noncontingent or negative response, or nonverbal attention Strategies Subsequent Subsequent Total Initiations Openers B G B G B G Exact Repetition — 2 3 1 3 (3.9) 3 (8.1) Modified repetition 15 8 9 9 24(31.6) 17(45.9) Elaboration 20 6 21 10 41(53.9) 16(43.2) Attention 5 — 3 1 8(10.5) 1 (2.7) # Re-attempted bids 40 16 36 21 76 37 89 Question 6: Do the blind twin and the sighted twin differ in the extent to which they respond to previous utterances when a response is obliged? The percentage of responses actually given in obligatory contexts (Table 3.31) ranged from 72.2% to 100% for B and 81.0% to 100% for G. The lowest percentage of required responses given occurred in Period One for both of the twins. Table 3.31. Number (and %) of responses in obligatory contexts actually given Code Period One Period Two Period Three Total B G B G B G B G Given $req 13 17 23 28 16 8 52(89.7) 53(93.0) Total $req 18 21 24 28 16 8 58 57 Question 7: Do the blind twin and the sighted twin differ in the extent to which they respond to previous utterances when a response is not obliged? The percentage of turns that were coded as nonobliged responses (Table 3.32) ranged from 24.1% to 27.7% for B and 33.3% to 40.7% for G. In general, the percentage of responses increased over time for both twins, with G always providing a greater percentage of responses than B. 90 Table 3.32. Number (and %) of nonobliged responses Period One Period Two Period Three Total B G B G B G B G # 27 77 56 47 40 42 123 166 % (24.1) (33.3) (25.3) (37.0) (27.8) (40.8) (25.8) (36.0) Summary of Results B and G differed with regard to bids for interaction. Although the number of interactions each twin attempted to initiate was similar, B devoted proportionally almost twice as many turns to initiations than did G. With regard to the use of verbal devices in discontinuous turns, the twins were more similar than different, with the exception of a few devices. B and G used agreement, cohesion, deictics, greetings, general addresses, locating devices, questions, repetitions and routines in similar proportions. The twins differed in their use of expressives, pretend play, prosodic devices and vocatives. B used a greater proportion of pretend play, prosodic devices and vocatives than did G; whereas G used a greater proportion of expressives than did B. G alone used greetings and labels in discontinuous turns. B and G were more different in their use of nonverbal devices than verbal devices. The only nonverbal device that B and G used in similar proportions was smiling. B alone used nonverbal action strategies, such as jumping up and down, and he used contextual support alone a greater proportion of the time than did G. As would be expected, G alone used eye contact. 91 G was more successful than B in getting a response to his discontinuous turns. The twins' relative success varied according to discourse category, with G markedly more successful than B with initiations. B was slightly more successful than G with openers. The twins were similar in their persistence after a failed initial initiation bid or opener, but their favored strategies differed. More than half of B's reattempts involved elaboration, whereas more than half of G's involved repetition. 92 CHAPTER IV DISCUSSION In Chapter I, the following seven questions regarding the conversational discourse of B and G.were asked: 1. Do the blind twin and the sighted twin differ in the extent to which they initiate conversational interaction? 2. Do the blind twin and the sighted twin differ in their verbal and nonverbal devices: (a) for initiating interaction? (b) for opening bids? (c) to gain attention? (d) to establish a turn? 3. Do the blind twin and the sighted twin differ in their success in obtaining a response to conversational initiations and openers? 4. Do the blind twin and the sighted twin differ in the degree to which they persist when an initiation fails? 5. Do the blind twin and the sighted twin differ in the strategies they use when an initiation or opener fails? 6. Do the blind twin and the sighted twin differ in the extent to which they respond to previous utterances when a response is obliged? 93 7. Do the blind twin and the sighted twin differ in the extent to which they respond to previous utterances when a response is not obliged? To answer these questions, transcripts sampling B and G's conversational discourse in a kindergarten classroom between the ages of 4; 11 and 5;6 were coded for interactional status, verbal and nonverbal utterance characteristics, and type of response received. Data Interpretation Considerations Certain contextual and situational factors may have influenced the children's individual profiles of conversational discourse. These will be mentioned briefly before an interpretation of the results. One possible confound is the inequality of turns within and across time periods. Although the number of the twins' turns differed within time periods, overall total turns were approximately equal (compare 477 turns for B and 461 turns for G). Differences in the number of turns for each twin in each time period, of course, impacted on the comparability of frequency counts of discourse codes. For this reason, comparisons were based on proportions. However, the small number of turns for such discourse codes as attention bids and turn bids impacted on the interpretations that could be drawn. Due to the nature of proportions, values are generated from a closed set. For every high value there must be a low value. A high value may indeed reflect frequent usage of that specific conversational unit, or a high value may be an epiphenomenon of the infrequency of another unit. As such, it is noted that the discussion in this chapter represents one interpretation of the meaning of these values. 94 Interpretation is also constrained by the nature of the study, which is both single-subject in design and preliminary. More research must be done before any definite statements can be made regarding the conversational competencies of blind children in kindergarten. In addition, B has been singled out as a particularly well-functioning blind child, who has no other disabilities. The figures quoted by Jan, Freeman and Scott (1977) suggest that this is atypical of blind children. They report that 50 - 90% of all blind children in North America have additional - often multiple - disabilities. Another issue is the variability of context. Different situations create differences in opportunities for specific kinds of turns. Furthermore, there was no control over the number or age of other participants in the conversations. These contextual issues will be discussed further below. Initiation of Conversational Interaction To answer the question, "Do the blind twin and the sighted twin differ in the extent to which they initiate conversational interaction?", both the number of interactions initiated and the proportion of turns devoted to initiation were analyzed for each twin. B attempted to initiate more interactions than G (41 and 30, respectively), and also devoted a greater proportion of turns to initiation than his twin did (17.2% and 10.2%, respectively), largely due to period three, where B devoted 31.3% of his 144 turns to conversational initiation, in contrast to G's 5.8% of 103. 95 B's greater proportion of turns devoted to initiation was, of course, partly due to the fact that he initiated more interactions. B may have initiated more interactions because he, unlike his twin, does not have access to what is going on around him. As such, B can not easily fall into multiparty conversations or follow topic switches and tangential remarks that result from events in the surrounding environment. Example 1, below, demonstrates the difficulties B has maintaining conversation when rapid topic shifts occur due to extraneous events in the classroom and his difficulty reinitiating when the attention of his conversational partner has strayed. The devices B uses, such as greetings in the middle of an established interaction (also reported by Andersen, 1984) and his extensive use of pretend play as a way to control the interaction, are not effective in this case. Example 1: (B's [4; 11] interaction with N has been interrupted by C who teases B. B talks briefly with C and then attempts to re-initiate interaction with N, who is still preoccupied with the activity at the bulletin board unbeknown to B). B: pretend I was a baby N (name) N no. N : no way. B [baby sounds]. B xxx [pretend play voice]. B hi. C xxx [mimics B]. B I'm in the house. (An art object falls off the bulletin board; C points to the fallen object and laughs; another child tries to put it back up). B: ha [= laughs]. B: wha...? B: <my> [>]. C: <haha> [<] [= laughs]. B: < hey # where' s > [ > ] my bottle? C: <haha> [<]. B: was that Billy talking? 96 Other likely contributing factors included differences in the situations the children were involved in. In period three, B entered a teddy bear hospital play area after most of his peers had already established pretend play interactions. B attempted to establish several of his own pretend play scenarios, but many of these scenarios were unsuccessful in gaining peer response. Example 2: (B [5;6] is playing doctor with his teddy bear in the corner of the room designated as the teddy bear hospital. B hears a classmate approach the cot where he is sitting.) B: give the [//] my bear something. B: say I was the doctor. (The classmate has already looked at B's bear and left). A hospital pretend play scenario involves rigid scripts in which specific instruments must be used for specific purposes and instrument placement is also important. Such precision is often difficult to obtain without instruction. In the following example, the teacher (T) gives B (5;6) instructions about how to find the teddy bear's mouth so that he can give the bear medicine. Example 3: I'm giving my bear some medicine. good # find his mouth. okay where's his mouth? I'll give him his mouth. there's his mouth. that's his tail. Peers will often not tolerate alternatives in scripts, such as medicine being administered to the bottom as opposed to the top end of the teddy bear. In addition, peers are often not conversationally mature enough or patient enough to guide B in such play interactions. 97 In contrast, in period three, G was involved with several children who were sitting on the floor drawing pictures on a large piece of cardboard. G's situation represented a ready-made conversational interaction. G maintained conversational interaction by making comments on what he and his neighbours were drawing. Consideration must also be given to B's success in gaining a response to his initiations. In period three, B made 20 attempts at initiation, and failed to gain a response in 15 of these initiations. B re-attempted initiation in nine of these 15 failed attempts. Of these nine re-attempts only four were successful, two after one re-initiation turn, the others after two and three re-initiation turns. B also tended to persist with re-initiation turns despite continually receiving no response. Thus, for example, in one instance in period three B made six re-initiation turns, none of which gained a response, and in two instances made four re-initiation turns, all resulting in no response. This is in keeping with reports that blind children have difficulty initiating conversational interaction (Dunlea, 1989; Kekelis & Andersen, 1984; Mulford, 1983; Rowland, 1983). A closer look at the means B used in his initiations attempts may shed some light on why this is so. Verbal and Nonverbal Characteristics of Conversational Beginnings Differences in B and G's use of verbal and nonverbal characteristics will be considered first in initiation bids, then in opening bids. Next, interesting observations regarding B's use of attention bids will be discussed. Finally, the differences between B and G's use of verbal and nonverbal features in all discontinuous turns will be considered in light of research reported in Chapter I. 98 There were interesting differences in the proportions of certain verbal characteristics of B and G's initiating turns. B used prosodic devices such as raised pitch and increased loudness, and pretend play to a greater extent than G. In fact, B used pretend play in half of his 41 initiation bids, whereas G did in only one of his 30 bids. B's extensive use of pretend play could be interpreted as a compensatory strategy. That is, it may be that B created pretend play scenarios to draw his peers into his play interactions that he controlled, making it unnecessary for him to access ongoing interactions, an often difficult feat when one is not able to see what is going on in the surrounding environment. In example 4 below, B (5;6) stands alone in the teddy bear hospital and offers that his bear is awake, perhaps indirectly suggesting that the bear is ready for any medical intervention a stray doctor may be willing to give: Example 4: B: bear # morning time. B: wake up. B: my bear's going to wake up now. B: see he woke up. B did not use greetings in initial initiations bids, but rather in the middle of established interactions, consistent with reports by Andersen (1984) (see example 6). With regard to nonverbal characteristics in initiations, B relied on context alone (i.e. the proximity of potential conversational partners) to a higher degree than G, not using nonverbal devices such as gesture. This compares with Mulford's (1983) observation of three blind children about the same age as B and G. G used eye contact in 20.0% of initiations, an option not available to B. This may seem low, but there is no quantitative data available for comparison with other sighted children. 99 B and G also differed in their usage of verbal and nonverbal characteristics in openers. B used the addressee's name, pretend play, prosodic devices and questions to a greater extent than G. In fact, B was more likely than his twin to use a specific verbal characteristic when introducing a new topic. B alone used the nonverbal characteristics of physical action and touch. Predictably, G used both eye contact and gesture in proportions similar to those for initiations. B and G used smiling to an equal degree for both initiations and opening bids. This is in conflict with reports by Parke, Shallcross and Andersen (1980) who report that blind students spend more time smiling than sighted students. This discrepancy may be due to Parke et al.'s older and larger sample of subjects. Not much can be said regarding the difference between B and G's use of verbal and nonverbal characteristics for gaining attention and turns, due to the small and discrepant number of turns in these categories. However, it is of interest that, for B, verbal attention-getting devices often served a "checking" function. That is, B used verbal devices such as vocatives and expressive to gain information regarding the identification and/or proximity of individuals in the surrounding environment. Such checking was necessary because of B's nonverbal behaviour, such as looking away from a conversational partner while talking to her, and because of long gaps in conversation, both of which often resulted in the loss of attention of the conversational partner. B may have used more attention bids than G due to having to check for partner's attention verbally during conversation. Due to the small number of attention-getting and turn bids, it is perhaps more valuable to look at the twins' use of verbal and nonverbal characteristics across all discontinuous turns. B does not have access to many nonverbal attention-getting devices, such as gaze direction and 100 eye-contact (used by G in 17.7% of his 113 discontinuous turns), smiling used in conjunction with eye gaze, or orientation towards a listener. B did make use of the nonverbal devices he has at his disposal, such as touch (used in 8.0% of his 125 discontinuous turns, as compared to 1.8% of G's 113 discontinuous turns) and proximity (used in 69.0% of discontinuous turns). However, such devices provide information regarding only the physical presence of a listener; attention is not guaranteed (Mulford, 1983), as example 5 illustrates. Example 5: (B [4; 11] and N are sitting together on a cot). B: the doggie's not comin' here today [laughs]. B: where's the doggie gone? B: I don't know where he is. B: I'll [/] I'll [/] I'll call him. (B's head is turned away from N and he is pressing his eyes. N's attention turns towards activity going on at the bulletin board). B: hey # come here. B: xxx pup puppy dog [high pitch]. B: eh the puppy dog's not comin' today. B: where's the puppy dog? B: N (name), xxx puppy dogs (long pause). B: then xxx N (name), xxx all the dogs here B: ah (pause). (N is talking to another child). B: N (name). N: what? Whereas G can use eye contact and head orientation to maintain and/or check attention status, B must necessarily rely on verbal means to check the attention status of his conversational partner. Thus, in general, B relies heavily on verbal means such as prosodic devices, pretend play, vocatives and directives. The use of prosodic devices may relate to B's dramatic and social nature. Recall the discussion of B's use of pretend play in initiations. 101 B's comparatively greater use of vocatives across all time periods and in total is highly adaptive. Vocatives function not only to gain general attention, but also to specify "who should attend," which B can not specify with eye contact. In example 6, B used an expressive, a verbal device which, without the combined use of eye contact, does not designate who is to attend. Example 6: (B [4; 11] is talking as he walks on his way to the playhouse corner of the classroom; suddenly he hears a group of children playing nearby and moves towards them.) (Note that the symbols < and > in the transcript indicate overlapped speech.) B: hey. R: you can back <up towards my > [>] voice. C: < what B (name) > [ < ]? B: I'm missing the house. B: C (name)? C: what? T: what B (name)? (T approaches in response to B's ambiguous expressive "hey.") Thus, vocatives are a highly functional verbal device for a blind child. Vocatives also function to check the status of a listener's attention and maintain attention (Keenan & Schieffelin, 1976). B frequently used vocatives to serve this "checking" function, as was demonstrated in example 5. B also used twice as many directives, proportionally, than G. This is consistent with reports by Dunlea (1989) that blind children use more directives than sighted children, and that such expanded usage represents an adaptive strategy used by blind children to increase participation opportunities. 102 No difference between the twins in the use of questions surfaced in the total counts for all discontinuous turns; however, quantitative differences in B and G's questions were reported for both initiations and opening bids. B's use of questions was high for opening bids but comparatively low for initiations, and vice versa for G, the differences cancelling each other out. This levelling out effect for the total discontinuous turns emphasizes the importance of comparing the twins' usage of verbal and nonverbal characteristics across all discontinuous turns as well as within discontinuous coding categories, especially when numbers are small, since it seems unlikely that either twin would reserve questions to mark one particular type of discontinuous turn. Success in Gaining a Response G was almost twice as successful than B in gaining responses to his conversational initiations, with B succeeding in fewer than half of his initiations. The ability to gain and maintain attention is essential when trying to begin a conversational interaction (Keenan & Schieffelin, 1976; McTear, 1986). As previous examples demonstrated, B had difficulty maintaining the attention of his conversational partner. B was successful in approximately half (52.6%) of his 19 attention bids. Although G was successful in none of his nine attention bids, he seemed to achieve attention more often with nonverbal means such as head orientation and eye contact, nonverbal characteristics that were difficult to code adequately due to the camera not always being focused on the child's face. 103 B's difficulty maintaining his listener's attention and the fact that he does not have access to what is going on around him may account for some of his difficulty in gaining a response to his initiation bids. B's large proportion of turns devoted to initiation and his lesser success than G in gaining a response to his initiation bids is consistent with other reports that blind children have difficulty initiating interaction (Dunlea, 1989; Kekelis & Andersen, 1984; Mulford, 1983; Rowland, 1983). B was more successful than G for initial opening bid attempts. Out of 52 openers, 61.5% were successful for B and 53.8% were successful for G. B may have been more successful in initial openers because he used a verbal device in addition to a supportive context in 75.0% of his 52 openers, as opposed to G, who used context and a verbal device(s) in only 59.7% of his 52 openers. In addition, G was more likely than B to change topic without the use of a specific verbal device. For subsequent opening attempts, G may have been slightly more successful (21.2%) than B (19.2%), but this difference is too small to be interpreted. Again, as with initiation bids, B has difficulty gaining a response to his subsequent attempts. The ability to take the necessary steps to reinitiate if the original initiation bid or opening bid is unsuccessful in gaining a desirable response is important for successful initiation of conversation and topic introduction (McTear, 1985). 104 Persistence G was more likely than B to make a subsequent attempt at initiation when an initial initiation failed. In contrast, with regard to openers, B was slightly more persistent than G if an initial opening bid failed. Why was this the case? The answer may have to do with the fact that one is more likely to re-attempt a bid if he has experienced previous success. Thus, each twin is more willing to make a subsequent attempt with the conversational bid with which he experienced the greatest success in gaining a response. Relatedly, consider the operational definitions of initiations and openers. Initiations were defined as a bid for interaction when there has been no prior verbal interaction. In contrast, openers occurred only in established interactions. As such, when making a topic bid, B knew that there always existed an addressee in the proximity, in contrast to initiations, for which an addressee may or may not be in the proximity. Consequently, it makes sense that B would persist more for openers, a type of discontinuous discourse turn in which a potential addressee was guaranteed, and, therefore, there existed the possibility of gaining a response to the bid. The number of turns each twin will persist also seems to be related to success. It usually took G only three turns to gain a response, and on the rare occasion when he did not succeed in three turns, he gave up. B, on the other hand, usually required several turns to gain a response. 105 Strategies used when an Initiation or an Opener Failed While B favoured elaboration as a strategy for subsequent attempts at discontinuous turns (i.e., initiations and openers), G was equally likely to use either elaboration or repetition. B's use of elaboration as a strategy for re-attempted discontinuous turns, in which he tended to use the verbal device of pretend play, may have been related to the phenomenon of verbal role play reported by Andersen et al. (1984) and Urwin (1978). In both B's instances of elaboration using pretend play and verbal role play, he took responsibility for both sides of the conversation. B's poor success at gaining a response to his re-attempted bids (whereas B was successful in 13 of 68, or 19.1%, of his re-attempted bids, G was successful in 23 of his 36 re-attempted bids) may be partly related to his frequent use of verbal role-play. B's lesser success in gaining a response to re-attempted bids may also be linked to his frequent reliance on contextual support alone in the nonverbal domain; compare 42 (or 61.8%) of 68 subsequent attempts for B and 16 (or 44.4%) of subsequent attempts for G. Note that in six of these subsequent attempts for G, nonverbal devices could not be coded; thus the possibility remains that a nonverbal device was used. The only specific nonverbal device used by B for re-initiations was physical action, such as jumping up and down. This is consistent with Mulford's (1983) report that some blind children develop alternative initiation devices, due to their limited repertoire of nonverbal devices. In re-attempts at opening bids, B used smile alone once and smile in conjunction with touch once. In contrast, G had the benefit of using a wider repertoire of nonverbal devices, including smiling, eye contact and gesture. The latter two devices were present in the majority of G's successful re-attempted bids. B's lesser use of repetition (35.5% of 76 re-attempted bids) than G (54.0% of 37 re-106 attempted bids) is not consistent with Prizant's (1984) conclusion that blind children use repetition more frequently than sighted children. This discrepancy may be due to the age differences between the blind children in previous studies and B. The studies that were reviewed looked at younger children. Even Prizant (1984), who suggests that repetition is often used as a language processing strategy by blind children, studied younger blind children (1;11 to 2;8). In contrast, B was observed at three time periods over the ages of 4; 11 to 5;6. Thus, in general, G's success with re-attempted bids seemed to be related to his use of a specific verbal marker and (a) nonverbal cue(s) (i.e. eye contact, gesture, action strategy) in addition to contextual support; however, it is difficult to make definite statements due to the high number of uncodable nonverbal strategies for G. In contrast, B seems to rely on general verbal strategies, such as pretend play and prosodic devices, as well as on context alone for support. Responses G was slightly more likely to respond to a previous utterance when a response was obliged in both Periods One and Two. In Period Three both twins responded on all occasions when a response was obliged. For both twins, the proportion of responses in obligatory contexts increased with time, consistent with reports of children's increased ability to give contingent and explicit responses to questions over time (Bloom, Rocissano & Hood, 1976; Ervin-Tripp, 1978). The percentage of responses when a response was not obligatory increased over time 107 for both twins, with G always providing a greater percentage of nonobliged responses than B. Fey (1986) suggests that a high number of topic maintaining or nonobliged responses is indicative of the child's attention to the detail of his partner's turn contribution and his willingness to take these details into consideration when formulating his own turn. That is, nonobliged responses indicate a higher level of consideration for the needs of one's conversational partner than obliged responses. G was able to use eye contact, eye gaze and orientation, all relatively effortless but effective nonverbal devices, to help regulate conversation and to check the attention status of his listener. In contrast, B necessarily relied more on verbal means, which are less automatic than many nonverbal devices, to regulate conversation and to monitor what was happening in the surrounding environment. Thus, B may have less time and energy to expend on the details of his partner's turn. Social Conversational Patterns of B and G Continuous versus Discontinuous Turns The proportion of continuous versus discontinuous turns for B and G reveals much about each twin's social-conversational pattern. Fey (1986) presents a classification system based on two continua; a responsiveness continuum and an assertiveness continuum. Conversational assertiveness involves a child's ability to take a turn when none is obliged. A child who falls high on the assertiveness continuum would frequently use such conversational acts as requests for attention, requests for interaction, and clarification 108 requests, as well as frequent topic initiations and topic extensions (i.e., openers). Thus, Fey's (1986) assertiveness continuum roughly parallels the present study's discontinuous coding category, with the exception of Fey's inclusion of clarification requests. Similarly, Fey's (1986) responsiveness continuum parallels the present study's continuous coding category. B has a roughly equal proportion of discontinuous turns, 40.9% of 477 turns, as compared to continuous turns, 41.9%. This pattern of social-conversational activity is most similar to Fey's "active conversationalist." B must exert more effort to be involved in social interaction. He cannot monitor another's attention with eye gaze, so he must use language or physical means (e.g. touch) to maintain and monitor attention. He also must do more to gain attention, hence the use of physical action strategies such as jumping up and down. B's high proportion of initiation bids and attention bids suggest that he is highly motivated to communicate with his peers. B is definitely a social child. B's active conversational style seems also to be related to how the other children chose to interact with him. For instance, one child was assigned weekly by the classroom teacher to be B's helper, responsible for escorting him to various locations in the classroom as indicated by the daily routine and choice during free play. Many of the girls took this responsibility very seriously and saw to it that B not only arrived in the right place but also that he was engaged in activity. For the children, the easiest way to ensure that B was actively participating was to control the content and direction of the play scenario. In such a situation where control is unevenly distributed, frequent topic initiations (i.e. openers) represent a good strategy for shaping the direction of the interaction. The children also frequently chose to interact with B in a structured manner. For example, during free play, N chose to request 109 from B the names of the small plastic animals she handed him. In such structured situations, B's frequent use of initiations and openers, as compare to G, may reflect an attempt to extend the interaction in a less structured and more playful manner. In comparison, G is lower on the assertiveness continuum than B, suggesting a large number of acknowledgements and agreement, or turns that contribute no new information. G is more of a passive conversationalist. He does not have to exert as much effort to initiate interaction; a simple eye gaze may do. G has a larger proportion of continuous turns (59.2% of 461) than discontinuous turns (33.0%). "Thus, G falls high on the responsiveness continuum, which Fey (1986) asserts is indicative of a child's attention to details of his partner's turn contributions and willingness to take these details into consideration when formulating his own turn. That B is more of an active conversationalist, as compared to G, who is more of a passive conversationalist is consistent with Zazzo (1977) who reports that identical twins in the same environment will emphasize social and personality differences to assert their individuality. Zazzo (1977) also submits that, in both monozygotic and dizygotic twin pairs, one twin will take primary responsibility for socializing. This twin will also demonstrate more highly developed social skills and social awareness. With regard to B and G, it would seem that B is the more outgoing twin. However, due to limitations imposed by his blindness, B does not present as the more socially aware twin. 110 Noninteraction Turns Noninteractional turns were defined as turns that were self addressed and had no apparent communicative intent. B used almost three times more noninteractional turns than G (compare 11.5% of discontinuous turns for B and 4.3% for G). Communicative intent was more difficult to ascertain for B than for G due to lack of eye-contact and lack of communicative use of orientation. Also included in this category were pretend play routines that were ascertained to be not addressed to anyone. It was often difficult to tell if B was merely talking aloud while engaged in pretend play or if he was using his pretend play as an elaborate strategy to engage his peers. For example the following was coded as self directed: B: hi doctor [= said while talking into a toy telephone]. B: hi. B: come to my bear will ya. B: bye. B: I did phone the doctor The last utterance was coded as an initiation but, alternatively, the whole excerpt could be coded as an elaborate pretend play routine to engage a peer into participation. Challenges of the Classroom and Peer Conversation Conversational interactions in a kindergarten classroom often involve peers and multiple participants. Both factors increase the complexity of the conversational skills needed as compared to those needed for dyadic conversation with adults. The nature of these complexities and the possible implications for a blind child are briefly considered in this section. I l l Peer conversational interaction is characterized by longer gaps than adult-adult conversation (Ervin-Tripp, 1979), so there is more time between turns for attention to stray. During a gap a child may get up and leave or may begin a side conversation. This is especially likely to occur if the listener is unsure whether the speaker is going to continue speaking or if the child is unsure whether the speaker's continued talk is still directed to him or her. Some of B's nonverbal and verbal characteristics, such as turning his head away from his listener and the use of continued verbal pretend play routines so as to engage a listener, may have added to the likelihood of losing his listener's attention during long gaps. Peer interaction also involves more conversational breakdown due to communicative immaturity, and thus necessitates more re-initiations (McTear, 1985). As was demonstrated, B has difficulty with re-initiations. Multiparty conversation, as compared to the dyad, is more complex, as one must (i) use complex monitoring skills to keep track of who is talking, (ii) wait longer for a turn, (iii) be specific when selecting a speaker and, (iv) compete for a turn (Ervin-Tripp, 1979). For a blind child, speaker monitoring in a multiparty conversation is even more challenging, because it must be done without the use of direction of eye gaze and eye contact. In addition, one must remember to use vocatives when selecting a speaker, as eye contact and orientation are not sufficient. With regard to competition for a turn, according to Sacks, Schegloff and Jefferson (1974), in multi-party conversations, the quickest person to speak earns the right to continue speaking. If one's cognitive load is taxed due to having to do all monitoring of who is speaking and predicting the next transition-relevance place all through the auditory channel, one is less likely to be first in getting a turn on the floor. 112 Due to these challenges of multiparty conversation, B's tendency to initiate one-to-one conversations as opposed to attempting to involve himself in ongoing conversations seems to be a compensatory strategy. It is difficult to follow any conversation that one has not been a part of from the beginning, and especially difficult if it is a multiparty conversation. Another related strategy was that B tended to interact more with his female peers than his male peers. Two reasons are submitted for this preference. First, it seemed that the girls were more accommodating and tended to exert more effort in making the interaction successful. A related point is that some of the girls adopted a "mothering" attitude towards B, such that these girls often sought B out during free play. Second, the girls tended to set up and evolve a play situation in one spot of the classroom (e.g. playing house). In contrast, the boys' play often involved much movement (e.g. building bridges and racing cars). Educational Implications Children who are blind should be encouraged to continue using effective strategies. For B, these included such devices as vocatives and the nonverbal device of touch. Although the use of touch alone does not guarantee the attention of a listener, teaching the child to turn his head in the direction of the other person when talking to him or her can increase the probability of maintaining listener attention. Informing the child of the importance of using specific, routinized attention-getting devices may also increase the probability of listener response. For example, B may have been more successful in gaining a response to initiations if he had used more expressives to mark his utterances as initiations as opposed to making general statements for the benefit of the other children. 113 There cannot be sufficient emphasis placed on the importance of peer interactions for the development of conversational competence. As was mentioned previously, B's classroom teacher did much to encourage peer interactions, such as assigning a specific child to act as B's buddy for that day. The buddy would accompany B during the day, often automatically including B in the same peer interactions involving his buddy. This inclusion technique is obviously valuable for a child who has difficulty entering new interactions. A related point is that peer education must be emphasized. Such peer education can be largely done through adult modelling. B's teacher was very good at modelling descriptive language; recall her description of how to find the teddy bear's mouth. The teacher and the aide could also model the use of open questions, so as to give B choices and let him decide what he wants to do. As was mentioned, the children, with the best intentions, often took control of interactions in which B was involved. Instead, the children could be encouraged to ask such questions as, "what do you want to do, B?" and "Let's play house, who do you want to be?" B's peers could also be taught to announce themselves when approaching B. For example, "Hi, B. This is N. Would you like to play?" In this way, B would hear the appropriate use of attention-getting devices. The children announcing themselves also would make it easier for B to use vocatives when addressing another child. Implications for Future Research Due to the small corpus, this study represents only a preliminary effort. Further research on conversational discourse with peers involving other children who are blind is indicated. 114 The coding taxonomy was devised to mark various kinds of conversational initiations (i.e. bids for interaction, bids for topic change, attempts to gain attention, and attempts to gain a turn), their success in gaining a response and responses. In addition, verbal and nonverbal utterances were marked. In general, the coding taxonomy worked well and allowed the necessary comparisons, as indicated by the research questions, to be made. Distinctions between different kinds of initiations led to some interpretative difficulties. For instance, re-establishment of a conversational interaction after an interruption (i.e. bids for interaction) often co-occurred with a change of topic. As well, difficulty ensued due to the fact that there were no detailed specifications outlining how large a topic extension needed to be to be coded as an opener, a problem common to all research on 'topic'. Part of this difficulty occurred due to an initial decision to try to de-emphasize the notion of topic as much as possible because of its inherent problems and to limit the scope of this research project. Suggestions for improvement on the present coding taxonomy include a more elaborated specification of topic information. Such information would also allow a distinction to be made between allowable topic extensions and tangential topic extensions so as to follow up on Kekelis and Andersen's (1984) report that some blind children make frequent tangential comments, and allow for a more detailed investigation of topical contingency in responses. Additional nonverbal characteristics should also be included in the coding system. Nonverbal actions such as the details of head orientation and eye-pressing behaviour seemed to impact on the organization of conversational turns and on attention status of the listener. As such, inclusion of this information in the coding system may provide valuable information. As well, further use of the coding taxonomy by other researchers would serve to identify and resolve remaining problems of categorization and reliability. 115 With regard to research methods in general, timing should also be considered as an additional measure of opportunity for interaction in general. Although turns and utterances were counted for each transcript, interactions were not timed. This modification would allow the investigator to avoid the limitation of interpreting purely proportional results. Consistent with other reports in the literature, B had great difficulty gaining a response to his bids for interaction. Further investigation of the verbal and nonverbal devices that other blind children use in their initial initiation bids and re-initiations and how these relate to success in gaining a response would be profitable. Such additional research would aid in advancing us further toward being able to definitively answer the question of why this difficulty occurs and perhaps offer explanations for how it may be resolved. Conclusion B and G differed greatly with regard to bids for interaction. Although the number of interactions each twin attempted to initiate was similar, B devoted proportionally almost twice as many turns to initiations as did G. The large number of turns B devoted to initiation seems to be related to the findings that 1) less than half of B's initiations gained a response; 2) B persisted for a greater number of turns than G after an initial bid failed, due to repeated failed attempts; and 3) B used a large proportion of nonspecific and nonroutinized attention-getting devices such as pretend play and prosodic devices, as well as relying on contextual support alone in the nonverbal domain. This is consistent with reports that blind children have difficulty initiating conversational interaction (Dunlea, 1989; Kekelis and Andersen, 1984; Mulford, 1983; Rowland, 1983). Comparison of B's conversational skills with those of his 116 sighted identical twin is a unique contribution to the research literature on blind children's language use. Individual differences among children on such measures are so great that it is difficult, in the usual case, to determine what differences should be attributed to the lack of vision, per se. The case of comparing identical twins is as good a control as is possible. However, due to the small number and size of samples and the single-subject nature of the study, definitive generalization cannot be claimed. 117 REFERENCES Andersen, E . (1984, November). Siblings' contribution to the interactions! competence of blind children. Paper presented at the American Anthropological Association annual meeting, Denver, Colorado. Andersen, E . , & Kekelis, L . S. (1982). The role of visual perception in early caregiver-child communication: Some "input" from the blind. Paper presented at the Seventh Annual Boston University Conference on Language Development. Bates, E . , Camaioni, L . , & Volterra, V. (1975). The acquisition of performatives prior to speech. Merrill-Palmer Quarterly. 21, 205-226. Blockberger, S. & Johnson, C . (1985, October). How big are you today? Paper presented at the Boston University Conference on Child Language, Boston. Bloom, L . , Rocissano, L . , & Hood, L . (1976). Adult-child discourse: Developmental interaction between information processing and linguistic knowledge. Cognitive Psychology. 8, 521-552. Brown, G . , & Yule, G. (1983). Discourse analysis. New York: Cambridge University Press. Bruner, J. S. (1977). Early social interaction and language acquisition. In H . R. Schaffer (Ed.), Studies in mother-infant interaction (pp. 271 - 289). New York: Academic Press. Camaioni, L . (1979). Child-adult and child-child conversations: An interactional approach. In E . Ochs & B. Schieffelin (Eds.), Developmental Pragmatics (pp. 325-337). New York: Academic Press. Chomsky, N. A. (1957). Syntactic structures. The Hague: Mouton. 118 Collis, G. M . , & Schaffer, H . R. (1975). Synchronization of visual attention in mother-infant pairs. Journal of Child Psychology and Psychiatry. 16. 315 - 320. Dunlea, A . (1989). Vision and the emergence of meaning: Blind and sighted children's early language. Cambridge: Cambridge University Press. Ervin-Tripp, S. (1978). Some features of early child-adult dialogues. Language in Society. 7(3), 357-373. Ervin-Tripp, S. (1979). Children's verbal turn-taking. In E . Ochs & B. Schieffelin (Eds.), Developmental Pragmatics (pp. 391-414). New York: Academic Press. Fey, M . (1986). Determining basic intervention goals. Language intervention with young children (pp.67-100). Pro-ed: Austin, Texas. Garman, M . (1983). The investigation of vision in language development. In A . E . Mills (Ed.), Language acquisition in the blind child: Normal and deficient (pp. 162 - 166). London: Croom Helm and San Diego: College-Hill. Goffman, E . (1976). Replies and responses. Language and Society. 5. 257-313. Grice, H . P., (1975). Logic and conversation. In P. Cole and J. Morgan (Eds.), Syntax and semantics. Volume 3: Speech acts (pp. 41-58). New York: Academic Press. Grusec, J. E . , & Lytton, H . (1988). Social development: History, theory and research. New York: Springer-Verlag. Hymes, D. (1971). Competence and performance in linguistic theory. In R. Huxley & E . Ingram (Eds.), Language acquisition: Models and methods (pp. 3 - 28). London: Academic Press. Jan, J . , Freeman, R., & Scott, E . (1977). Visual Impairment in children and Adolescents. New York: Grune and Stratton. 119 Johnson, C , & Kearns, K. (1991, October). General nominals and categories in the early lexicon: Evidence from blind children. Paper presented at the Boston University Conference on Language Development. Keenan, E . O. , & Schieffelin, B. B. (1976). Topic as a discourse notion: A study of topic in the conversation of children and adults. In C. N. L i (Ed.), Subject and topic (pp. 335-384). New York: Academic Press. Kekelis, L . , & Andersen, E . (1984, February). Family communication styles and language development. Journal of Visual Impairment and Blindness. 78. 54-65. MacWhinney, B. (1991). The CHILDES Project: Computational Tools for Analyzing Talk, Version 1.0. Pittsburgh: Carnegie Mellon University. McGinnis, A . R. (1981). Functional strategies of blind children. Journal of Visual Impairment and Blindness. 75, 210-214. McTear, M . (1985). Children's conversation. New York: Basil Blackwell. Mulford, R. C. (1983). Referential development in blind children. In A. E . Mills (Ed.), Language acquisition in the blind child: Normal and deficient (pp. 89-107). Croom Helm, London and College Hill, San Diego. Murphy, C. M . , & Messer, D. J. (1977). Mothers, infants and pointing: A study of gesture. In H . R. Scaffer (Ed.), Studies in mother-infant interaction (pp.325 -354). New York: Academic Press. Ochs, E . , Schieffelin, B. B., & Piatt, M . L . (1979). Propositions across utterances and speakers. In E . Ochs & B. B. Schidffelin (Eds.), Developmental pragmatics (pp. 251-268). Academic Press: New York. Parke, K. L . , Shallcross, R., & Anderson, R. J. (1980). Differences in coverbal behavior between blind and sighted persons during dyadic communication. Journal of Visual Impairment and Blindness. 74, 142-146. 120 Prizant, B. (1984). Towards an understanding of language symptomatology of visually-impaired children. In A. W. Sykanda, J.E. Jan, S.J. Blockberger (Eds), Insight in sight. Proceedings of the Fifth Canadian Interdisciplinary Conference on the visually impaired child, CNIB, Vancouver. Prizant, B. , & Booziotis, K. Language acquisition strategies in a blind child. Unpublished manuscript. Rowland, C. (1983). Patterns of interaction between three blind infants and their mothers. In A. E . Mills (ed.), Language acquisition in the blind child: Normal and deficient (pp. 114-132). Croom Helm, London and College Hill Press, San Diego. Sacks, H . , Schegloff, E . A . , & Jefferson, G. (1974). A simplest systematics for the organization of turn-taking for conversation. Language. 50 (4). 696-735. Schaffer, H . R. (1977). Early interactive development. In H . R. Scaffer (Ed.), Studies in mother-infant interaction (pp. 271 - 289). New York: Academic Press. Schaffer, H . R., Collis, G. M . , & Parsons, G. (1977). Vocal interchange and visual regard in verbal and preverbal children. In H . R. Schaffer (Ed.), Studies in mother-infant interaction (pp. 291 -324). New York: Academic Press. Schegloff, E . A . , & Sacks, H . (1973). Openings and closings. Semiotica. 8. 289-327. Snow. C. E . (1977). The development of conversation between mothers and babies. Journal of Child Language. 4, 1-22. Sugarman, S. (1983). Empirical versus logical issues in the transition from prelinguistic to linguistic communication. In R. Michnick Golinkioff (Ed.), The transition from prelinguistic to linguistic communication. Hillsdale: Lawrence Erlbaum. Trevarthen, C. (1977). Descriptive analysis of infant communicative behaviour. In H . R. Schaffer (Ed.), Studies in mother-infant interaction (pp. 227-270). New York: Academic Press. 121 Urwin, C. (1978). The development of communication between blind.infants and their parents. In A . Lock (Ed.), Action, gesture and symbol: The emergence of language (pp. 79-108). Academic Press: New York. Urwin, C. (1983). Dialogue and cognitive functioning in the early language development of three blind children. In A. E . Mills (Ed.), Language acquisition in the blind child: Normal and deficient (pp. 142-161). Croom Helm, London and College Hill: San Diego. Wellman, H . M . , & Lempers, L . D. (1977). The naturalistic communicative abilities of two-year-olds. Child Development. 48, 1052-1057. Zazzo, R. (1977). Genesis and peculiarities of the personality of twins. In W. E . Nance (Ed.), Progress in clinical and biological research: Vol.24a. Twin Research: Proceedings of the Second International Congress on Twin Studies (pp. 1-11). New York: Alan R. Liss. 42 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0098947/manifest

Comment

Related Items