t)! "W 29%“ I‘JJJJ‘JJ‘I'JJ 'J‘J'J-JM‘JW ’:,3 f JJJJJHLUJJJJJ JJ“ I" .J| J. J" ‘3“ m", ‘ ‘ 'th 3'“ “In 'H TJJJJJJ. JflnaJJ - JJ: 1“ fiJhflJJJthhJJ; q f”. 1! , J “:LJJ ”W“ H" :J MW" JJ‘JJJJ- .: J,J.‘J.J,J, "'J'J‘JJJJ J JJJJJ . ' .w J ,J, JIJJIJJ “VJ , ‘ :‘JJ JJJJJJJfit . NJ”: ~o _-:_J _. 'J in“ ’3" MI J szJ-H"! MJLJ J. ’4 Jill“; w ' }IJ 'J‘El J'H‘H “A”! I“: “1"“ "I “H“ J‘ JJJLJJJ :1...:—“:tj -.—_.-—_q—\_’- M F! H: .- f —— Au. ‘ .A... -——-- -.;. . J¢ - , “4:":— .q,. _ .-. '3,“ . = 2 ’ J 39?.“ ' . ‘9‘??? - VAA. a o n. _. $7.11? tax... ' - ’i-r . 51:33? . I 1":“33354' :Q n ”:3 D A?! - ’5 . 32:: ”.13 m own—6 ‘ ”vow.— . V .g.-_.-_V ‘ . - “-40-on . :IEIJ “4: ' Wm W J W . ‘WIJA‘JJI “5' NJ ': 37L PM!" 343% iJJJJJJxl} ’ “2': {Ln '1. zJiJ! .-, 15115, J ‘J . ”1' {J 2., 7,1: J Jill’s: El; ‘ ’-f . I“ 3’ "4‘ 8i". '_ 31?: 1! J‘IJJJJ .5114 J “Ev—4%. . J Y. J JJJ. JM JJJ‘JJJI J J "J; JAM-Jab I. “BIB-'1‘ FWJ” J THESIS This is to certify that the thesis entitled PRAGMATIC AND SOCIAL BEHAVIORS OF LANGUAGE-IMPAIRED ADOLESCENTS presented by Karen Ann Hux has been accepted towards fulfillment of the requirements for Master of Arts degree in Speech Pathology Major profe so Sandra A. Hayfield Date 4624441 (/7; ”f3 0-7539 ‘ MS U is an Affirmative Action/Equal Opportunity Institution J'N“ RH“; .- ka‘ “.3?".‘ I? run a".-. _ g £7»: m M t MSU LIBRARIES n RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. item a» «ma-e , - . _ y {, *r'rWF . '1‘ -' ‘ . ' r. , . _ . _ _, rt . , ‘ 35 it "'~ ‘4': e ‘. I. 3' "'g 3‘. II; 7:.“- unjl‘ ”i g ii “‘1 F7 ""7 ‘?‘ a” "‘7 9'7“ :“i g... ’f'f‘ "' ;. -. ~ ,' r} ' 4 ‘ 1’“ ."' 7 L .57- - a: warm a A! n. /4//~ a: 7d“; PRAGMATIC AND SOCIAL BEHAVIORS OF LANGUAGE-IMPAIRED ADOLESCENTS By Karen Ann Hux A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Audiology and Speech Sciences 1983 Copyright by KAREN ANN HUX 1983 ABSTRACT PRAGMATIC AND SOCIAL BEHAVIORS OF LANGUAGE-IMPAIRED ADOLESCENTS By Karen Ann Hux This study investigated the pragmatic behaviors and the actual and perceived social status ratings of language- impaired adolescents. Subjects included nine normal- language and three language—impaired l3— and 14-year-olds who were grouped into six dyads. Each dyad participated in an instruction—giving taSk. Each subject also completed either a Peer Rating or Predicted Peer Rating Scale. There were four major findings of the study. First, the language- impaired speakers made more frequent accuracy and clarityi pragmatic violations than did the normal-language speakers. Second, semantics was the underlying contributor to the difficulties. Third, post7hoc analysis revealed that the occurrence of confused anaphoric pronoun referents resulted from underlying semantic disorganization. Finally, language-impaired adolescents did not receive low social status ratings from normal-language peers. The results con- cerning the language-impaired adolescents' perceptions of their social status were inconclusive. ACKNOWLEDGMENTS I wish to express my sincere appreciation and thanks to my thesis advisor, Dr. Sandra Mayfield, for her support and guidance in the completion of this research project. I thank her for so willingly sharing her time and knowledge with me. Her encouragement, friendship, and enthusiasm during the extent of my graduate studies have helped me tremendously. I also wish to thank the other members of my thesis committee, Dr. Michael Casby and Dr. Leo V. Deal. I am grateful for their friendship, as well as the support and‘ interest they have expressed in this research. A special note of thanks is extended to Dr. Ida Stockman for her comments and assistance concerning the final writing of this manuscript. To Maria Fotias, my friend and colleague, I extend special thanks for assisting me in obtaining reliability measures and, more importantly, for keeping me laughing throughout the last year. I wish to thank the students, parents, and school personnel whose cooperation and participation permitted the completion of this study. I also thank my friends and colleagues who assisted me in locating potential subjects and who participated in various preliminary activities ii which were crucial to the development of this study. My appreciation is extended to Robin Blais for so patiently and expertly typing this manuscript. Finally, I would like to give special thanks to my parents for their continued love and support when it was needed the most. iii TABLE OF CONTENTS LIST OF TABLES. LIST OF FIGURES . . . . . . . . . . INTRODUCTION. REVIEW OF THE LITERATURE. Definition of Terms. Pragmatics. Functions Speech acts Pragmatic Taxonomies Description Problems. Pragmatics in Language-Impaired Children Social Status Ratings and Communication. Lowered social status ratings Explanations for lowered social status ratings. STATEMENT OF THE PROBLEM. METHOD. Subjects Groupings . . . . . . . . . . . . . Selection and screening Materials. Puzzle designs. iv vii ix ll 13 18 18 21 29 32 32 32 33 36 36 Equipment and setting Social status ratings Procedures Instruction-giving task Social status ratings Transcriptions and Data Analysis Procedures. Utterance functions Communication acts. Maxims. Listener behavior Non—pragmatic factors "Teacher-student" responsibility. Inter-judge reliability RESULTS AND DISCUSSION. Pragmatic Behaviors: Communication Act Descriptions. Number and nature of communication acts Number of unsuccessful communication acts Responsibility for success or failure Pragmatic violations. Pronoun use Non-pragmatic factors Social Status. Ratings Predictions SUMMARY AND CONCLUSIONS APPENDICES A. Parent Questionnaire V 37 37 38 38 ”1 U1 7+2 H2 as us 49 51 53 5H 5H 75 79 82 96 105 121 121 123 126 133 B. Education Levels of Subjects' Parents. . . . . 135 C. Raw Scores and Age—Equivalency Scores on the PPVT-R. . . . . . . . . . . . . . . . . . . 136 D. Designs for Instruction-Giving Task. . . . . . 139 E. Peer Rating Scale and Predicted Peer Rating Scale . . . . . . . . . . . . . . . . . . . 151 F. Drawings of Facial Expressions Used to Represent the One— to Five-Point Peer Rating Scales . . .‘. . . . . . . . . . . . 152 G. Directions.for Instruction—Giving Task . . . . 153 H. Definitions and Examples of Dore's (1978) Conversational Act Categories . . . . . . . 155 I. Percentages of Inter—Judge Agreement on Pragmatic Analyses Across Dyads and for Each Individual Dyad. . . . . . . . . . . . 155 J. Number of Conversational Acts for Each Utterance Function Category . . . . . . . . 166 K. Type and Number of Conversational Acts Within Each Utterance Function Category for the Experimental and Control "Teach- ers" and "Students" . . . . . . . . . . . . 168 L. Analyses of Anaphoric Pronoun Use. . . . . . . 170 REFERENCE NOTES . . . . . . . . . . . . . . . . . . . 179- REFERENCES; . . . . . . . . . . . . . . . . . . . . . 180 vi Table 10. LIST OF TABLES Number, Mean, Variance, Standard Deviation, and Range of Successful or Unsuccessful Communi- cation Acts in the Control and Experimental Groups Number of Successful and Unsuccessful Acts for which "Teachers" or "Students" were Respon- sible in the Control and Experimental Dyads. Ranked Percentages of Pragmatic Violations for the Control and Experimental Groups. Number, Mean, Variance, Standard Deviation, and Range of Pragmatic Violations for the Control and Experimental Groups. Ranked Number of Pragmatic Violations for Individual Dyads Number of Anaphoric, Cataphoric, and Exophoric Pronouns Used by Control and Experimental "Teachers" Number, Mean, Variance, Standard Deviation, and Range of Anaphoric or Exophoric Pronouns Used by Control and Experimental "Teachers". Number of "Teacher" Anaphoric Pronouns that were Confused or Not Confused in the Control and Experimental Dyads Number, Mean, Standard Deviation, and Range of "Ambiguous" or "Incorrect" Word Choice Errors for the Control and Experimental Groups Actual and Predicted "Teacher" Social Status Ratings in the Control and Experimental Groups and in Individual Dyads Analysis of Teacher C-M's Anaphoric Pronoun Use. vii Page 78 81 85 86 88 98 99 103 107 122 170 Table Analysis Use. Analysis Use. Analysis Use. Analysis Use. Analysis Use. of of of of of Teacher Teacher Teacher Teacher Teacher C-Fl's Anaphoric Pronoun C-F2's Anaphoric Pronoun E-M's Anaphoric Pronoun E-Fl's Anaphoric Pronoun E-F2's Anaphoric Pronoun viii Page 172 173 179 177 178 LIST OF FIGURES Figure 1. Percentage of utterance function categories produced by the control and experimental groups . . . . . . . . . . . . . . 2. Percentage of conversational acts within the control and experimental group's "Request" and "Response" category functions. 3. Percentage of conversational acts within the control and experimental group's "Organiza- tional Device" and "Miscellaneous" category functions. . . . . . . . . . . . . . . . . . H. Percentage of conversational acts within the control and experimental group's "Statement" and "Description" category functions 5. Percentage of conversational acts within the control and experimental group's "Performa- tive" and "Acknowledgment" category functions. . . . . . . . . . . 6. Percentage of utterances produced by control and experimental "teachers" and "students" for each function category . . . . . . 7. Percentage of control and experimental "teacher" utterances in each conversational act of the "Request" and "Response" cate- gories and in the collapsed "Other" cate- gory . . . . . . . . . . . . . . . 8. Percentage of conversational acts within the control and experimental "teachers'" "Request" and "Response" function cate- gories . . . . . . . . . . . . . . . . . . . 9. Percentage of control and experimental "stu- dent" utterances in each conversational act of the "Request" and "Response" categories and in the collapsed "Other" category. . ix Page 58 60 62 6Q 66 67 68 71 Figure Page 10. Percentage of conversational acts within the control and experimental "students‘" "Request" and "Response” function cate— gories . . . . . . . . . . . . . . . . . . . 72 11. Percentage of successful and unsuccessful com- munication acts for the control and experi- mental groups and for each individual dyad . 76 12. Percentage of successful communication acts for which "teachers" or "students" were respon- sible in the control and experimental group. . . . . . . . . . . . . . . . . . . . 80 13. Percentage of unsuccessful communication acts for which "teachers" or "students" were responsible in the control and experimental groups and in each individual dyad . . . . . 83 19. Percentage of pragmatic violations for each control and experimental dyad. . . . . . . . 87 15. Percentage of anaphoric and exophoric pronouns used by "teachers" in the control and ex- perimental groups and in each individual __ dyad . . . . . . . . . . . . . . . . . . . . 100 16. Percentage of "teachers'" anaphoric pronoun referents that were confused or not con- fused in the control and experimental groups and in each individual dyad . . . . . 10a 17. Percentage of "ambiguous" word choice and "incorrect" word choice errors for the "teacher" — responsible communication act failures in the control and experimental groups and in each individual dyad . . . . . 106 18. Percentage of "teachers'" "Quantity," "Clari- ty," and "Topic" violations that were a result of "ambiguous" word choice or "incorrect" word choice errors in the control and experimental groups and in each individual dyad . . . . . . . . . . . . 109 19. Percentage of "teachers'" "Accuracy" violations that were a result of "ambiguous" word choice or "incorrect" word choice errors in the control and experimental groups and in each individual dyad . . . . . . . . . . . . 110 Introduction ‘ A major purpose of communication is the establishment and maintenance of social interactions and relationships (Watzlawick, Beavin, 8 Jackson, 1967). Pragmatics is the area of language that incorporates a speaker's intents with the social rules and varying contexts of which he must be cognizant to be an effective communicator (Lucas, 1980; Prutting, 1982; Rees, 1978). It is not surprising, there- fore, that a disruption in an individual's pragmatic skills will affect his overall communicative competence and, in turn, his social interactions with peers. _‘ Recently, a variety of taxonomies to describe child- ren's language use (pragmatics) have been proposed (Dore, 1978; Halliday, 1975; Tough, 1977). These systems primarily emphasize the acquisition of language use in the pre-school and early elementary school years (Chapman, 1981). Some have compared the frequency of particular pragmatic fea- tures in the speech of language-impaired versus normal- language children (Coggins 8 Carpenter, 1978; Gallagher 8 Darnton, 1978; Miller, 1978; Tough, 1977). Taxonomies for older children and adults are also available, although they primarily focus on social rules regulating conversation (Grice, 1975) and politeness forms (James, 1978; Lakoff, 1973) rather than specific utterance or discourse level 1 analyses. Lacking, however, is the application of these systems to language-impaired adolescents. This is not sur- prising since the study of adolescent communication dis- orders is a relatively recent phenomenon with many questions left unexplored (Boyce 8 Larson, 1983). In all cases, however, the pragmatic taxonomies are irooted in Searle's (1969) Speech Act Theory (Chapman, 1981; Prutting, 1982). The recent interest in pragmatic aspects of communication stems from a recognition that the speech act is a theoretical unit of communication that involves both a speaker and a hearer--that is, a dyad (Austin, 1962; Searle, 1969). The speech act, therefore, has evolved as the basic unit of pragmatics (Lucas, 1980); and many researchers now consider dyad pairs as the minimal unit_for analysis of human communication patterns (Brinton 8 Fujuki, 1982; Prutting, 1982; Retherford, Schwartz, 8 Chapman, 1981). However, once again the research involving language- impaired adolescents has only touched the surface of the dyad concept for communication analysis (Boyce 8 Larson, 1983). Also lacking is sufficient information about the social consequences of violating various pragmatic rules for language-impaired adolescents (Boyce 8 Larson, 1983). The little information that is available on selected aspects of children's social behavior in communication-oriented tasks generally is limited to school—age learning-disabled children. For example, learning-disabled children have been found to receive lower social status ratings from normal-learning peers (Bruininks, 1978; Bryan, 1979a, 1976; Goldman 8 Hardin, 1982; Scranton 8 Ryckman, 1979). Although these studies have not identified these youngsters as hav- ing specific linguistic or pragmatic impairments, it is probable that at least some of the subjects did manifest communication difficulties. Such an assertion is plausible since the U.S. Office of Education defines learning dis- ability as an impairment that involves at least one of the fundamental psychological processes necessary for the com- prehension and use of spoken and/or written language in the absence of emotional, intellectual, or sensory deficiency (1977). While the literature documents the existence of lowered social status ratings for school-age learning- disabled children, the phenomenon has not been explored for either school-age or adolescent youngsters classified as oral-language-impaired. Finally, to this writer's know- ledge, it also is not known how quickly normal-learning or normal-language youngsters form social impressions about learning~disabled or language—impaired peers. Given the limited amount of available data on the lin- guistic, pragmatic, and social behaviors of language- impaired adolescents, both descriptive and experimental research are warranted in this area (Boyce 8 Larson, 1983). The purposes of this study, therefore, are (1) to describe pragmatic behaviors of language-impaired adolescents in a dyad situation, (2) to determine whether language-impaired adolescents receive low social status ratings from normal- language peers, and (3) to determine whether language- impaired adolescents are less accurate than their normal- language peers in predicting their social status. Review of the Literature The following literature review will (1) define basic terminology essential for discussing pragmatic behaviors, (2) discuss major pragmatic taxonomies revelant to this study, (3) describe the pragmatic behaviors of language- impaired children observed in dyad situations, and (H) summarize findings regarding the social status ratings of children with language-based learning disabilities. Definition of Terms Pragmatics. Several authors have attempted to define the term "pragmatics." As a result of this, the words and sentence forms used to define pragmatics vary considerably. However, the general content or meanings of the definitions are strikingly similar. For example, Watzlawick et a1. (1967) state that pragmatics is the behavioral effects of communication. Others rather loosely define pragmatics as the functions or uses of language (Bloom 8 Lahey, 1978; Halliday, 1975). Still others define pragmatics as the use of language in context (Bates, 1976; Prutting, 1982). In many cases, pragmatics is viewed as an integral part of competence for communication and for social interaction (Hymes, 1971; Prutting, 1982; Searle, 1969). Such broad definitions allow several areas to fall within the realm of "pragmatics," including studies of mother—child interaction (Lasky 8 Klopp, 1982), contextual effects on language (Haniff 8 Siegel, 1981), discourse or conversational rule analyses (Gallagher 8 Darnton, 1978), and communicative intent or language functions (Blank, Gessner, 8 Esposito, 1979; Halliday, 1975). For the purposes of this study, pragmatics is defined as the social use of language in context to express a variety of communication intents or functions to accomplish a particular goal in an efficient and effective manner. Functions. When discussing pragmatics or pragmatic functions, a distinction needs to be made between the func- tions or purposes which individual utterances may serve for a speaker and the broader functions which language may serve for communication. Such a distinction is necessary because a speaker's intended meaning may be conveyed by a variety of grammatical forms--that is, the function of the utterances (Rees, 1978). Equally true is that one gram— matical form may convey several different meanings--that is, the function of language (Rees, 1978). For example, the grammatical form of the declarative utterance, "I'm getting hungry," serves the utterance func- tion of reporting an internal state (Dore, 1978). However, the utterance also may serve at least two other broader language functions. One broader function is reporting information, and the other is an indirect request for action. Furthermore, the language function of reporting information is the same as the utterance function of report- ing internal state. In other words, a sentence can serve more than one utterance function and it can simultaneously serve both utterance functions and language functions. Such multiple functions for a single sentence demonstrate that the study of pragmatics is a complex endeavor that should not be limited to descriptions of utterance func- tions only or of language functions only (Chapman, 1981). Speech acts. Searle's (1969) Speech Act Theory en- couraged the distinction between utterance functions and language functions by introducing the speech act as the theoretical unit or minimal unit of analysis for the study of communicative intent rather than the sentence. Searle (1969) proposed that every speech act contains a locutionr ary act, a propositional act, and an illocutionary act. The locutionary act involves the actual motoric act of speaking. The propositional act refers to the content and semantic meaning of the utterance. The illocutionary act refers to the speaker's intended meaning. A speech act also may contain a perlocutionary act which is simply any effect a speech act may have on a listener‘s behavior. For example, "I'm getting hungry," conveys a proposi- tion that the speaker is experiencing a sensation of hunger (propositional act). The utterance also indicates that the speaker intends to convey that he is hungry to the listener (illocutionary act). Once the speaker articulates the mes- sage, the speech act has been uttered (locutionary act). One possible effect of this message on the listener (per- locutionary act) would be for the listener to go immediately to the kitchen and begin to prepare a meal for the speaker. This resulting action behavior from the listener is consid- ered the perlocutionary act or the effect of the speaker's speech act on the listener. Obviously, the speech act concept requires that at least two participants are present for a speech act to occur. The pair of participants who carries out the speech act is often referred to as a dyad (Prutting, 1982). The dyad, therefore, is considered the minimal unit for the analysis of communication (Austin, 1962; Prutting, 1982; Searle, 1969; Watzlawick, 1967). To analyze communication behaviors within dyads, investigators have attempted to operationalize boundaries for a single "speech act." The sizes of the "speech act" units have varied depending on the level of analyses used by the investigators (Prutting, 1982). For example, some authors have analyzed functions at the utterance level (Dore, 1978; Searle, 1976), while others have analyzed functions at the broader conversational (i.e., discourse) language level (Garvey, 1975). Some investigators also have defined the speech act unit by the number of turns taken by the speaker and the listener. Investigators' definitions of "speech act" by number of turns have ranged from two-turn exchanges (Dore, 1979) to three- (Garvey, 1975) and four-turn (Gallagher, 1981) sequences or interchanges. Thus, the boundaries of a single speech act as a unit of analysis vary with the investigator and his particular research goals. Pragmatic Taxonomies Description. Based on Searle's (1969) Speech Act Theory, a variety of taxonomies have been developed to describe the communicative functions of adults and child- ren (Dore, 1975; Halliday, 1975; Tough, 1977). One such taxonomy was developed by Grice (1975), who established four maxims for conversation. The maxims describe the con- versational rules that adults use to maintain a sense of cooperation during communication. Grice's (1975) first maxim pertains to the quantity of information that the speaker gives to the listener. The._ speaker needs to provide enough information to be informa— tive but must not supply irrelevant or redundant informa- tion. Grice's (1975) second maxim involves the quality of the speaker's utterance. The speaker must attempt to be accurate and truthful in what he says. He should not speak about things for which there is insufficient evidence or about things which he does not know to be true. The third maxim, relation, concerns the topic of and situation for the speaker's utterance. The speaker should make his con- tributions relevant to the topic being discussed and appro- priate to the on-going situation. Grice's (1975) final maxim, manner, refers to the clarity of the speaker's utterance. The manner of a speaker's conversation is appropriate when he maintains clarity in his message. Clarity is achieved by avoiding ambiguity, by being orderly, and by being as brief as possible without sacrificing impor- tant information. Another taxonomy based on Searle (1969) was developed by Dore (1978). Dore (1978) classified the utterances of normal pre-school age children on the basis of grammatical form, interactional accomplishment (social context), and illocutionary function (intended meaning). Specifically, the grammatical form of an utterance and the social situa- tion in which it occurred are used to determine what the speaker intended to mean in his message. For example, a question such as, "Is he finished with it?" would be classi- fied as a Yes/No Question within the Request category. This classification emphasizes the grammatical form of the utter- ance (yes/no question) and its illocutionary function (to request information). Conversely, classification of a ques- tion such as, "Can I fix your tie?" as a Permission Request relies heavily on the influence of social context in deter- mining the intended meaning. In this case, the social con- text must be one in which the speaker needs to gain permis- sion from the listener before engaging in a specified activity. . A final example of a taxonomy based on Searle (1969) is that of Ellyn Lucas (1980). Lucas (1980) approached the study of pragmatics by describing a set of underlying se- mantic rules that specify the assumptions that speakers 10 make and the knowledge they use during the act of communi- cating. Although her work is not directed towards a spe- cific age group or population, Lucas (1980) emphasizes the 'importance of a child's acquisition of four underlying se- mantic rules as the basis for communicating pragmatic intent. These rules, which speakers and listeners use along with underlying conceptual knowledge to achieve the purposes of speech acts, are outlined within Searle's (1969) consti— tutive rules--that is, speaker conditions necessary for communicating to a listener. The first semantic rule or speaker condition is the Propositional Content or Meaning Rule. The Meaning Rule refers to the speaker's expectation that the listener will comprehend the content of the utterance and realize that_he will be expected to respond to that content in some manner. The second semantic rule is the Preparatory or Assumption Rule and it involves variables of presupposition. Specifi- cally, the speaker assumes or presupposes that the listener is able to do what the speaker asks and that without the speaker's verbal or nonverbal message, the listener would not act. Also, the speaker's assumptions about the relative status of the listener and the appropriate forms of address or degree of formality for that listener's status are addressed by this rule. The third semantic rule, the Sin- cerity Rule, states that the speaker must attempt to be honest and truthful except in cases of joking and sarcasm. 11 The last rule is the Essential Point Rule of the speech act. The Essential Point Rule refers to the listener's attempt to carry out or satisfy the goals or requests of the speaker. .The speech act is not completed until the listener attempts to respond appropriately to the intended meaning of the speaker's message. For example, if the speaker says, "Get the book," the essential point of the speech act will be satisfied when the listener actually attempts to ful- fill the speaker's request. The speech act is not complete until the listener tries to satisfy the essential point of getting the book. Lucas' (1980) taxonomy assumes a close relationship between pragmatics and semantics. This assumption forms the basis of Lucas' thesis that pragmatic disorders result from underlying semantic deficits in language-impaired children. Unlike Dore's (1978) taxonomy which is heavily influenced by syntactic or grammatical aspects of language, Lucas' (1980) taxonomy is heavily influenced by semantic and cognitive processes. Problems. Chapman (1981) cites several problems which may be associated with the use of current pragmatic taxono- mies. One problem apparent in taxonomies such as Dore's (1978) involves the differentiation of classification cate- gories on the basis of grammatical form. When a child's linguistic development is severely limited, it is inappro- priate to categorize his communicative intents on the basis of the functions expressed by the grammatical form of his 12 individual utterances. An analysis of the child's general language functions, rather than utterance functions, may provide a more accurate description of the child's communi- cation abilities since grammatical form alone would not dictate the pragmatic or functional classification. Fail- ure to classify by language function may lead to the identification of linguistically delayed children as being "pragmatically" different (Chapman, 1981, p. 132) when, in fact, they are not. In other words, their delayed grammati- cal systems may result in limited utterance function classification. A second problem with pragmatic taxonomies is the fact that a speaker's communicative intents can only be inferred (Chapman, 1981). Thus, although a group of listeners may_ agree in their classification of a given utterance, they can not say with certainty that their interpretation matches the speaker's original intent. Interestingly, this issue of differences in interpretation can also be described as "general semantics" (Deal, Note 1). A final problem with pragmatic taxonomies is that multiple codings are possible for utterances. As discussed earlier, this is largely a result of utterances being defined in terms of grammatical units rather than by their underlying propositions; conse- quently, one utterance can serve several functions. The problems associated with current pragmatic taxono- mies should not be overlooked by investigators who are attempting to select a pragmatic classification system for 13 research purposes. If the problems are ignored, investi- gators may lose important information about language func— tions. One way information may be lost is by classifying functions solely on the basis of grammatical form. Another way to lose information is by misinterpreting a speaker's intent. Finally, information may be lost by forcing single classifications for utterances that may be coded in more than one category (Chapman, 1981). Investigators, there- fore, should cautiously select pragmatic taxonomies that best meet the goals and needs of their particular project. To summarize, several pragmatic taxonomies have been developed since the introduction of Searle's (1969) Speech Act Theory. Problems are associated with the use of any one system since investigators' biases and assumptions will __ influence their classification categories. For example, use of one system rather than another may prevent multiple cod- ing of utterances or speech acts that serve several func- tions. In addition, depending on the taxonomy used, indi- vidual utterances rather than complete discourse units may be classified. The most feasible solution to such diffi- culties is either to select the taxonomy that most closely meets the research goals of the investigator (Chapman, 1981) or to use more than one taxonomy to describe the multiple levels of speech acts to be studied. Pragmatics in Language-Impaired Children With the increased interest in the pragmatic develop- ment of young children's language, attempts have been made 19 to describe the speech acts of children demonstrating diffi- culty with the acquisition and use of language; however, these attempts have been limited to pre-school and school- age children. Rom and Bliss (1981) compared twenty language-impaired children in Brown's Stages III and IV with twenty normal-language children of comparable linguistic development and twenty normal—language children matched on the variable of chronological age. The mean ages of the language-impaired subjects and the normal-language subjects matched on chronological age were 9.29 and 9.28 years, respectively. The younger, linguistically similar normal- language subjects had a mean age of 2.79 years. The frequency of speech acts produced by each child was recorded while he played with a familiar peer-~that is, in a dyad situation. Each child's language functions were analyzed using a combination of the pragmatic taxonomies of Dore (1975), Halliday (1973), and Schachter, Kirschner, Klips, Briedricks, and Sanders (1979). However, Rom and Bliss (1981) did not include transcript samples of the spe- cific function types discussed in their study. In any case, similarities were noted among the three subject groups' uses of Interjecting, Repeating, Requesting, Practicing, Self-Expressing, and Protesting Functions. All of the subjects used the Describing Function most fre- quently. The language—impaired and linguistically similar children, however, produced significantly fewer utterances overall and fewer Describing and Acknowledging Functions 15 than did the normal-language children matched on chronoé logical age. From this, the authors concluded that language—impaired children were quantitatively, rather than qualitatively, different on pragmatic variables from normal- language children of the same age. Soenksen, Flagg, and Schmits (1981) completed a prag- matic analysis of the functional communication skills of two learning-disabled eight-year-old boys in comparison to two normal—learning boys of the same age. Each subject participated in a 30—minute dyadic interaction with each of the other boys. Thus, a total of six interaction periods were analyzed. Conversations obtained while the children played together were coded using Halliday's developmental language functions (1978). The mean length of utterance_~ (MLU) was calculated for each child during an initial Adult-Child language sampling procedure and during each of the three Child—Child interaction periods. The results of the pragmatic analysis indicated that all of the children used the Regulatory, Personal, and Informative Functions most frequently during the dyadic interactions. Each child demonstrated a particular conver- sational style that was unique to him. However, each of the learning-disabled children relied heavily on the use of only one language function rather than shifting among the functions demonstrated to be in his repertoire. The normal- learning students showed more flexibility in using language functions by frequently alternating their use of functions. 16 Comparisons of the MLU calculations revealed that the learning-disabled children did not change their utterance lengths significantly between the Adult-Child and Child- Child interactions. The normal-learning students, however, modified their MLU's with respect to the person with whom they were interacting. The lack of change in utterance length indicates that the learning-disabled students were less likely to engage in "code-switching" (p. 285)--the modification of language production as a function of the situation and listener characteristics—-than were the normal-language children. Additionally, when compared with normative data, the MLU's produced by the learning-disabled subjects were representative of younger children's speech. Spekman (1981) conducted an investigation of the com- munication skills of twelve learning-disabled and thirty- six normally-achieving fourth- and fifth-grade boys. Boys with comprehension, syntax, or word retrieval problems were excluded from the study in order to insure that differences found between the learning-disabled and normal groups would be attributable to pragmatic competence. The subjects were grouped into twenty-four dyads, twelve of which paired a learning-disabled and normally-achieving child together and the remaining twelve composed of normally-achieving chil- dren only. Each subject participated in problem-solving tasks in which he performed in both Speaker and Listener roles. Auditory and visual feedback between dyad members was varied in three different ways. In Condition 1, the 17 dyad members were allowed to converse freely with one another and could see each other's faces. In Condition II, the subjects again were permitted free verbal exchange but could no longer view one another. Condition III further restricted the available feedback by permitting only the Speaker to talk and by eliminating visual cues as in Condi- tion II. Dyadic interactions were analyzed with respect to overall success, task efficiency, amount of interaction, information content, and task completion time. Analyses of the Speaker's utterances included information content, Gestalt information, and response type ratios. Listener variables included Listener adequacy, total questions, needed requests, and redundant requests. The results of_- these analyses indicated that differences existed between learning-disabled and normally—achieving youngsters in lan- guage comprehension and production skills, particularly with respect to general efficiency. The learning-disabled children tended to supply a smaller amount of pertinent information as Speakers and to ask fewer relevant questions as Listeners. Although the learning-disabled subjects demonstrated knowledge of spatial and attribute terminology required for the task during the pre-test, they had diffi- culty using this vocabulary appropriately in the interaction context. Thus even though they apparently were not identi- fied as language-impaired by formal assessment procedures, 18 they were hindered by language difficulties in the comple- tion of the communication task of this study. In summary, young language-impaired children have been found to differ both quantitatively and qualitatively from their normal-language peers on pragmatic variables. Quanti- tative differences are seen in language-impaired children's production of utterances in frequencies resembling those of younger, normal-language children. Qualitative differences are seen in language-impaired children's lack of flexi- bility in using a variety of utterance functions and in their failure to modify their language production to best meet the listener and situation needs-~that is, to engage in code-switching or register shifts. Additionally, chil- dren with language-based learning-disabilities have been-1 noted to supply a smaller quantity of pertinent information to their listeners, to ask fewer relevant questions, and to be less efficient overall in their use of language to com- plete a task. Social Status Ratings and Communication Lowered social status ratings. Oral communication skills are crucial in the establishment and maintenance of satisfying social interactions (Prutting, 1982; Watzlawick et al., 1967). Since learning-disabled children often exhibit deficits in the ability to acquire and use lan— guage (Soenksen et al., 1981; Spekman, 1981; U.S. Office of Education, 1977), it is not surprising that these children are often described as having poor interpersonal 19 relationships (Wiig 8 Semel, 1980). In an attempt to study learning-disabled youngsters' social perceptiveness, Bruininks (1978) investigated their perceptions of their self-concepts and social status among their peer group. The investigation also explored their interpersonal needs and friendship preferences. The self— and peer—ratings of 23 mainstreamed, elemen- tary school-age, learning-disabled children were compared to the ratings of 23 normal-learning classmates. Measures of perceived and actual social status were completed by use of the Peer Acceptance Scale (Bruininks, Rynders, 8 Gross, 1979). The Scale was adapted from Rucker's (1967) modifi- cation of the Ohio Social Acceptance Scale (Fordyce, Yauck, 8 Raths, 1996). This is a forced-choice scale on which __ peers within an identified group rate the degree to which they wish to be friends with other group members. Self- concept was measured by the Coopersmith Self-Esteem Inven- tory (Coopersmith, 1959). Interpersonal needs for the expression and desire for inclusion, control, and affection were determined by completion of the Fundamental Interper- sonal Relations Orientation-Behavior (Schutz, 1967). Results of the study showed that the learning-disabled students received significantly lower social status ratings than did their normal-learning peers. Furthermore, the learning—disabled students perceived their status as being much higher than it actually was; they predicted that their ratings closely resembled those of their normal-learning 20 peers. Normal-learning students did not demonstrate this discrepancy between perceived and actual status. There were no significant differences found in whom the two groups of subjects tended to choose as friends. Apparently, the learning-disabled children, like the normal-learning youngsters, preferred normal-learning friends rather than learning-disabled friends. Measures of personality and interpersonal needs showed that the learning-disabled sub- jects had significantly lower self-concept scores and sig- nificantly higher needs to express control. No differences were found between subject groups in the need to express inclusion and affection nor in the desire to receive con- trol, inclusion, and affection from others. Goldman and Hardin (1982) investigated the differences in social perception between 57 learning-disabled and normal-learning youngsters, ages nine to eleven years. The effect of sex on social perception, as well as the inter- action between the variables of sex and learning-disability, also were examined. Four social perception measures were administered: (1) The Children's Embedded Figures Test (Karp 8 Konstadt, 1971), (2) The Piers-Harris Self Concept Scale for Children (Piers, 1969), (3) The Nowicki-Strickland Internal-External Control Scale (Nowicki 8 Strickland, 1973), and (9) the Socialization Subtest of the California Psychological Inventory (Gough, 1975). The investigators found that the learning-disabled children were less accurate in their perceptions of social interactions than were 21 normal—learning youngsters. No significant differences were found between the abilities of male and female chil- dren in this skill and the sex of the learning-disabled child had no effect on how others perceived him. Instead, the determining variable was limited to the learning- disabled/normal-learning status of the child. In summary, school-age learning-disabled children consistently received lower social status ratings from classroom peers than did normal-learning children. Further- more, although learning—disabled students showed no differ- ences in selecting friends and in needing to be included and accepted by a group, they did demonstrate poor self- concepts and high needs to express control. Also, accu- rately perceiving their relative social status among peers has been demonstrated to be difficult for learning-disabled students. The children with learning disabilities tended . to perceive their status as being very similar to that of normal—learning students, when, in fact, it was signifi- cantly lower. Explanations for lowered social status ratings. A possible explanation for the lower social ratings of learning-disabled children is the stigma associated with being labeled as a learning-disabled or slow-learning stu- dent (Scranton 8 Ryckman, 1979). By investigating the sociometric status of children within an open-classroom school setting, Scranton and Ryckman (1979) minimized the effects of such labeling. The "open concept" (p. 50) was 22 such that students frequently moved from one room to another and did not necessarily know the destination of their class- mates. Thus, those students receiving special education services were not readily apparent. Scranton and Ryckman (1979) asked 92 first-, second-, and third-grade students to name a maximum of three class- mates in response to each of six questions. Three ques- tions pertained to positive characteristics, and three pertained to negative characteristics. Significant results, again were obtained between the learning-disabled and normal-learning students. This difference was established primarily by the contrast between the superiority of the control females on items referring to positive characteris- tics and the frequent identification of the learning- _‘ disabled females on the negative items. Since the effects of labeling were significantly reduced by the type of school program in which the students were enrolled, the authors concluded that peer status must be due at least partially to factors other than labels applied by educators. However, in spite of the absence of labels, it is possible that students had observed teacher-student interactions that suggested that the children with learning disabilities were in fact different in some way. To identify specific variables that may contribute to poor peer interactions, Bryan and her colleagues (Bryan, 1979a, 1979b, 1976, 1978; Bryan, Wheeler, Felcan, 8 Henek, 1976) completed a series of studies investigating the verbal 23 and social skills of elementary school-age learning- disabled children. Among the issues addressed in their research were the following: (1) ‘measurement of attitudes and behaviors demon- strated by teachers and peers towards learning— disabled children and the behaviors that these children direct back to them, (2) determination of specific behaviors that alert people to the existence of a problem and allow for the differentiation of learning-disabled and normal—learning children, and (3) determination of learning-disabled children's comprehension and use of verbal and nonverbal communication modes, as well as the complexity_- and patterns of their communication interactions with peers. Measurement of teachers' and peers' attitudes and behaviors towards learning-disabled children in a classroom setting were made through an observational study using a time sampling technique (Bryan, 1979b). The subjects were five learning-disabled and five average-achieving third- grade boys. Four major categories of behavior were coded during the observation periods: Task-Oriented Behavior, Non-Task-Oriented Behavior, Social Interactions, and Wait- ing. Significant differences were found between the amount of time the two groups of students spent in Task-Oriented 29 and Non-Task-Oriented Behaviors. The learning-disabled children spent significantly less time in Task-Oriented Behavior and significantly more time in Non-Task-Oriented Behavior than did the control group students. The results of the social interaction analyses showed that the learning- disabled and normalelearning children spent approximately the same amount of time interacting with teachers and peers, but the interaction periods were qualitatively different between the two groups of children. Specifically, class- room teachers gave equal amounts of positive reinforcement to all of the children but twice as much negative reinforce— ment to the learning-disabled students. Also, the children with learning-disabilities were twice as likely as their normal-learning peers to be ignored by both teachers and_‘ classmates when they initiated social interaction. In Bryan's next two studies (1979a; 1976), measurements of peers' attitudes towards learning—disabled children in classroom settings were made through the administration of sociometric scales. Included in the measures were the following: (1) the choice of three classmates as friends, class- room neighbors, and invitees to a birthday party, (2) the choice of three classmates who were not desirable in the three categories mentioned above, and (3) the Guess Who Technique (Garry, 1963), in which students select peers to correspond with various 25 descriptions. The subjects selected for the 1979 study were 89 learning- disabled third-, fourth-, and fifth-grade students who were matched with an equal number of normal-learning children on the variables of sex, race, and classroom assignment. A total of 62 classrooms, including 1,930 children, parti- cipated in completing the sociometric scales. The results revealed that the learning-disabled children were more likely to receive higher social rejection and lower social attraction scores than their normal-learning peers. Class- mates of learning-disabled students described them as undesirable playmates because of such traits as being "unhappy," "scared," and "worried" (p. 623). White, learning-disabled females were particularly likely to be__ associated with these characteristics. The 1976 study was a replication of the earlier research and used 25 of the original 89 learning-disabled subjects. These students were now in fourth- and fifth- grade classrooms. In addition to confirming the results obtained in the previous study, the 1976 results documented that the social rejection of learning-disabled children was maintained over time. Furthermore, the same rejection pattern occurred despite the fact that the majority of the normal-learning raters were different from the 1979 study. To determine whether learning-disabled children were being rejected by their peers because of differences in their communication patterns, Bryan et a1. (1976) analyzed- 26 the functional content of the utterances of 17 learning- disabled third-, fourth-, and fifth-grade children while they interacted with both learning-disabled and normal- learning peers. Eight categories of communication were established, including rejection statements, requests for information, self-image statements, cooperation statements, competition statements, helping statements, consideration statements, and intrusiveness statements. These categories were developed by the authors based on previous social interaction studies (Bryan, 1979a, 1979b, 1976; Hartup, 1970; Hartup, Glazer, 8 Charlesworth, 1967) and on Piaget's (1955) observations of young children's language develop- ment. Utterances which could not be reliably classified within these categories were eliminated from the analysis. Considerable differences were noted between the two subject groups in the frequency with which utterances occurred in the various statement categories. The analysis showed that significantly more competitive statements and significantly fewer consideration statements were made by the learning-disabled students. Also, the learning- disabled children uttered and received a greater number of rejection statements than did their normal-learning peers. Bryan (1978) next explored the behaviors and attitudes expressed by 89 fourth- and fifth-grade learning-disabled children to younger, normal-learning peers during dyadic interactions. Second-grade students, matched with the learning-disabled subjects on the variables of sex and race, 27 were used as the normal-learning subjects. The dyadic interaction task involved having the learning-disabled sub- ject teach a bowling game to the normal-learning youngster while the experimenter left to run an errand. Previous research has indicated that there is a rela-. tionship between children's popularity and social approval and their use of positive and negative reinforcement with peers (Hartup, Glazer, 8 Charlesworth, 1967). Bryan's study revealed, however, that there were no significant differences in the way reinforcement was used by learning- disabled and normal-learning students. Neither were there significant differences in any of the other behaviors investigated. Interestingly, there was a difference noted in the number of task-related statements made by the differ- ent groups formed according to sociometric status. The salient and rejected students made significantly more task-related statements than the popular and isolated groups. This result differed from expectations since it was thought that task-related statements would be appreci- ated and viewed as helpful behavior. Further analysis indicated that this behavior may have been viewed by peers as "meddlesome and intrusive" (p. 62) and Bryan (1978) suggested that the helping behavior may have been an attempt on the part of the learning-disabled child to repair an image damaged by social and academic failure in the class- room. Aside from identifying the possible adverse effect of producing too many task-related statements, the results 28 did not disclose any behaVioral differences which would help in explaining the sociometric rejection of learning- disabled children. Therefore, Bryan (1978) expanded her analysis in an attempt to identify specific behaviors that observers may use to differentiate learning-disabled children from normal- learning children. College students were used as judges who observed and rated the learning-disabled and normal- learning students under one of three different conditions. The judges had no knowledge of the nature or goals of the study. Condition one consisted of one-and-one-half to four minute videotapes of the bowling game interactions. Condi- tion two consisted of written transcripts only. In condi- tion three, the judges listened to the audio portion of ._ the videotapes. The judges rated eleven of the learning- disabled and eleven of the normal-learning children in all three of the conditions. Seven-point scales were used for rating each child's physical appearance, speech and lan- guage, expected academic achievement, attractiveness to other children, thoughtfulness, and ability to express an idea. Under all three conditions, the learning-disabled subjects were rated significantly lower than the control children on all of the variables included in the analysis. Since the different judging conditions yielded the same results, Bryan (1978) concluded that the learning-disabled children demonstrated expressive language deficiencies which strongly contributed to their identification. 29 Summarizing the results of Bryan's series of studies, learning-disabled children were found to have difficulty in social development and, particularly, in forming positive interpersonal relationships with peers. Negative impres- sions of these children were formed very quickly by teach- ers, peers, and strangers. This was determined to be at least partially due to the learning-disabled child's less developed expressive language abilities. Thus, there is evidence that certain language differences noted in language-disabled children affect social interactions and development as well as academic skills. However, such find- ings have not been documented in adolescents and they have not been documented for adolescents who have been classified as "language-impaired." Furthermore, the specific nature_ of the pragmatic (i.e., functional) differences possibly contributing to the negative ratings of children has not been described. Therefore, it appears appropriate that fu- ture research focus on identifying and describing the prag- matic behaviors that may impact on the social status ratings of language-impaired adolescents. Statement of the Problem Existing pragmatic literature documents the importance of studying pragmatic behaviors in normal-language and language-impaired populations. The literature also implies a strong relationship between social interaction behaviors and communication behaviors. However, this relationship has not been documented in children specifically classified as 30 language-impaired. Furthermore, the literature still contains little in- formation about the communication behaviors of language- impaired adolescents. However, it is well documented that there is a need for more information and research on this population (Boyce 8 Larson, 1983). Thus far, few investi- gators have looked specifically at the crucial areas of pragmatics or social behaviors in language-impaired ado- lescents. Investigators have continued to ignore language- impaired adolescents in spite of evidence that younger language-impaired children are known to demonstrate prag- matic communication difficulties. This oversight is even more surprising when one considers that there is evidence that some language-based learning disabilities may be asso- ciated with broader problems involving both communication and social skills. The present study, therefore, was designed (1) to describe pragmatic behaviors of language-impaired ado- lescents in a dyad situation, (2) to determine whether language-impaired adolescents receive low social status rat- ings from normal-language peers, and (3) to determine whether language-impaired adolescents are less accurate than their normal-language peers in predicting their social status. Specifically, this study was designed to answer the following questions: (1) During a fifteen minute period, do dyads with one (2) (3) (9) (5) (6) (7) 31 languagevimpaired and one normal-language ado- lescent use more speech acts to complete a verbal instruction task than do dyads with two normal- ' language adolescents? Do dyads with one language-impaired and one normal-language adolescent produce more unsuc- cessful speech acts than do dyads with two normal-language adolescents? Do language-impaired speakers assume more or less responsibility than normal-language speakers for producing successful or unsuccessful speech acts in a dyad situation? During unsuccessful speech acts, do dyads with one language-impaired and one normal-language adoles- cent violate different pragmatic or conversational rules than do dyads with two normal-language adolescents? What underlying semantic, syntactic, or phono- logical factors contribute to language-impaired speakers' and normal-language speakers' pragmatic violations in unsuccessful speech acts? Do language-impaired speakers receive lower social status ratings from normal-language listeners than do normal-language speakers? Do language-impaired speakers predict their 32 social status ratings as accurately as do normal- language speakers? Method Subjects The subjects were 12 thirteen- and fourteen-year old seventh- and eighth-grade students. Four subjects were male and eight were female. Nine of the 12 subjects, three males and six females, were considered normal-language adoles- cents. Three of the 12 subjects, two females and one male, were considered language-impaired. Two of the language- impaired subjects were in the seventh grade. The one re- maining language-impaired subject and the nine normal— language subjects were in the eighth grade. All subjects were judged to be grossly equivalent in‘ educational environment. The criterion for equivalency was the education levels of parents. At least one parent must have received a high school diploma; no more than one parent can have earned a degree higher than a baccalaureate. This information was obtained through a Parent Questionnaire (Ap- pendix A) that was administered during a telephone inter- view. In all cases, the mothers supplied the questionnaire information. The education levels of the subjects' parents are listed in Appendix B. Groupings. The 12 subjects were paired into six dyads. The control group consisted of three dyads with two normal- language youngsters in each; two dyads were female and one was male. The experimental group consisted of three dyads 33 with one normal-language youngster and one language-impaired youngster in each; two dyads were female and one was male. The chronological ages of the subjects in each dyad were within three months (i) of one another in order to avoid possible social status effects associated with age differ- ences. The control dyads were labeled as Control-Male, Control-Female One, and Control—Female Two (C-M, C-Fl, C-F2). The experimental dyads were labeled as Experimental- Male, Experimental-Female One, and Experimental-Female Two (E-M, E-Fl, E-F2). To insure that the subjects' social ratings of one another were limited to the time spent together during the dyad interaction, it was necessary to establish that members of each dyad did not know one another prior to the research project. This was done by pairing together only those stu- dents who attended schools in different local districts within the Ingham Intermediate School District. Also, dur- ing a telephone interview, parents were asked to verify that their son/daughter had not had any previous contact with the adolescent with whom he/she was to be paired. In addition, when subjects arrived for the dyad session, they were asked if they knew one another. None of the paired subjects reported knowing one another prior to the dyad session. Selection and screening. All subject volunteers were students from Ingham Intermediate School District, Ingham County, of the Lansing Area of Michigan. In three cases, however, subjects were solicited from the investigator's 39 colleagues and friends in the Lansing area. After receiv- ing permission from school principals and classroom teach- ers, the investigator made announcements to eighth-grade English classrooms. The investigator briefly described the tasks involved in the study and distributed parent consent forms to interested students. School speech-language clinicians were asked to identi- fy thirteen- and fourteen-year old language—impaired stu- dents on their caseloads. Once language—impaired students were identified, the investigator met with each student individually. The tasks of the study were described and parent consent forms were given to interested students. After the parent consent forms were returned, volun- teers were scheduled for screening of language comprehen:m sion and production skills. The language screening results were used to identify language-impaired and normal-language subjects. The Peabody Picture Vocabulary Test-Revised (Form L) (PPVT-R) (Dunn, 1981) and four subtests of the Detroit Tests of Learning Aptitude (DTLA) (Baker 8 Leland, 1959) were included in the language screening and were administered to all subject volunteers individually. The testing was completed at each student's school or home. The PPVT-R (Dunn, 1981) was included in the language screening as a measure of single-word receptive vocabulary. The following four subtests of the DTLA (Baker 8 Leland, 1959) also were administered: the Social Adjustment A, Social Adjustment B, Auditory Attention Span for Related 35 Syllables, and Oral Directions Subtests. According to the test manual, the two Social Adjustment Subtests measure com- prehension and reasoning skills. They were included in this screening to provide a sample of the adolescents' language formulation skills. The Auditory Attention Span for Related Syllables Subtest is a sentence imitation task and the Oral Directions Subtest involves following increasingly longer directions. Both subtests rely heavily on auditory percep- tion, comprehension, and memory. In addition, visual attentive abilities, motor ability and precision, and practical judgment are reportedly measured by the Oral Di- rections Subtest. "Language-impaired" was defined as language comprehen- sion and/or production age-equivalency scores that were _ two or more years below chronological age on the PPVT-R and on at least two subtests of the QTLA. "Normal-language" was defined as an age-equivalency score within two years of chronological age on the PPVT-R and on at least three of the four DTLA subtests. The chronological ages, sexes, raw scores, and age-equivalency scores for the 12 subjects are listed in Appendix C. Providing parental consent was obtained, the school record of each subject was reviewed by the investigator to confirm that cognitive/intellectual abilities and hearing sensitivity were within normal limits. Reports of previ- ously administered intelligence tests, achievement tests, and grades were used to make judgments about normal 36 cognitive abilities. When this information was not made available, parent and teacher reports of age-appropriate classroom performance were used as the criteria for normal cognitive/intellectual abilities. Volunteers with suspected or confirmed intellectual deficits or hearing impairments were excluded from the study. A total of 17 students volunteered to participate in the study, but five volunteers were excluded from the actual study because of subject selection criteria. One female was excluded because achievement test scores and classroom grade reports suggested that intellectual abilities were impaired; two females were excluded because of age differ- ences that were greater than three months from the other volunteers; one female was excluded because her racial background was different from the other volunteers; and one male was excluded because of reports in his school record indicating that he had difficulty interacting with teachers and peers. One female subject volunteered from a regular eighth grade classroom; however, her screening test perfor- mance classified her as language-impaired. She, therefore, was classified as language-impaired and placed in the experimental group (E-Fl). Materials Puzzle designs. Puzzle designs were used for the instruction-giving task of the study. The investigator con- structed ten puzzle designs that each consisted of eight pieces in three colors: two blue triangles, one red 37 triangle, and one yellow triangle; one red rectangle and one yellow rectangle; one red square and one yellow square. All colored shapes were made of foam rubber. The puzzle designs are presented in Appendix D. Equipment and setting. A Sony Betamax Video Recorder (#SLO-323), a Sony color camera (#HVC-2000), and Sony video- cassettes (#L-SOO) were used to video and audio record the fifteen-minute instruction-giving task of each dyad. The setting for the instruction-giving task was a small room containing two tables, a lectern, and one chair. The chair was placed in front of one of the tables with the second table and lectern located slightly behind and to the left of it. . Social status ratings. A Peer Rating Scale (Appendix E) was developed by the investigator for subjects to rate the social status of their dyad partners (e.g., "How much would you like to play on a team in gym with ?"). A Predicted Peer Rating Scale (Appendix E) was developed for subjects to indicate how they believed their dyad partners rated them in social status (e.g., "How much do you think would like to play on a team in gym with you?"). Each scale consisted of ten questions that were rated on a one-to-five-point scale. A rating of "one" represented the most negative reaction to participation in a future activ- ity with the dyad partner. A rating of "five" represented the most positive reaction to further participation with the dyad partner in the specified activity. Drawings of 38 facial expressions with captions were used to represent the range of points on the five-point scale (Appendix F). A total score of one to five for each scale was obtained by averaging the student's rating responses to the ten ques- tions. Procedures All sessions were held at Michigan State University's Communication Arts and Sciences Building. Each pair of sub- jects was scheduled separately and seen for approximately 95 minutes. When the subjects arrived, the investigator introduced them to one another and confirmed that neither subject had met nor seen the other before their arrival at the session. To minimize the effects of dress-related social vari; ables during the dyad interactions, attempts were made to reduce individual dress variability. When subjects were contacted for scheduling, they were told to wear a tee- shirt and jeans to the dyad session. Therefore, when the subjects arrived for the dyad session, the investigator con- firmed that they were wearing the specified dress style before escorting them to the experimental room. Instruction-giving task. The subjects in each dyad were assigned to either a "teacher" or "student" role. The language-impaired subjects were always assigned the role of "teacher." In dyads comprised only of normal-language subjects, the "teacher" and "student" roles were randomly determined prior to the session. The subject assigned the 39 role of "teacher" was told to stand behind the lectern where he could view the other subject's work area. The "teacher" always was told to keep his hands inside the lec- tern to insure that he did not use his hands to explain the task. The subject assigned the role of "student" was seated at the table in front of the lecturn and with his back to the "teacher." The instruction-giving task and rules were explained to both subjects simultaneously. The instructions that were given to the subjects are listed in Appendix G. Ten puzzle designs (Appendix D), each drawn on a separate sheet of paper, and one set of puzzle shapes were located on top of the lectern. A second set of shapes was placed on the other table. After the investigator an- swered all of the subjects' questions, the pair was instructed to begin work on the instruction-giving puzzle task. The investigator remained in the room to videotape the interaction; however, subjects were not permitted to ask questions or seek clarification from her once the task began. The subjects were instructed to work for fifteen minutes. The instruction-giving puzzle task involved the sub- jects' participation in a modified over-the-shoulder activ- ity (Muma, 1978). This is a variation on the barrier activity, derived from suggested intervention procedures for developing skills in the dump and play operations of the communication game (Muma, 1975; Longhurst 8 Reichle, 90 1975). Originally, the barrier activity was presented by Glucksberg and Krauss (1967) as a research paradigm for investigating the acquisition and development of referen— tial communication in children. Specifically, the experi- mental task involved the "student" constructing abstract puzzle designs by relying on the verbal instructions of the "teacher" member of the dyad. Rules for completion of the task were specified accord- ing to the two roles portrayed. The "teachers" were expected to provide verbal instructions for the arrangement of the puzzle designs. They were allowed to observe the placement of shapes by the "student," thus encouraging repe- titions, modifications, and/or revisions of directions as felt necessary. However, the "teachers" were not to tough or point to any of the "students'" shapes. The "students" could not move their chair, stand up, or engage in any other movement that would allow them to view the "teachers'" design models. "Students," however, could manipulate their own set of shapes in following the "teachers'" directions, and they were encouraged to ask questions when confirmation, clarification, or further explanation was needed. A puzzle was considered to be com- pleted when the "teacher" judged the design constructed by the "student" to be identical to the original model. The' order of presentation of the designs was randomly pre- determined for each dyad. 91 Social status ratings. FolloWing completion of the instruction-giving task, the two subjects were separated and individually instructed in the completion of either the Peer Rating or Predicted Peer Rating Scale. "Student" sub- jects always completed the Peer Rating Scale (Appendix E) to rate the "teacher" subject who gave the instruction. "Teacher" subjects always completed the Predicted Peer Rat- ing Scale (Appendix E) to state how they believed their peer had rated them. The scales were administered on an individual basis with the questions read aloud by the inves- tigator. The subject was asked either to point to one of the drawings or to verbalize one of the drawing captions in response to each question. At the end of the rating session, the investigator __ answered questions the subjects had about the procedures; however, rating results were not shared with the subjects. Transcriptions and Data Analysis Procedures The videotaped dyad sessions were transcribed by hand on 19" x 17" pads from an RCA color television monitor (XL-100). All verbal utterances, relevant non-verbal behaviors, and contextual puzzle manipulation information were recorded in traditional orthography. An "utterance" was defined as any remark representing a complete unit of thought; this definition was based on Loban's (1976) con- cept of a communication unit. Hesitation behaviors were captured by noting pauses between words within an utter- ance. A "pause" was defined as silence lasting for more 92 than 3/9 of a second. Utterance functions. The pragmatic function or intended meaning (i.e., communicative intent) of each utter- ance was classified according to Dore's (1978) conversa— tional act (i.e., speech act) categories. For example, "Put the red triangle above the yellow square," was classi- fied as expressing an "Action Request" function or intent. "Will you say that again?" was classified as "Clarification Request." The utterance, "Right there," was classified as "Description of Location." Dore's (1978) classification categories for conversational acts and examples for each classification are listed in Appendix H. Communication acts. Once each utterance function was identified, it was evident that several speech acts or con- versational acts, involving both participants, were usually included in the "teacher's" and student's" attempts to satisfy a single communicative intent or goal expressed by the "teacher." The achievement of a "teacher's" communica- tive intent was evidenced by the fulfillment of the essen- tial point condition of the speech acts (Chapman, 1981; Lucas, 1980; Searle, 1969). The essential point condition requires that the listener make some attempt to carry out the speaker's intent. In this study, the essential point condition was met when the "student" attempted to select the puzzle piece intended by the "teacher" and/or when the "student" attempted to move the puzzle piece to the location intended by the "teacher." Thus, several speech acts could 93 occur by the "teacher" and "student" before the essential point condition or the communicative goal was met. For example, three utterance exchanges or speech acts occur before the "teacher's" communicative intent is satis- fied in the following exchange: Teacher: Now take the yellow triangle and put it on the blue one. Student: To make a square? Teacher: So it's on top but not to make a square. The communicative intent was expressed in line one but the essential point condition of the "teacher's" communicative intent was not satisfied until line three, when the "stu- dent" finally complied with the "teacher's" intended com- mand (i.e., request for action). Such an exchange was viewed by the investigator as one communication act. A variety of speech acts, including "Action Request, "Yes/No Request," and "Clarification Response" were performed for the completion of this communication act. Therefore, the minimal unit for conversational analy- sis of communicative intent or language functions during the dyad interactions was described as a "communication act." A "communication act" was defined as a speaker's verbal or nonverbal expression of a communicative intent and the listener's verbal or nonverbal response attempt to satisfy the initial communicative intent or goal of the speaker and, thus, the essential point condition of the speech acts. Specifically, an act was not complete until 99 the "student-listener" attempted to move a puzzle piece to the position he believed the "teacher” intended. That attempt served to fulfill the essential point condition and mark the completion of the act. The introduction of another intended puzzle piece and/or another intended location for a puzzle piece served as the beginning of a new communica- tion act. In summary, a communication act consisted of the speech acts and/or relevant nonverbal behaviors that were produced to achieve a single communicative intent regarding a specific puzzle piece and its location-—that is, the accomplishment of the essential point of the "teacher's" communication goal. Each separate communication act was identified and marked on the transcript. The total number of communica;_ tion acts completed by each dyad was tallied. In addition, each communication act was classified as either "success- ful" or "unsuccessful" in satisfying the speaker's intended goal, and,therefore, the essential point condition of the communication act. A communication act was classified as "successful" when the essential point condition was met. In other words, once the "student-listener" successfully complied with the "teacher's" initial intent regarding placement of a puzzle piece, the act was judged as "successful." For example, in the following exchange, the "student" successfully placed a puzzle piece in the design: Teacher: Turn the blue triangle so the bottom is on 95 the red rectangle. Student: Okay, the bottom is towards it. Teacher: Yeah. Since the "student" correctly placed the blue triangle, the act was classified as "successful" in meeting the essential point condition of the "teacher's" intended meaning. A communication act was classified as "unsuccessful" when the "student" made an attempt to comply with the "teacher-speaker's" intent but failed to select the intended puzzle piece or failed to place the piece in the intended position. The following exchange is an example of an "unsuccessful" communication act: Teacher: Right below the red square, put the red rectangle. _‘ Student: Okay. In this example, the "student" placed the red rectangle per- pendicular to the red square rather than parallel to it, as was intended by the "teacher." In essence, the "stu- dent" did not fulfill the essential point condition of the speaker's intended meaning and, therefore, the act was considered "unsuccessful." Four descriptive procedures were completed to explain why the essential point condition was met in successful communication acts but not met in unsuccessful communication acts. Maxims. The first procedure completed to explain the success or failure of the communication acts consisted of a 96 pragmatic description of the "teacher" utterances only. Since this study was designed to focus on the essential point conditions associated with the communicative intents expressed by the subject in the instruction-giving "teach- er" role, the maxim violation classifications were limited to the utterances of the "teachers." All "teacher" utter- ances for each communication act were checked for either appropriate use or violation of Grice's (1975) four prag- matic maxims for effective, cooperative conversation. In all cases, only one of the four maxim categories was used to explain the success or failure of a single communication act. Specifically, each communication act's classification as successful or unsuccessful was related to the "teacher- speaker's" appropriate use or violation of either the Quantity, Manner (Clarity), Quality (Accuracy), or Relation (Topic) maxim (Grice, 1975). Consistent with Grice (1975), a "Quantity" violation was defined as insufficient information from the "teacher— speaker" for the "student-listener" to realize the "teacher's" intended meaning. For example, in the utter- ance, "Put the yellow triangle on the side of the red square," the "teacher" does not tell the "student" on which side of the red square to place the yellow triangle. Thus, the error was coded as a "Quantity" violation because the "teacher" failed to supply enough information for the "stu- dent" to carry out the request. A "Clarity" (Manner) violation was defined as ambiguous 97 information that resulted in the "student-listener" being unclear or confused about the "teacher-speaker's" intended meaning (Grice, 1975). The "student's" response, therefore, had to indicate that for some reason the "teacher's" intent was not clear and, thus, the "student" was confused about what the "teacher" expected from him. For example, the "teacher's" utterance, "The straight piece should be on the bottom," provided the "student" with ambiguous information concerning the placement of a triangle in the puzzle design since all three sides of a triangle are straight. In this example, the "student's" confusion was evident by her re- quest for confirmation of her actions (i.e., "Yes/No Ques- tionz" "You mean like this?"). The unsuccessful communi- cation act, therefore, was coded as a "Clarity" violation: An "Accuracy" (Quality) violation, consisted of inac- curate information (Grice, 1975) that resulted in the "student-listener" responding in a way that was not intended by the "teacher-speaker." "Accuracy" violations were differentiated from "Clarity" (Manner) violations by examining the truthfulness of the utterance. Thus, if the utterance contained accurate (truthful) information but was stated in such a way that there could be more than one valid interpretation, the violation was coded as "Clarity." If the "teacher's" utterance contained false or inaccurate information, the violation was coded as "Accuracy." For example, "Put the yellow square to the right of the red square," was coded as an "Accuracy" violation since the 98 "teacher" intended for the yellow square to be placed to the leg: of the red square. A "Topic" (Relation) violation was defined as a "teacher-speaker's" failure to maintain the topic of the puzzle design task. For example, a "Topic" violation would have occurred if a "teacher" had said, "Put the yellow square next to the red square. Did you see the football game yesterday?" The second sentence concerning the foot- ball game is unrelated and irrelevant to the topic of the puzzle design task; therefore, the "teacher" would have been coded for a "Topic" violation and a new communication act would have begun when the "student" or "teacher" returned to the topic of the puzzle design task. Listener Behavior. The second descriptive procedure_ used to explain the success or failure of the communication acts involved only "student" behaviors. All "student" verbal responses, verbal comments, and puzzle manipulation responses were classified as either appropriate or inappro— priate pragmatic "Listener Behaviors." This category was included in an attempt to account for at least some of the "students'" behavior and responsibility in the dyad inter- actions. "Listener Behavior" was considered appropriate when the "student" followed the directions provided by the "teacher," regardless of the accuracy, clarity, and/or thoroughness of the "teacher's" instructions. For example, following the "teacher's" utterance, "Put the blue triangle 1+9 underneath the yellow square," the "student" placed the blue triangle below the yellow square. This was appropriate "Listener Behavior" since it complied with the "teacher's" instructions. ”Listener Behavior" was considered inapproprate when the "student" did not do as he was instructed by the "teach— er." For example, following the "teacher's" utterance, "Put the blue triangle on the left side of the red square," the "student" placed the blue triangle to the left of the red triangle. Thus, the "Listener Behavior" was considered inappropriate because the "student" failed to follow the "teacher's" instructions. Non-pragmatic factors. The pragmatic explanation for each communication act's success or failure was either ope of the four maxim categories or the "Listener Behavior" Category. The third descriptive procedure was used to offer additional non-pragmatic explanations for the failure of the unsuccessful communication acts. This was done by coding every unsuccessful communication act as either semantic, syntactic, or phonological in nature. In other words, these components were considered underlying factors that contributed to the pragmatic violations and, thus, the unsuccessful communication acts. "Semantic" factors were subdivided into two categories based on Lucas (1980). The first subcategory was "Semantic Content," and it was defined as ambiguous word choices. For example, the word "by" in the utterance, "Put the yellow 50 square by the red triangle," does not provide the listener with clear instructions concerning the placement of the yellow square with respect to the red triangle. This was considered an ambiguous word choice and, thus, a "Semantic Content" error. The second type of "Semantic" error involved "Vocabu- lary.” A "Vocabulary" error consisted of an incorrect word choice. For example, the incorrect lableing of "red triangle" as "red rectangle" in the utterance, "Put the 33d rectangle above the yellow rectangle," was judged to be a "Vocabulary" error. Thus, "incorrect" was differentiated from "ambiguous" word choice on the basis of the accuracy of the information. An underlying "Syntactic" error was defined by Luca§_ (1980) as "an unconventional ordering of the lexicon (vocabulary)" (p. 58). Thus, "Syntactic" errors involve the grammatical word order of utterances. When word order is altered, the listener will often have difficulty deter- mining the speaker's intended meaning and will have to rely heavily on contextual cues to interpret the speaker's utterance. Underlying "Phonological" errors were also defined on the basis of concepts presented by Lucas (1980). An error was considered phonological when a phoneme or syllable omission resulted in the omission of a grammatical morpheme or functor word or of a grammatical morpheme ending. For example, the omission of a phoneme marker for the 51 grammatical morpheme "is" would be considered a functor word or grammatical morpheme omission. Production of "move" for "moved" would be considered an omission of a grammatical morpheme ending. In other words, phoneme and syllable omissions were considered phonological when they were "morphophonemic in nature" (Lucas, 1980, p. 70). "Teacher-student" responsibility. The final descrip- tive procedure used to explain the success or failure of the communication acts consisted of assigning either the "teacher" or the "student" as responsible for the success or failure of the act. The repsonsibility of the "student" or "teacher" was always assigned on the basis of "Listener Behavior" and maxim behaviors, respectively. The responsi- bility for the success of a communication act was assigned to the "teacher" when he appropriately used the four con- versational maxims. Conversely, responsibility for the failure of a communication act was assigned to the "teach- er" when he violated one of the four conversational maxims. The responsibility for a successful act was assigned to the "student" when the "teacher" violated one of the maxims but the "student" was able to correctly guess the intended puzzle piece or location. Thus, the "student" satisfied the essential point condition in spite of the "teacher's" maxim violation. The responsibility for an unsuccessful act was assigned to the "student" when the "teacher" appropriately used the maxims but the "student" failed to comply with the "teacher's" instructions and, 52 thus, failed to fulfill the essential point condition of the act. Therefore, for all communication acts, "teacher" responsibility classifications always were associated with "teacher" violation of maxims, and "student" responsibility classifications always were associated with "student" "Listener Behaviors." For example, "teacher" responsibility for a successful communication act occurred in the following example: Teacher: Put the red square on the right hand side of the yellow square. The "teacher" supplied adequate and accurate information for the "student" to correctly position the red square in the puzzle design. If the "student" had selected a puzzle piece other than the red square or placed it in a location other than to the right of the yellow square, the communi- cation act would have been unsuccessful and the "student" would have been assigned the responsibility for the commu- nication act failure. "Teacher" responsibility for an unsuccessful communication act occurred in the following example: Teacher: Below the red square, put the blue triangle. In this example, the "teacher" violated the "Quantity" maxim since she failed to specify that the base of the triangle needed to be placed along the bottom edge of the red square. Thus, she supplied an insufficient quantity of information and the "student" was not able to place the blue triangle in the correct position. If the "student" 53 had guessed the correct triangle placement, the communica- tion act would have been successful and the "student" would have been assigned the responsibility for the commu- nication act success. To summarize, all communication acts received either a pragmatic maxim classification or a "Listener Behavior" classification as well as either a "student" responsibility or a "teacher" responsibility classification. Consequently, all "teacher" responsibility classifications reflect use of the maxims, while all "student" responsibility classifica- tions reflect "Listener Behavior" codings. Only the prag- matic "teacher" violations of the unsuccessful communication acts were described by the underlying semantic incorrect word choice, semantic ambiguous word choice, syntactic, and phonological factors. Inter-judge reliability. To determine inter-judge reliability of the coding systems, a second judge was trained in their use. Working independently from the pre- pared transcripts, she analyzed one-third (five minutes) of each dyad's interaction on all of the variables with the exception of the "teacher-student" responsibility for successful communication acts. Reliability for "teacher- student" responsibility for successful communication acts was determined by the second judge's ratings of all of dyad E-M's and C-F2's successful communication acts. Her results were compared with those of the investigator. Inter-judge reliability for the seven coding systems ranged from .87 59 to .99, with an overall average of .93. The results of the overall percent of agreement for each coding system across dyads and for each individual dyad are listed in Appendix I. Results and Discussion The three major goals of this study were (1) to describe pragmatic behaviors of language-impaired ado- lescents in a dyad situation, (2) to determine whether language-impaired adolescents receive low social status ratings from normal—language peers, and (3) to determine whether language-impaired adolescents are less accurate than their normal-language peers in predicting their social status. Pragmatic Behaviors: Communication Act Descriptions .. Number and nature of communication acts. Initially, the investigator expected that the experimental group would use more communication acts than the control group to com- plete the puzzle design task. It was hypothesized that the language impairments of the experimental "teachers" would contribute to difficulties in producing the verbal instruc- tions necessary for completing the puzzles. Consequently, the language-impaired "teachers" would need to revise more messages than would the control "teachers" to accomplish the task. More communication acts, therefore, would be used by the experimental "teachers" to give effective verbal instructions. Interestingly, the opposite occurred; the control group 55 produced slightly more communication acts than did the ex- perimental group (909:356). The number of communication acts ranged from 118 to 198 for the control group and from 113 to 128 for the experimental group. Although the differ- ence between the total number of communication acts pro- duced was only 98, this is a large enough number to indi- cate that the control group had a tendency to use slightly more communication acts than did the experimental group. It was believed that part of the explanation for the differences noted between the control and experimental groups' frequency of communication act use might involve the amount of work that the two groups accomplished during the interaction sessions. However, both the control and ex- perimental groups completed the same number of puzzle designs--that is, both completed 19 puzzles plus three ’pieces of the 20th puzzle for a total of 155 puzzle pieces. The ranges of the groups' completed puzzles also were remarkably similar. The control group arranged 97 to 61 puzzle pieces, and the experimental group arranged 96 to 61 puzzle pieces. A small difference did occur in the total number of puzzle design errors made by the two groups. In the control group, two of the 19 completed puzzles contained errors. Five of the 19 puzzles completed by the experimental group contained errors. While the difference suggests a trend for the experimental group to make more puzzle errors than the control group, the total difference is only three puzzles 56 and, therefore, should not be considered substantial. Over- all, the results of the number of communication acts pro- duced, of puzzles completed, and of puzzle design errors made suggest that the verbal instructions and nonverbal behaviors of the two groups' communication acts did not differ enough to substantially affect their performances on the puzzle task. Since task performance outcomes did not differ enough to explain the control group's production of more communi- cation acts than were produced by the experimental group, an attempt was made to identify possible qualitative differ- ences in the nature of the communication acts themselves. The possibility of qualitative differences was explored by describing the utterance function categories and converse: tional acts of the communication acts used (1) by the con- trol and experimental groups, (2) by the "teacher" subjects within the control and experimental groups, and (3) by the "student" subjects within the control and experimental groups. The control group used 152 more conversational acts to carry out their communication acts than did the experimental group. The control group produced a total of 776 conversa- tional acts while the experimental group produced 629. The average number of conversational acts expressed in a single communication act, however, was similar for the two groups. The control and experimental groups expressed an average of 1.92 and 1.75 conversational acts, respectively, for each 57 communication act. Also similar were the utterance function categories expressed in the communication acts of the control and ex- perimental groups. Both groups used the "Response" and "Request" categories most frequently. Forty-seven percent of the control group's and 96% of the experimental group's utterance functions were "Responses." "Requests" accounted for 30% and 36% of the control and experimental groups' utterance functions, respectively. "Organizational Device" and "Miscellaneous" categories were produced less frequently than the "Response" and "Re- quest" categories. The "Organizational Device" category accounted for 11% of the control group's functions and 10% of the experimental group's functions. The "Miscellaneous" category accounted for 6% of the utterance functions of both the control and experimental groups. The remaining four categories of "Descriptions," "Statements," "Performa- tives," and "Acknowledgments" each accounted for less than 9% of the utterance functions used by the control and ex- perimental groups. Percentages of the control and experi- mental groups' uses of each of the utterance function categories are graphically represented in Figure 1. The numbers associated with each utterance function category are listed in Appendix J. The specific types of conversational acts expressed within each utterance function category also were similar for the control and experimental groups. The most frequent 58 .mmsosw Hopaofimnomxo Ufim H9980 map \E couscous mofihomopmo £3555. oofifiofib mo omopcoohom .H am macaw Hmpcflfisoaxm m macaw Hoflboog Knox woof/om Eomowwo .3929» .Enomnom 35:68pm dwaomon .HHoomE . nowpo mpmoscom momcommom cowuogm o m CH ma om mm afieiuearad om mm or m: om 59 "Request" type for both groups was an "Action Request" (e.g., "Put the yellow square in front of you."). "Action Requests" accounted for 69% of the "Request" functions of the contrOl group and for 65% of those of the experimental group. "Yes/No Requests" (e.g., "Like that?") were the next most frequent "Request" type for both the control group (28%) and the experimental group (31%). The percen- tages of the various request types within the "Request" category are shown in Figure 2. The number of conversa- tional acts which occurred within each utterance function category are listed in Appendix J. The conversational acts of "Clarification Responses" (e.g., "Move it up.") and "Yes/No Responses” (e.g., "Yeah, like that.") were distributed relatively equally for the-” control and experimental groups. This similarity is shown in Figure 2. Fifty percent of the control group's and 96% of the experimental group's "Response" functions were "Clarification Responses." Forty-one percent of both the control and experimental groups' "Response" functions were "Yes/No Responses." Thus, the combined conversational act categories of "Clarification Responses" and "Yes/No Responses" accounted for almost all of the "Response" utterance functions which occurred in both the control and the experimental dyads. The similarity in function patterns for the two groups also is apparent in the other 35 conversational act types. In both groups, 15 of Dore's (1978) conversational acts did 60 L .mcoapocsm >nomono aomcommom= cam zumosvom= m.m:onw Hmvco8HQoaxo cam Hochoo may cease: mpom HMCOflpmmnw>coo mo mmMpcwonwm .m madman QSOQw HopcuEflnwmxm mmmw .Qsomm Hoanooummm "zox momcommom memoswom 88.23 .8ng A98 058» .388 .8315 MEMO 05mm» 833. 8.9. OH om om 0: cm afiexueoaed .om on om om OOH 61 not occur: Disapproval/Disagreement Acknowledgment, Return Acknowledgment; Event Description, Time Description; No Answer-Miscellaneous; Accompaniment Device, Calling Device, 'Speaker Selection Device; Claim Performative, Teasing Per- formative, Warning Performative; Qualification Response; Permission Request, Rhetorical Question Request; and Attribution Statement. Of the 20 remaining conversational acts that did occur, "Uninterpretable" or incomplete sen- tences were the most common act within the "Miscellaneous" category for both groups. "Boundary Markers" accounted for the majority of the "Organizational Devices" used by both groups. These and other similarities, as well as a few minor differences, are represented by percentages for the 29 occurring conversational acts shown in Figure 3, Figure 9, and Figure 5. The normal-language and language-impaired "teachers'" utterance function categories and conversational acts were explored as the second way of attempting to identify quali- tative differences in the nature of the communication acts of the control and experimental groups. The utterance function categories used by the "teachers" were separated from those of the "students." Consistent with the "teacher- speaker" design of the study, the vast majority of the utterance function categories were used by the "teachers" in both the control (651 / 776) and the experimental groups (988 / 629). As was the case in comparisons of the control and experimental groups' total number of communication acts 62 .maowvocsm >nowouoo :msoocmHHoomfiz: cam :oofl>mm Hmcowwmwflcmmno: m.asonw Hopcmewnmmxm cam Honpcoo mnp aflnpwz wpom choflpmwnw>coo mo ommucmOme .m mcdwfim anonw Hopcmsflnmaxm mmmw agonw HOQHCOOfimmw "hex msoocmaaoomflz mooa>mm Hmcowemwmcmwno coaumamaoxm .ponmwn#CflcD mwocopflaom mnmocdom po< 0H om om o: om om ~ afielueoaed on om om OOH 63 .mcoavocsm aowmpmo acoflmfinomona mam :vcgapmfim: m. asonw fiasco—Ego EB HoHEoo ME, canvas mpom Hmcoflmmhoauoo mo oprcoQHom .: warm 919nm Hmpcofimhomxm m 9.5% Hoficoog Sex meowpawcaomon mpgpmvm c3983 hag .oEchoUH soapmfimamxm mam cowwmsamSm Pygmy“ .LhH po< afieiueoaed 69 mpcméwooazgaolw .893 .89 monopmmooee .mcoa “EEG ?omopoo ..#5:&U0H39b_o<: can zo>flmfiompoma m .msoam Hopgnmmxm one HofiEoo was c.2393 meow HmcoapMma>coo mo owmpgm .m madman 939% 32.958me m we?” ugompom moxoh. Hwopoam e3. OH om cm 0: om ow on om om ooa aBmuezued 969m HoHEoo @ "Knox 65 and total number of conversational acts, the control "teachers" used more conversational acts than did the ex- perimental "teachers" (651:988). Thus, the language- impaired "teachers" clearly expressed fewer conversational act functions than the normal-language "teachers." In spite of this difference, the average number of conversa- tional acts produced in each communication act was similar for the control "teachers" (1.61) and the experimental "teachers" (1.37). 1 Another similarity between the control and experi- mental "teachers" was the frequency with which they used the utterance function categories. Figure 6 shows the per- centages of "teacher" and "student" contributions to the total percentage of utterance function category occurrences for the control and experimental groups. The percentages of "teachers'" utterances classified in the various acts of the "Request" and "Response" categories and of the remain- ing "teachers'" utterances classified in the category of "Other" are shown in Figure 7. The "teachers'" conversa- tional act composition of the "Request" and "Response" categories is shown by percentages in Figure 8. The types and numbers of conversational acts within each utterance function category for all "teachers"-are listed in Appendix K. The graphs show that both the control and experimental "teachers" were most likely to use either "Responses" or "Requests" (Figure 6). When the "teachers" produced a 66 .3828 .>nomopmo cospoczm zoom now :mwcopSPm: can :mnwnomop: Hmpcmsflnomxm cam Hoppcoo ma ooosconm mmocmnmppz mo omMHcooamm .m onswwm 99:839. Sea a 98688... .emafimm .8899? .ESB 38880:: 98% ">91 muowopmo mo0fl>on mpmosvom momcoamom cospocsm .HHoomflz .cmwno .Euomnmm mpcoempMPm .anomom eSequeoaed 67 .aowmpmo .1059. commmfioo mfi. CH one moflmowovmo :omcommom: can .Lmosqoma o5. mo pom dmcoapmmno>coo zoom 5” 805.83: :Lmnooop: flaw—fig Em HQHEoo mo owopcooomm .N. a: :mpogomop: Hopcofiaoefl m :mnwgomop: HQHEOQE Sax momdommmm mpmmsvwm 88.0 .8523 .9918 .emamm 058» News News 058.» .mmaoué 83% Ba. 9821118 acted 68 .mofisomopmo :oflogm :omcoamom: _ one .Lmosvom: : .mpoaomop: Hopcofinomxo qu. HofiEoo may FEM: mubm Hmcoflumngeoo mo owmmEooHom .m mama :wgoaomop: Hopcgflfiog m :mponomow: HoHEoo a Knox momcommom memosvom .9330 .8812: 81.3.3 was 35.5» 49:8 #88 029% 5.88 058w ahuaared 69 "Response," it was likely to be either a "Clarification Response" or a "Yes/No Response" (Figure 8). When the con- trol or experimental "teachers" produced a "Request," it was almost always an "Action Request" (Figure 8). The one exception to the "Action Request" pattern was the "teach- er" of E-M, who used one "Wh-Request." Since no substantial differences were evident in the "teachers'" utterance func- tions and conversational acts, "teacher" functions alone do not account for the control group's production of more communication acts than the experimental group's to complete the same number of puzzle designs. The "student" behaviors within the control and experi- mental groups also do not account for the larger number of communiCation acts produced by the control group. This is true in spite of a few small differences between the con- trol and experimental "students." The "students" of the experimental group produced 11 more utterance functions than did the control "students" (136:125). The experimental "students" also produced an average of .6 or approximately one-half more utterance functions per communication act than did the control "students" (2.62:3.23). The fact that these differences are so slight indicates that the general trends of the control and experimental "students" were basically the same. The "students'" similarity can also be seen in examin- ing the specific conversational acts which they tended to use. The percentages of "students'" utterances classified 70 in the various acts of the "Request" and "Response" cate- gories and the percentage of remaining "student" utterances classified in the collapsed category of "Other" are graphically represented in Figure 9. The "students'" con- versational act composition of the "Request" and "Response" categories is shown by percentages in Figure 10. In Figures 9 and 10, it can be seen that both groups of "stu- dents" tended to use either a "Yes/No Request" (e.g., "Like this?") or a "Compliance Response" (e.g., "Okay.") most frequently. The repeated use of these conversational act types by the "students" appears to be a reflection of the "teacher-speaker" design of the task. Although the overall patterns are very similar, one small, but seemingly interesting, difference was noted in the ratio of the experimental and control "studentS'" total "Clarification Requests" (e.g., "Will you say that again?") (Figure 9). This 2:1 ratio suggests that the normal-language "students" were twice as likely to request clarification of the language—impaired "teachers" than were the normal-language "students" of the normal-language "teachers." However, the actual numbers associated with the "Clarification Request" act are small (6 and 3). Given a total of less than ten occurrences of this conversational act, such a conclusion appears questionable. One additional similarity between the control and ex- perimental groups involved the manner in which the puzzle design tasks were completed. Specifically, in both groups, 71 .hpomopmo :sofio: vomamfloo 93 FA paw mownomome =omcommom= new apmoscom: 93 mo pom. Hggpg>coo zoom Gm moofimsofi: .Lcmvspma Housmaflpomxo mam. HDUEOU mo owmquQHum .m 9»de :mHEman: Hmwcofihmgxm m =mHEoUBm= HQHEOQ g ">9." momcamom msmosvom o I . ma om mm om amuaazad mm o: m: om mm 72 .moanomovmo cosponsm zomcoamom= cam :pmodwom: =.mp:oU:pm: HopcmEAnoaxo mam Honpcoo one canvas muom Hmcoapmmgo>coo mo owmwcmogom .OH osdwflm =mpcopdpm= Hopcoeflnwaxmnmmmm =mvcmpspm= Honpaoo mmw "hex mpmosvom momcoamom Salaam 4918 :99» Sara: .988 058s 99. .399}: .988 058s afieiueaaed 73 the "teacher" produced an "Action Request" and then relied on visual cues and "student" responses to carry out that "Action Request." This can be seen by comparing both the control and experimental "teachers'" uses of "Action Re- quests" with their uses of "Yes/No Responses" and "Clari- fication Responses" (Figure 7). The approximately 2:1 ratio of "Requests" to "Responses" suggests that the "teachers" relied heavily on responding to, rather than requesting, "student" behaviors. For example, a typical "teacher" "Clarification Re- sponse" was, "Turn it around," to a "student's" "Yes/No Request" of, "Like this?" In most cases, "Yes/No Re- sponses" and "Clarification'Responses" were used by a "teacher" as a response to the "student's" search for and/ or correct guess of the intended puzzle piece or location. Thus, rather than depending on the use of only "Action Re- quests" to complete the task, both normal-language and language-impaired "teachers" also responded to "student" comments and behaviors. The control and experimental "stu- dents" also carried out both requesting and responding roles. Thus, the manner in which the two groups used con- versational acts to complete the task was similar and showed no qualitative differences. In conclusion, there is no strong evidence that either "student" or "teacher" utterance function behaviors were substantially different enough to contribute to qualitative differences in the communication acts of the groups or to 79 account for the larger number of communication acts pro- duced by the control group. Consequently, the only other plausible explanation for the control group's use of a larger number of communication acts in completing the puzzle task is risk-taking behaviors. The control group apparently was willing to take more verbal instruction risks than was the experimental group--especially the experimental group "teachers." .Since the normal-language "teachers" took more verbal risks, they made more verbal attempts and, in turn, produced more conversational acts than did the language-impaired "teachers." The normal-language "teach- ers'" expression of a larger number of conversational acts led to the production of more communication acts than were produced by the language-impaired "teachers" in the fifteen minute intervals. It is conceivable that the normal-language "teachers'" facility and confidence in verbal language use made them more willing than the language-impaired "teachers" to take risks in communicating the instructions necessary for com- pleting the task. Specifically, if a communication act was unsuccessful, the normal-language "teachers" may have felt more confident about repairing or revising their messages. Hence, the normal-language "teachers" talked more and pro- duced more communication acts than did the language- impaired "teachers" to complete the same number of puzzle designs. 7Conversely, the language-impaired "teachers'" reluctance to take verbal risks led to their talking less 75 and their production of fewer conversational acts and fewer communication acts. Interestingly, Shipley and McFarlane (1981) suggest that since efficient readers are willing to read quickly and risk making mistakes in order to compre- hend the content, risk-taking behaviors should be encouraged in children with reading problems. Number of Unsuccessful Communication Acts. It was hypothesized that dyads containing a language—impaired sub- ject as the "teacher-speaker" would produce more unsuccess- ful communication acts than dyads with a normal-language "teacher-speaker." It was believed that the language difficulties of the language-impaired subjects would con- tribute to a greater number‘of communication breakdowns and, thus, more unsuccessful communication acts. The oppo- site occurred: The control group produced a slightly larger percentage of unsuccessful communication acts than did the experimental group (59%:50%). The experimental group was equally likely to produce either successful or unsuccessful communication acts. While the control group showed a slight tendency to produce more unsuccessful acts, generally, it was also almost as likely to produce a suc- cessful act as an unsuccessful act. The percentages of unsuccessful and successful communication acts for the con- trol and experimental groups and for each individual dyad are shown in Figure 11. A slight exception to the trend of producing an equal number of successful and unsuccessful communication acts is ooooooooooooooooooooooooooooo ooooooooooooooooooooooooooooo ooooooooooooooooooooooooooooo ............................. ooooooooooooooooooooooooooooo ooooooooooooooooooooooooooooo ooooooooooooooooooooooooooooo oooooooooooooooooooooooooooo 0000000000000000000000000000 OOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOOO ............. mmmmmmmmmm r—l OOOOOOOOOOO 77 apparent when the communication acts of the individual dyads are viewed. The three control dyads each had higher per- centages of unsuccessful communication acts than only two of the three experimental dyads. The remaining experimental dyad, E-M, produced a greater percentage of unsuccessful communication acts (60%) than any of the other five dyads. The proportion, mean, standard deviation, variance, and range of unsuccessful and successful communication acts for the control and experimental groups are listed in Table 1. While the amount of variation associated with the number of unsuccessful acts is similar for the control and experi- mental dyads, the variance for the successful communication acts differs for the two groups. The amount of variation associated with the number of successful communication acts for the experimental group is three times that of the con- trol group. The smaller percentage of successful acts in one experimental dyad, E-M, appears to largely be responsi- ble for the greater variation within this group. The control dyads' production of a slightly larger num- ber of unsuccessful communication acts than two of the three experimental dyads appears to be related to the fact that they also produced a greater number of utterances and commu- nication acts. Since they produced more communication acts, they increased their number of opportunities for making errors. Thus, they produced more communication acts than the experimental dyads; and they also produced more unsuccessful communication acts. 78 .wmm u z A . .90: u 28 mmnmm mo.m m.mo m.mm Amna muim: mm.mH H.mmH o.mm puma Hmvcoeflpomxm Hmnmm sm.oH m.omH m.mm mbam soumm Hm.: m.om m.mm owma Honpcoo owcmm .Q.w mm x * omcmm .Q.m mm x s @3090 muo< HammmoUUSmca muo< Hammmooosm mmsopw Hopcmefimoaxm paw Hopscou any as muo< coquOACSEEoo HammmoOOSmCD so Hammmmoosm mo mmcmm can .coawma>mn caduceum .oocmfind> .coox .popfisz H QHQMH 79 Responsibility for success or failure. For both the control and experimental groups, "teachers" and "students" were almost equally likely to be responsible for the suc— cess of a communication act. Forty-eight percent of the control group's successful communication acts occurred as a result of adequate "teacher" verbal direction. A slightly lower percentage of the experimental group's suc- cessful communication acts (99%) were a result of the "teachers'" adequate verbal directions. The percentage of successful acts for which the control and experimental "teachers" or "students" were responsible is shown in Figure 12. In both groups, the remaining successful communication acts--52% or 56%-- occurred because the "students" correct- ly guessed an intended puzzle piece or location in spite of insufficient or faulty directions from the "teachers." Qualitatively, the control and experimental "students" per- formed in a similar fashion. The experimental "students" were responsible for slightly more (9%) of the successful communication acts than were the control "students." How- ever, this minor difference is because of one control dyad only (C-M), and thus, it should not be concluded that the experimental normal-language "students" contributed more to the success of the communication acts than did the con- trol "students."- The number of successful and unsuccessful acts for which "teachers" or "students" were responsible is listed for each control and experimental dyad in Table 2. 80. 100 ..".".".n-fi. ‘I—I' fii fif‘v ooooooooooooooooooooooooooooooooo ooooooooooooooooooooooooooooooooo ooooooooooooooooooooooooooooooooo ooooooooooooooooooooooooooooooooo 1 90 ................................. oooooooooooooooooooooooooooooooooo h 00000000000000000000000000000000 1 oooooooooooooooooooooooooooooooooo 000000000000000000000000000000000 1 oooooooooooooooooooooooooooooooooo 30 .................................. ooooooooooooooooooooooooooooooooo i oooooooooooooooooooooooooooooooooo 0000000000000000000000000000000000 oooooooooooooooooooooooooooooooooo 7 oooooooooooooooooooooooooooooooooo 0 b 0000000000000000 b 0000000000000000 00000000000000000000000000000000 oooooooooooooooooooooooooooooooooo 0000000000000000000000000000000000 oooooooooooooooooooooooooooooooooo 60 OOOOOOOOOOOOOOOOO p 0000000000000000 ................................. b ................ t ................ ooooooooooooooooooooooooooooooooo ................. l................ t 00000000000000000 p oooooooooooooooo 50 OOOOOOOOOOOOOOOOO p OOOOOOOOOOOOOO l- oooooooooooooooooooooooooooooooooo OOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOO 1:\\ Percentage Group Control Exper1mental Group Group Key: "Teacher"-responsible "Student"-responsible Figure 12 . Percentage of successful cormmnication acts for which "teachers" or "students" were responsible in the control and experimental group. 81 . .omH n zo .sma u 20 .sHN u 2n .sma u zm omH Hopoe puma HovOH OOOH HMFOH obs Hopoe .mlII SIm flmIII 97m WImIII SIm ImmlI EIm : I HmIm Hm I HmIm mm I HMIM um I HmIm Hopcoeflnoaxm : I zIm mm I zIm mm I zIm 5H I zIm cam HopOH AmmH Hmwoe. Mum HopOH com Hovoe mII CIo AMII SID MImlI and NNII Ed 3 I HMIU mm I HmIo o: I HmIo ma I HmIo Hopscoo me I zIo as I 2.0 em I zIo me I qu mwwaflnfimcommom Nmfiawnwmpommmm zpwaflnflmcomwom mpwawnflmQOMmom apcooSpm: :ponomohz avcooSpm: :hmnomoea mpo< HawmmooosmCD mpo< Hammmooosm momzo HmvcoEwhmmxm 0cm Hoppcoo onu cw oHnwmcommom who: :mpcoospm: no :mnoaomoea noes: Low mvo< Hammmooosmcz ocm wsmmmooosm Mo Lonssz m oHQMH 82 In Table 2, it is clear that the "teachers" of both groups usually were responsible for the failure of a commu- nication act. The percentage of unsuccessful acts for which "teachers" or "students" were responsible in the two groups and in each individual dyad is shown in Figure 13. As for the successful communication acts, the control and experimental dyads were quite similar in "teacher" respon- sibility for the unsuccessful acts--that is, 89% and 93%, respectively. Again, the 9% difference between the two groups is due to the slightly lower percentage of dyad C-M. Thus, given the design of the study, the normal-language and language-impaired "teachers" contributed almost equally to the failure of their dyads' unsuccessful communication acts. The remaining question, therefore, is whether the- communication failures of the language-impaired "teachers" differ from those of the normal-language "teachers" in (1) the type of pragmatic violations made, or (2) the under- lying semantic, syntactic, or phonological factors that contribute to the pragmatic violations. Pragmatic violations. The pragmatic violation pat- terns of the experimental and control groups' unsuccessful communication acts were similar in two behaviors. For both groups, "Quantity" violations were the most common type of error that contributed to unsuccessful communication acts. There was only one instance of a "Topic" error, and it occurred in an experimental dyad (E-F2). Thus, "Topic" errors were non-existant in the control group and almost so 89 . in the experimental group. There were three pragmatic behaviors in which the con- trol and experimental groups differed--"Clarity," "Accu- racy," and "Listener Behavior" violations. "Clarity" and "Accuracy" violations were greater for the experimental group than for the control group. In addition, the amount of variability associated with the "Clarity" and "Accuracy" violations was greater for the experimental group. "Lis- tener Behavior" violations were slightly greater for the control group than for the experimental group. The ranked percentages for the pragmatic violation types in the con- trol and experimental groups are listed in Table 3. The proportion, mean, standard deviation, variance, and range of the types of pragmatic violations for the experimental and control groups are summarized in Table 9. Explanations for the groups' differences in pragmatic violations and in the variability associated with the vio- lations can be explained best by examining the behaviors of the individual dyads within the experimental and control groups. The percentages of types of pragmatic violations for each individual dyad are shown in Figure 19. The ranked number of pragmatic violations for each individual dyad are listed in Table 5. There was only a 5% difference-between the percentage of "Clarity" violations (Table 3) made by the experimental (12%) and the control group (7%). Within the control group, dyads C-M and C-Fl were similar in that approximately 85 Table 3 Ranked Percentages of Pragmatic Violations for the Control and Experimental Groups Control Group Experimental Group . Pragmatic , Pragmatic Rank Violation % Rank Violation % 1 Quantity 79 1 Quantity 69 2 List. Beh. 11 2 Clarity 12 3 Clarity 7 3 Accuracy 11 9 Accuracy 3 9 List. Beh. 7 5 Topic 0 5 Topic 1 86 .0eH u 2a .0Hm u 20 II 0.0 0.0 0.: nae mHIe 00.0 0.xm 0.0 mam coa>mamm pocopqu H I0 0m. 00. m. be II 0.0 0.0 0.0 80 oaaoe NHI: 00.: m.H0 0.0 A00 0 IN 0m. 00. 0.N we somcaoo< mHIm 0H.0 0.0m m.s ham 0 I0 00.: 0.HN 0.0 had schemao mmImm :0.HH 0.0:H m.ae name 00I00. 00.0H H.sHH 0.x0 mama seepemao mmcmm .m.m m x * mmemm .a.m a x a aoeemaoa> N m 00pmfimohm mzopw Hmwcoefisomxm @5090 Honpcou mmsopw Hmecoefipomxm oam Hogwaoo one how mCOHHMH00> oflmeMMQm mo mmcmm ocm .cowpmfl>ma Unmocmwm .mo:mapm> «coo: nmonasz : MHQMB 87 .omzo 30.559098 one HQHEOO zoom mom mcoapmaow> owwflcmmnm mo ommvcooaom .1; warm 90.333 .3501qu a owmoH. a Egon/w E bflfimao m bflpfinoa "hex Nulm Hulm , Elm NMIU HMIO ZIU UMhQ 0 OH o N cm 0+. on 3891113319,; ow on om om 8H 88. mm mm 00 0: 00 H0 Hmpoe .wlll . owaoe dill. OHQOH .mlll. UMQOH .mlll OHQOH mill. owmog mill , canoe 0 :9 .33 m 3:98 : .90 30.3 0 3390 N 3982 m 3:909. 3 3098:: 3 .60 .33 z 3:98”: N 3m98< : :mm .33 0 33:8 2 3390 NH 33.9833 0 3320 m .49 .303 0 3330 3 :3 303 mm 33:90 8 33:90 3 33:90 8 33:90 2. 33:90 a, 33:90 w aoflwwH0fl> w coavmaow> w cowpmaow> w COflpmH0fl> w GOHpMHow> m cOHpMH0fl> Sum Elm znm 8-0 8-0 2.0 . mvmza Hasnfl>wccH pom mcoHPMHOH> owvmemmnm mo gmnspz noxcmm m manme 89 10% of their unsuccessful communication acts were coded as "Clarity" violations. The one remaining control dyad, C-FZ, did not make any "Clarity" violations. In the ex- perimental group, all three dyads made "Clarity" violations. While two of the experimental dyads' percentages of "Clari- ty" violations were less than the control dyads', one ex- perimental dyad, E-F2, made more than twice the percentage of "Clarity" violations than the control dyads or the other two experimental dyads made. Thus, the larger percentage of "Clarity" violations in the experimental group was the result of one dyad, E-FZ (Figure 1a). For "Listener Behavior" violations, the percentage difference between the control (11%) and experimental (7%) groups also was small (u%). However, as shown in Table 3, the number differences between the two groups suggests that the control group was twice as likely as the experimental group to make "Listener Behavior" violations (2n:12). In Table 5, it can be seen that this difference was the result of one control dyad. Dyad C-M was responsible for almost three-fourths of the control group's "Listener Behavior" violation (IS/29). As a result of C-M's large number of errors, the control group made twice as many "Listener Behavior" violations as the experimental group. With the exception of dyad C-M, the control and experimental dyads were similar in "Listener Behavior" violations. This simi- lar "Listener Behavior" pattern for five of the six dyads and the more frequent occurrence of "Listener Behavior" 90 violations for dyad C-M can also be seen in the pragmatic violation percentages shown for each dyad in Figure 1H. Differences in the number of "Accuracy" violations were responsible for the most appreciable distinction between the experimental and control groups. The experimental group made almost three times as many "Accuracy" errors as did the control group (20:7). In addition, the amount of variability associated with "Accuracy" violations was much greater for the experimental group (82 = 21.3u; SD = H.62) than it was for the control group (s2 = .35; SD = .59). An even larger amount of variability was seen, how- ever, in the percentage of "Accuracy" violations made within the experimental group. While the control dyads produced either a small percentage of "Accuracy" violations <3%-u%) or no "Accuracy" violations at all, the percentage of "Accuracy" violations made by the experimental dyads ranged from 6% and 7% to 22% (Figure 1a). The majority of the experimental group's "Accuracy" violations, therefore, were made by one dyad--E-Fl. However, even the percentage of the other two experimental dyads' "Accuracy" errors was almost twice as great as that of the control dyads. Furthermore, all experimental dyads made several "Accuracy" _violations, but only two of the control dyads made even a small number of "Accuracy" violations. The results of comparisons of the control and experi- mental groups' pragmatic violations indicate that "Clarity" and "Accuracy" violations were the outstanding errors made 91 by the experimental group. Since these errors were made by the language-impaired "teachers" of the experimental group, it can be concluded that this group of language-impaired adolescents had difficulty providing non-ambiguous and accurate verbal instructions for completing the puzzle task. In spite of the language-impaired "teachers'" difficulties, their normal-language "student" partners were able to com- plete the same number of puzzles as did the normal-language "students" in the control group. Explanations for the language-impaired "teachers'" "Clarity" and "Accuracy" vio- lations, yet their successful completion of the puzzle task, can be found by examining the outstanding behaviors of each of the three experimental dyads. The two language-impaired females produced fewer unsuc- cessful communication acts than any of the other dyads (Figure 11), and they made fewer "Quantity" violations than any of the other dyads (Figure 19). It appears that these two language-impaired subjects took more care to provide their listeners with a sufficient amount of information to complete the task. They both, therefore, were able to suc- cessfully compensate for their difficulties in providing accurate and non-ambiguous information. However, the two subjects did differ somewhat from each other in the nature of their pragmatic violation patterns. Teacher E-Fl's major error was "Accuracy" violations, while Teacher B-F2's major error was "Clarity" violations (Figure 1n). Teacher B-Fl made three times as many "Accuracy" violations as Teacher 92 E-F2. Teacher B-F2, however, made three times as many "Clarity" violations as Teacher E-Fl. Teacher E-Fl's high proportion of "Accuracy" violations appeared to occur for two reasons. First, she tended to accept inaccurate puzzle piece placements more frequently than did the control "teachers." Second, she tended either to inaccurately label the shape and color names of the puzzle pieces or to reverse directional terms such as "left" and "right." Specifically, Teacher E-Fl judged four puzzle pieces of one design as accurate when, in actuality, they were mis- positioned. Although these errors accounted for four of the twelve "Accuracy" violations, even elimination of these exchanges from the analysis resulted in more than twice as many "Accuracy" violations by Teacher E-Fl as the control "teachers." When position-judgment "Accuracy" violations were removed from the analyses of all dyads, Teacher E-Fl still made a greater proportion of "Accuracy" violations than the normal-language "teachers." This suggests that the inaccurate labeling of puzzle shapes, colors, and direc- tional cues was the main reason that this language-impaired female made a substantial number of "Accuracy" violations. The mislabeling of puzzle piece shapes, colors, and directional cues was a relatively common phenomenon for all "teachers." However, the normal-language "teachers" usually corrected their mistakes without feedback from their "students." Teacher E-Fl, however, did not correct 93 her mistakes. Thus, she differed from the normal—language "teachers" in that she rarely corrected her inaccurate statements or naming errors. Teacher E-FZ differed from the control "teachers" and from the other language-impaired "teachers" because of her frequent "Clarity" violations (Figure 1a). She tended to make ambiguous statements which lacked sufficient clarity for the "student" to correctly identify the puzzle piece or its location. Unlike Teacher E-Fl, Teacher E-F2's state- ments generally contained accurate information. However, her word choices were often vague and, therefore, ambiguous, since more than one meaning could be interpreted by the "student." For example, in the "Clarification Response," "The straight piece should be on the bottom," Teacher E-FZ made a true statement concerning the placement of a triangle in the puzzle design. However, the statement is ambiguous since a triangle has three "straight" pieces. Therefore, more than one interpretation of the "teacher's" meaning were possible. If Teacher B-F2 had used a less ambiguous word or phrase, her information would have been adequate for the "student" to place the puzzle piece in the appropriate position. The second remarkable behavior of Teacher E-F2 was her frequent hesitation between and within utterances. She often paused (83 times) from one to nine seconds before say- ing key words that named colors, shapes, and locations. The resulting slowness and deliberateness with which she 9” gave instructions appeared to be a beneficial compensatory strategy for this language—impaired "teacher." In general, Teacher B-F2 appeared to be aware of the necessity of making clear, complete, and accurate state- ments. Her hesitation technique appeared to be most effec- tive in situations where the word choice was limited to specific color or shape names. This strategy, however, lost its effectiveness when several alternatives were available for stating the direction. For example, Teacher E-FZ often used general locational terms, as in the utter- ance, "It's along in front of you," instead of providing more precise instructions such as, "It goes to the right of the other pieces." In such ambiguous cases, Teacher B-F2 frequently did not choose the most appropriate option. The result was an increase in the ambiguity of her instruc- tions and, thus, an increased number of "Clarity" viola- tions. The third language-impaired subject, Teacher E-M, differed from both the control "teachers" and from the other two experimental "teachers." First, dyad E-M produced a larger percentage of unsuccessful communication acts than any of the control or experimental dyads (Figure 11). Second, Teacher E-M produced a larger percentage of "Quan- tity" violations than either of the other two experimental "teachers" and than two of the three control "teachers" (Figure 1a). Also, with the exception of his "Accuracy" violations, Teacher E-M's violation trends were more similar 95 to those of the control "teachers" than to those of the two remaining experimental "teachers." Thus, although Teacher E-M made both "Clarity" and "Accuracy" violations, as did the other experimental "teachers," he also made substan- tially more "Quantity" violations than were made by the other two experimental "teachers." Since the proportion of maxim violations were both similar to and different from the other five "teachers,'" other factors were explored to explain the investigator's sense of a qualitative difference in Teacher E-M's perfor- mance. Closer examination of his pragmatic violations indi- cated that Teacher B-M's use of pronouns for shape names appeared to be a contributing factor to both "Quantity" and "Clarity" violations and, thus, possibly to the investi-- gator's sense of a qualitative difference. Specifically, in numerous "Quantity" violations, Teacher E-M used a pro- noun when use of a specific shape name would have been more appropriate. The "student," however, was able to guess the correct shape because of other visual cues. For example, "the red one" obviously referred to the "red triangle" when the only red shape on the table was a "triangle." In such cases, the "teacher's" violation was coded as a "Quantity" violation when other information-~such as location informa- tion-~was omitted and the pronoun confusion was not the primary reason for the pragmatic violation. In cases where contextual cues did not readily identify the intended shape name for the "student," the "teacher's" use of an ambiguous 96 pronoun referent was coded a "Clarity" violation. Such "Quantity" and "Clarity" violations involving Teacher E—M's use of pronouns appeared to result in a larger amount of "student" confusion than was observed in the other five dyads. The occurrence of more confusion in dyad E-M appeared to explain the investigator's sense of a qualita- tive difference in Teacher B-M's performance. To determine whether these observations could be documented and sup- ported, two post hoc anaylsis procedures were completed on the pronouns produced by all "teachers." Pronoun use. Since contextual cues often allowed a "student" to correctly guess shape names in spite of a "teacher's" ambiguous use of pronouns, successful communi- cation acts as well as unsuccessful communication acts could contain ambiguous pronoun referents. Therefore, it was necessary to analyze all of the pronouns produced by "teach- ers," rather than only those used in unsuccessful communi- cation acts. The first analysis was a tally of each "teachers'" use of anaphoric (referent precedes pronoun), cataphoric (referent follows pronoun), and exophoric (pronoun refers to external reference) pronouns. Since there were only four instances of cataphoric pronoun use, cataphoric pronouns were excluded from further analyses. Exophoric pronouns also rarely occurred; however, they were used much more fre- quently than cataphoric pronouns for both the experimental (22) and the control (56) groups. Anaphoric pronouns were 97 the most frequent type of pronoun used by both the control (221) and experimental (180) groups. The number of each pronoun type used by the experimental and control "teachers" is listed in Table 6. The proportion, mean, variance, standard deviation, and range of anaphoric and exophoric pronouns used by the control group and experimental group "teachers" are listed in Table 7. The percentage of ana- phoric and exophoric pronouns used by the experimental and control groups and by each individual "teacher" are shown in Figure 15. The percentage of exophoric pronouns found in the normal-language "teaChers'" speech (20%) was almost twice that of the language-impaired "teachers" (11%). This find- ing is different from the findings of Tough (1977). Tough (1977) found that normal-language disadvantaged children used more exophoric pronouns than did normal-language chil- dren from advantaged backgrounds. Tough (1977) described the use of exophoric pronouns as a less effective means of communication because the listener was forced to assume a much more active role in following and interpreting the speaker's intended meaning from contextual and gestural cues. However, in this study, normal-language subjects used exophoric pronouns primarily as final judgmental statements that confirmed the position of a puzzle piece. For example, "That's right," and "Like that," were used to confirm that the "student" had placed a puzzle piece in the appropriate position. In such cases, confusions or additional demands 98 Table 6 Number of Anaphoric, Cataphoric, and Exophoric Pronouns Used by Control and Experimental "Teachers" Control Experimental "Teachers" "Teachers" C-M - 111 E—M - 109 Anaphoric C—Fl - 70 E-Fl - 51 C-F2 -__30 E-F2 7__20 Total 221 Total 180 C-M - 2 E-M - 0 Cataphoric C-Fl - 0 E-Fl - 0 C-F2 -___1 E-F2 -___1 Total ' 3 Total 1 C-M - 9 E-M - 11 Exophoric C-Fl - 33 E-Fl - 6 C-F2 -___11 E-F2 '- 5 Total 56 Total 22 99 .mom u z n . .000 u zm Ha um mm.m 0.0 m.0 ANN mmnm :m.o~ m.woa 0.ma Mom capocamxm moauom mm.mm m.omma 0.00 nMom; Haauo: H.mm 0.02m >.m> MHNN caponmmc< mwcwm .Q.m mm x * mwcmm .o.m mm x * mmza Gnocchm :mhmzomwh: Hmucweflpmmxm =mpm£owme: Hoppcoo :mpmgomoe: Hopcmeammxm cam Hompcoo ho pom: masocomm Uflaonaoxm ho camonmmc< mo mmcwm cam .cowpmfi>ma vampCMpm amocmwnm> .cmmz .mmnEDZ 0 mHAMH n H in the 001112101 and 101 on the listener were avoided because of the visual contex- tual Cues afforded by the task situation. I It appears that for the situational demands of this study's task, exophoric pronouns were appropriately and effectively used by normal-language subjects to complete the puzzle design task. For some reason, the language-impaired subjects did not use exophoric pronouns as frequently as the normal-language subjects to effectively confirm their "student's" positioning of puzzle pieces. In any case, the use of exophoric pronouns appears to be related to more than socioeconomic class variables. Situational demands or contextual variables, rather than socioeconomic variables only, appear to influence the frequency and appropriateness of using exophoric pronouns. The frequency of anaphoric pronoun use varied con- siderably within the dyads of both the experimental (20-109) and control (HO-111) groups (Table 6 and Table 7). Although both male dyads used more anaphoric pronouns than the other female dyads (222:181), they did not produce substantially more unsuccessful acts or more "Clarity" violations than did the females (Figure 11 and Figure 14). Thus, frequency data alone does not appear to explain how the "teachers'" use of anaphoric pronouns affected the "Clarity" of their messages. In addition, it appeared that cataphoric and exophoric pronouns did not reduce clarity since cataphoric pronouns were rarely used and exophoric pronouns seemed to enhance the efficiency of the puzzle 102 completion task. The subjects' frequency of pronoun use, therefore, appeared to say more about individual style differences and situational demands of the task than it did about subjects' clarity in giving directions. A second post hoc analysis was completed on anaphoric pronouns to determine whether there were confusions of anaphoric pronoun referents by either the control or experimental group. One complete puzzle design interaction was randomly selected from each dyad for analysis of con- fusions of anaphoric pronoun referents. Each anaphoric pronoun was tallied according to (l) the presence or absence of anaphoric referent confusion--that is, more than one possible pronoun referent; (2) the number of regressive markers which occurred between the pronoun and its anaphoric referent; and (3) the type of linguistic or contextual information that prevented confusions from occurring. The detailed results for this analysis are listed in Appendix L. The number of anaphoric pronoun referent confusions for experimental and control "teachers" are listed in Table 8. The percentage of anaphoric pronoun referent confusions are shown in Figure 16. Only three of the six "teachers"--C-M, C-Fl, and E-M-- used anaphoric pronouns whose referents were unclear or ambiguous. Of these three "teachers," only Teacher E-M used more than one ambiguous anaphoric pronoun. Teacher E-M's frequent use of ambiguous anaphoric pronouns (Table 8) appears to be responsible for a qualitative difference in 103 Table 8 Number of "Teacher" Anaphoric Pronouns that were Confused or Not Confused in the Control and Experimental Dyads Control Dyads Experimental Dyads C-M - 1 E-M - 6 Pronoun C-Fl - 1 E-Fl - 0 Confusion . C-F2 - 0 E-F2 - 0 Total 2 Total 6 C-M — 19 E-M - 10 No Pronoun C-Fl — 5 E-Fl - 7 Confusion ' C-F2 - 6 E-F2 - 2 Total 30 Total 19 109 gumbo 35332.." :08. Gun 98 339% Hmohméflgmaxm can. HofiEoo 93 fi oomsmcoo yo: no Ummdmcoo v.33 “:30. 35.3me 50:89 33938.”: : .mfimnommp: mo mmmucmvumm .ma mgmflm 0.509963 mo £03938 02 a 3:98.39 mo counmdmcoog :82 adopw asoaw :. H: o 2-0 .pmaxm Ho:3:oo I. oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo n333333”33533 om 000000000000 OOOOOOOOOOOOOOOOOOOOOOOO IIIIIIIIIIII OOOOOOOOOOOOOOOOOOOOOOOOO 000000000000 000000000000 OOOOOOOOOOOOOOOOOOOOOOOOO 333333“"333333 om 000000000000 ............ OOOOOOOOOOOOOOOOOOOOOOOO ............ OOOOOOOOOOOOOOOOOOOOOOOOO 00000000000000000000000000 n333333:33an o: OOOOOOOOOOOOO :...........uxxxuxx om ooooooooooooooooooooooo oooooooooooooooooooooooo ooooooooooooooooooooooo OOOOOOOOOOOOOOOOOOOOOOO ooooooooooooooooooooooo ooooooooooooooooooooooo ooooooooooooooooooooooo ooooooooooooooooooooooo 00000000000000000000000 ooooooooooooooooooooooo ooooooooooooooooooooooo ooooooooooooooooooooooo ooooooooooooooooooooooo ooooooooooooooooooooooo eBauaoaed 33333” .............. om OOOOOOOOOOOO ............. OOOOOOOOOOOO 000000000000 OOOOOOOOOOOO 000000000000 mmmmmm imam mmmmmm s nunnxx ”33333 333333 333333 om ooooooooooooooooooooooo oooooooooooo ooooooooooooooooooooooo oooooooooooo 000000000000 OOOOOOOOOOOO OOOOOOOOOOOO 000000000000 OOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOO Db)”’£|--3 IIIIIIIIII I .1 ........... 1 . (144(1i OO— 105 his behavior as compared to that of the other experimental and control "teachers." Teacher E-M's ambiguous pronoun use appeared to force the "student" to use other linguistic and contextual cues to determine his intended meanings regarding shape names. Non-pragmatic factors. Three non-pragmatic factors were investigated as possible underlying contributors to the pragmatic violations committed by the "teachers" in those unsuccessful communication acts for which "teachers" were responsible. The three non-pragmatic factors were syntactic errors (i.e., word order), phonological errors (i.e., morphophonemic), and semantic errors. Two types of semantic errors were investigated--semantic content errors (i.e., ambiguous word choice) and semantic vocabulary errors (i.e., incorrect word choice). §yntactic and phonological errors did not occur. Both "ambiguous" and "incorrect" semantic word choice errors occurred in both the control and experi- mental groups. The percentage of "ambiguous" and "incor- rect" word choice errors for the "teacher"-responsible failures in the control and experimental groups and in the individual dyads is shown in Figure 17. The number of "ambiguous" and "incorrect" word choice errors for the "teacher"-responsible failures in the experimental and control groups and in the invididual dyads is shown in Table 9. The control and experimental dyads were similar in the proportion of "ambiguous" and "incorrect" word choice W /: n "Incorrect" word choice OOOOOOOOOOO MMMMMMMMMM Figure 17. Percentage of "ambiguous" word choice and "incorrect" word choice errors for the " C‘t S "Ambiguous" word choice Key: teacher" — ntrol and experimental groups and in each failures in the co responsible conmunicat ion a dyad . ind ivid 107 sma n 22 .mmH n 20 maum :5.m o.oH Dom ltoHIH mo.m >.m 05H :pomppoocH: mmnmm :>.m s.m: numfi omnom :m.m s.mm Mama =msosmwnfi<= mwcmm .Q.m x , * 0wcmm .Q.m x * momhm mofionu 0&03 «mm090 Hopcoswnmmxm .mmomw Honpcoo mmsopw Hopcmeflpomxm 0:0 Hoppcoo on“ pom whomhm ooflozo 0903 zpomppoocHz Lo :mZOSMflQE<: mo mmcwx paw acowumw>mm Uhmochm .cmoz .anasz m mance 108 errors. "Ambiguous" word choice errors were substantially more frequent than were "incorrect" word choice errors. With the exception of two experimental dyads, EeFl and E-F2, "ambiguous" word choice errors accounted for at least 85% of each "teacher's" semantic errors. Only 71% of Teacher 'E-F2's semantic errors were "ambiguous" word choice errors, and 82% of Teacher E-Fl's semantic errors were "ambiguous." Teacher E-M's error pattern, however, closely resembled that of the control "teachers." Thus, Teacher E-F2 was largely responsible for the experimental group's slightly less per- centage of "ambiguous" errors (82%) than the control group's (89%). The relationship between the semantic errors and the pragmatic maxim violations of the two groups and of the individual "teachers" can be appreciated best by viewing Figure 18 and Figure 19. Figure 18 shows the percentage of "Quantity," "Clarity," and "Topic" pragmatic violations that were either "ambiguous" or "incorrect" word choices for the control and experimental groups. Figure 19 shows the percentage of "Accuracy" pragmatic violations that were either "ambiguous" or "incorrect" word choices for the control and experimental groups and for the individual dyads. For all "teachers," "ambiguous" word choice errors contributed to all of the "Quantity" pragmatic violations; an "ambiguous" word choice error also was responsible for the one "Topic" violation. For all "teachers," "incorrect" word choice errors contributed to all of the "Clarity" 109 .Omse HMOOO>HOcfl sumo Ow cam masogw Housman? cam HOHEOO may 5 9698 030:0 p.53 :uomgoofi... .Ho 030:0 0903 :mdoswfleam... mo PHdmwQ w mhw3,HMSP mfiOfl#MHOH> :UflQOB: Uflm :«hHfithU: =n%#fl#fim50: :.mhM£UMmP: MO 0mmpcwohmm .mH mLSMHh mowozo Ogoz OpoognoucH= ”mm mOflono Ono; zmsosmwga .gfi H9508 .mexm Honchoo 68¢me choo @3090 ...uuuuu4.4.uuuuuuuu.1.1u4.uufiuu O m z 1% ”3333333 ”83333”; mwmwmwmwmw //////////////////// mmmmm m mmm m m mmm OH ”333333; . ”nuuuuuuunnunn"nun”uunnuunufl mmmmmmmmmmmmmmmm m 14 .:-:_:l:; OOH 110 .39 9899905. fifia0finuusmmadewHwflh¥fl9gemvusmao9flfixv0gp a mgofimw0Qfl¥fi.b§¥s:vgyhfioqw.90 . 00Ho£0 U903 =msosww950= mo “H3009 0 0903 H099 mcowpmaow> :>009500a H050fi>fiocH ca 090 005090 H09C0EO90QXM 090 Ho9pcoo 0£H :9 mmcwpmm mSHMpm Hmfloom =90£000B: U0wofio09m 0:0 H0590< 0H 0HQMH 123 One possible explanation for the unexpected similar social status ratings is the design of this study. The design allowed for the examination of social perception skills and peer relationships from a somewhat different view than that of previous studies. Previous studies (Bruininks, 1978; Bryan, 1979a, 1976; Goldman 8 Hardin, 1982; Scranton 8 Ryckman, 1979) focused on children be- tween the ages of six and eleven. This study, however, used an adolescent population. Furthermore, the other investi- gators examined peer relations within classroom settings to insure that the subjects knew one another relatively well and that they had observed one another in a variety of situations. In this study, the subjects were not acquainted and their interaction was limited to one specified task m during a brief fifteen-minute period. This was done to minimize the likelihood that "students" would over- or under-rate "teachers" because of impressions from previous encounters or because of fears of possible future inter- actions with them. If the normal-language subjects did develop negative feelings about the language-impaired sub- jects during the short interaction period, they certainly did not express these feelings in their social status ratings of the language-impaired subjects. Predictions. The language-impaired subjects were simi- lar to the normal-language subjects in their predictions of the ratings that their normal-language peers gave to them. The normal-language "teachers" predicted that their peers 12” gave them a mean social status score of 2.77. The language- impaired subjects responded likewise; they predicted a mean social status score of 2.87. The language-impaired and normal-language subjects' predictions of their status differed by only 1/10 of a point. The small variability of their predictions also was similar (Table 10). Thus, the language-impaired and normal-language subjects appear to have predicted their ratings in a similar manner. The status scores predicted by the normal-language "teachers" and the language-impaired "teachers" for each individual dyad and for the control and experimental groups are listed in Table 10. However, the accuracy of the two groups' predictions was slightly different. The difference between the pre-. dicted and actual scores was slightly larger for the language-impaired group (1.5) that it was for the normal- language (.8). This small difference of 7/10 of a point reflects the fact that the actual received scores of the language-impaired subjects were slightly higher than those of the normal-language subjects. The language-impaired group did rate itself in a manner similar to that of the normal-language group, but the language-impaired group's prediction was 1.5 point lower than its actual received score. The normal-language group's prediction was only u/s of a point lower than its actual received score. Thus, it appears that the language-impaired subjects were slightly less accurate in predicting their status ratings than were 125 the normal-language subjects. However, the difference in accuracy between the two groups was only 7/10 of a point. The slightly less accurate predictions by the language- impaired subjects, however, may not reflect anything unusual about their perceptions; instead, they may be more a reflection of the normal-language "students'" behaviors. Specifically, the experimental "students" may not have been truthful when rating the "teachers," or they may not have said or done anything during thesession to lead the "teachers" to expect a high social status rating. Since the language-impaired "teachers" did not have any unusually positive feedback from the "students," they did not have any reason to expect high ratings. Thus, they predicted rather neutral ratings just as the normal-language "teach- ers" did. In other words, the experimental "students" may have contributed to the slightly less accurate predictions by actually not giving signals that would lead one to expect high ratings. If this were so, discrepancies between the normal-language "students'" feedback signals and their rating behaviors were more responsible for the difference in the actual and predicted ratings than were the percep- tions of the language-impaired subjects. In any case, the reasons for the "students'" slightly higher status ratings of the language-impaired subjects and for the language-impaired subjects' less accurate predic- tions of their ratings are not clear. Thus, any interpre- tations of the results should be made cautiously. At most, 126 the language-impaired subjects showed a slight tendency for less accurate predictions of their social status ratings but not necessarily less accurate perceptions of those ratings. The results, therefore, are considered inconclusive because they neither confirm nor negate other results (Bruininks, 1978; Goldman 8 Hardin, 1982) which have found language- impaired children to be less accurate in their perceptions of their social status among peers. To summarize, the results of this study indicate that (1) these language-impaired adolescents did not receive low social status ratings from normal-language peers, and (2) they were not substantially less accurate than their normal- language peers in predicting their social status ratings. Summary and Conclusions There were four major findings of this study. The first finding was that the pragmatic behaviors of language- impaired adolescents differed from those of normal-language adolescents in two ways. The two differences involved the accuracy and clarity of the language-impaired adolescents' messages. The language-impaired speakers of this study gave inaccurate or ambiguous information to their listeners more frequently than did the normal-language speakers. One research implication of the finding of different pragmatic--conversational--behaviors in language-impaired and normal-language adolescents involves the use of taxono- mies for coding pragmatic behaviors. In this study, every unsuccessful communication act for which the 127 "teacher-speaker" was responsible was coded for one of four pragmatic violations based on Grice's (1975) maxims for conversation. The coding of only one violation per unsuc- cessful communication act implied that only one pragmatic violation was responsible for the failure of an act. How- ever, it is possible for a combination of violations to occur within one utterance and result in a communication act failure. Thus, Chapman's (1981) warning concerning the multiple coding of utterance functions can also be applied to the coding of violations of Grice's (1975) maxims for conversation. A second implication from the finding of differences in language-impaired and normal-language adolescents' prag- matic behaviors concerns the design of the instruction- _ giving task of this study. During the puzzle task, the "student-listeners" were permitted to manipulate their puzzle pieces in an attempt to determine the "teacher- speakers'" intended meanings concerning the completion of puzzle designs. As a result, the "students" frequently guessed at the position of puzzle pieces before receiving complete instructions from the "teachers." This allowed the "teachers" to rely on visual cues from the "students'" puzzle piece manipulations rather than being forced to pro- vide accurate verbal directions. Hence, much of the burden on the "teacher" to give complete, accurate, and clear instructions was reduced. To obtain a more accurate mea- sure of language-impaired and normal-language adolescents' 128 pragmatic violation behaviors, it is recommended that future studies require the "student" to not touch or move a puzzle piece until after the completion of the "teacher's" instructions concerning that piece. The second major finding of this study was that the underlying contributor to pragmatic violations was always semantic in nature. This supports Lucas' (1980) theory that semantics is the underlying basis of pragmatics. Spe- cifically, either ambiguous word choices or incorrect word choices were responsible for all of the pragmatic violations that occurred. The frequency of the language-impaired adolescents' ambiguous and incorrect word choices suggests that their errors may be associated with word-finding difficulties. ‘In addition, qualitative differences in the nature of the semantic errors implies that there are two possible underlying causes for the language-impaired sub- jects' word-finding problems--difficulty with the retrieval of words and difficulty with the semantic organization of messages. The nature and underlying causes of the language- impaired subjects' word-finding problems supports Lucas' (1980) distinction between retrieval and semantic disorders. The finding of underlying semantic causes to pragmatic violations has implications for both clinical and theo- retical research. First, the relationship between semantics and pragmatics needs to be explored further. Second, in individuals who demonstrate word-finding difficulties, a distinction needs to be made between underlying retrieval 129 problems'and underlying semantic organizational problems. In other words, it must be determined whether the indivi- dual's problem is one of recalling specific words or if it is one of not knowing what information needs to be included in an utterance. Only ater the underlying cause of a word-finding problem has been determined can an appro- priate and effective treatment program be established. An incidental finding of this study involved the use of anaphoric pronouns and their referents. One of the language-impaired subjects was found to differ qualitatively from the other subjects in his use of anaphoric pronouns. Specifically, the referents of the male language-impaired subject's anaphoric pronouns often were confusing for the listener. This resulted in an increase in the ambiguity_of his messages and in the amount of burden placed on the listener to use other linguistic and contextual cues to determine his intended meanings. Again, the underlying cause of the pragmatic violations that resulted from the use of confused anaphoric pronoun referents was semantic in nature. The study of language-impaired youngsters' use of anaphoric pronouns has received little attention in previous research. Considering the strong impact of confused anaphoric pronoun referents on the clarity of utterances and the possible relationship between the use of pronouns and word-finding difficulties, the study of anaphoric pronoun use and their-referents is an important area for 130 future research. Also, the relationship between pronoun use and underlying semantic difficulties raises questions con- cerning the processes of semantic organization in language- impaired youngsters. For example, the use of pronouns instead of noun labels has implications for research con- cerning semantic network theories and memory retrieval processes. The final major finding of this study was that the language-impaired speakers did not receive lower social status ratings than the normal-language speakers. This finding differed from those of other researchers (Bruininks, 1978; Bryan, 1979a, 1976; Scranton 8 Ryckman, 1979) who examined the social status of learning-disabled children whose primary disability was in language. The t results concerning the language-impaired adolescents' per- ceptions of their social status among peers was inconclusive because of the small number of subjects used in this study. It was difficult to determine whether the language-impaired adolescents accurately perceived their social status because of two reasons. The first reason was the unexpected finding that the language-impaired subjects actually received high social status ratings from normal-language peers. The second reason was the similarity of the predicted social status ratings from the language-impaired and normal- language subjects. The contradictory social status findings of this study and those of other researchers (Bruininks, 1978; Bryan, 197Ha, 1976; Scranton 8 Ryckman, 1979) and the 131 inconclusive findings concerning the prediction of social status ratings indicate that further research using a larger subject population is warranted in these areas. It should be remembered that the sample size used in this study was small. The results, therefore, should not be generalized to all language-impaired youngsters. Future research involving larger sample sizes is needed before generalization about any of the major findings of this study can be made. The results of this study have several implications for the assessment of language-impaired adolescents. The subtle nature of the subjects' language impairments and the spe- cific involvement of word-finding and anaphoric pronoun use difficulties suggest a need for assessment instruments directed specifically at measuring abilities in the areas of word-finding and pronoun use. Also, assessment proce- dures are needed which focus on adolescents' overall conver- sational effectiveness. Likewise, treatment approaches should emphasize enhancing the conversational effectiveness of language- impaired adolescents. Strategies for organizing the content of messages and strategies for more effectively storing words and word meanings for retrieval should be incorporated in intervention procedures in an attempt to improve the language-impaired adolescent's ability to relay accurate and clear information to a listener. For example, a focus on semantic network theories and the establishment of 132 numerous associations for retrieving specific words may have _potential as an effective intervention procedure. Another psosible treatment approach directed toward improving the accuracy and clarity of messages through underlying semantic organization involves presupposition behaviors. The find- ings of this study suggest that one possible way of working on presupposition skills may be through pronoun use. APPENDICES 133 Appendix A Parent Questionnaire Date: Parent Interviewed: Mother7Father/Both 1. Does know ? son/daughter's name dyad partner's name Please ask if he/she thinks he/she son/daugher's name knows ? dyad partner's name 2. Where did you go to junior and senior high school? 3. How many years of school did you complete? Did you receive a high school diploma? Did you receive a college degree? - Did you receive a graduate or professional degree? If so, what degree? So, counting everything, how many years of school did you complete? u. Where did your spouse go to junior and senior high school? How many 13“ years of school did your spouse complete? Did your Did your Did your degree? spouse receive a high school diploma? spouse receive a college degree? spouse receive a graduate or professional If so, what degree? So, counting everything, how many years of school did your spouse complete? When comes to the Communication Arts son/daughter's name and Sciences Building at Michigan State University for the interaction session, he/she should wear jeans and a tee-shirt. Education Levels of Subjects' 135 Appendix B Parents Subject Father's Education Mother's Education C-M "Teacher" 12 years 12 years "Student" 18 years 16 years C-Fl "Teacher" __a 12 years "Student" 16 years 15 years C-F2 "Teacher" 12 years 15 years "Student" 18 years 16 years E-M . "Teacher" 12 years 12 years "Student" __a 12 years E-Fl "Teacher" 1n years 12 years "Student" 1n years 12 years E-F2 "Teacher" 12 years 12 years "Student" 16 years 12 years aSingle-parent family. 136 OMOH OHOH OHOH .OOOHH H339 HocOHO>HOOm-mm< OH OO OO OO OOH mpoom sum HHHHOHO =HOOOOHO= HO-O OHOH OOOH OmOH OMOH kmOH HOOOHO>HOOm-mm< OH OO OO :O OOH mpoom 30m HOHHOHV =nmaomme= HO-O OMOH OHOH OHOH AOMHH kmOO HocmHm>HOOm-mm< OH OO OO HO OOH 0906O 30m HHHMOHO =HOOOOHO= z-o OmOH OMOH OMOH OMOH OmOH socmHm>HOOm-mm< HO OO OO OO OOH mgoom 30m HOMOHO zpmgommaz z-O .9Ho .c099< m.flO< < .OO< m-O>mm mpommnsm ngo spmwmms< Hmmmmm Hmmmmm HU:H 909 mHm0PQSm mm 0:9 :0 m09oom >0:0H0>flsvmu0w< 0:0 m09oow 30m 137 OOOH OMHH OOOH OMNH OmON HocmHm>HOOm-mm< ON OO OO OO OOH 0900O sum HOHOHO =9cmO29O= z-m OMHH OOMO OOOHH QOMO OHMOH soamHm>HOO9-mm< OH OO O: H: OOH mgoom 30m HOHOHO =9mnomma= z-m OHOH OHOH OMOH OMNH OHHOH HocmHm>HsOm-mw< OH HO OO OO OOH mpoom 3mm HOmOHV =9emospm= Nm-o OOOH OMOH OOOH OmHH OHmOH socmHm>HOOm-mm< ON HO OO OO OOH mpoom 3mm .HOHOHV =pmnomme= 99-0 .9HO .:099< m .OO< < .OO< m-O>mm mpommnsm HO9O spmwwms< Omwwmm Omwwmm «HBO 0:0H0>flsvmu0mHsOm-OO< OH OO OO O: OOH Opoom 30m HOHMOHO =9OOO29O= 99-9 OmOH HOMO OOMHH OOOO nOOHmOH socmHm>HOOm-mm< OH OO O: OO . OHH macaw 30m HOMOHO =90n0009= mm-m OMOH OmOH OmOH OOMHH mem -No:mHm>HmOm-mw< OO OO 9O OO OOH mgoom zmm AOmOHV =9amuapm= Hm-m OOMHH OMOH OmOH nOmHH OHmHH OOOOHO>HOOm-mm< OH OO O: HO OHH mgooO zmm HOMOHO zamnomma= Hm-m .9HO .:099< O .OO< < .Oue m-9>OO Opowfinsm Hmpo O9mwmms< Hmwme Hmmwmm . 53. 3,. , .OO 62. Oz .9 02. 00900m m:990m 900m 9:9om|0>9m 09 10:0 0:9 9:0009Q0m 09 000: 0:090009axm 909009 90 mw:az09m m 390:0am< 153 Appendix G Directions for Instruction-Giving Task To both subjects: You will be putting together puzzles using these shapes. Included are triangles, squares, and rectangles (point to each) in three colors--red, yellow, and blue (point to each). You both have identical sets of puzzle pieces. (Student), these are your pieces and, (teacher), these are yours. The puzzle designs are on these papers. «I will show you how to use these after we go over the rules you must follow while completing the puzzles. Listen carefully. (Teacher) will be the "teacher" and (student) will be the "student." To the "student:" (Student), you must put puzzles together following (teacher)'s directions. He/She is going to tell you where each piece goes. You can ask questions if you do not understand where to put a piece or which piece (teacher) is wanting you to use. However, you can not look at his/her designs or get out of your chair. Do you under- stand? To the "teacher:" (Teacher), you must tell (student) where each piece goes so that his/her puzzle looks exactly like yours. You can look at what (student) is doing to make sure that he/she is putting the pieces in the right places, but you can not point to or touch any of (student)'s puzzle pieces. To make sure that you don't use your hands, you 159 must keep your hands inside this box while you are telling (student) how to make the puzzles. Do you understand? To both subjects: Only (teacher) is allowed to look at these papers of the puzzle designs. We will do two puzzles together to make sure that you understand what to do (Appendix D). (Allow ten minutes, maximum, to complete practice puzzles.) To the "teacher:" After you and (student) have fin- ished a puzzle, remove the top sheet of paper and set up the next puzzle. Remember to put your hands back inside the box before you start to tell (student) where to put his/ her pieces. I To both subjects: Do either of you have any questions .about what you are supposed to do? Do you understand all of the rules? I will be busy running the camera while you are doing the puzzles, so I won't be able to answer any questions after you have started working. Remember, (teacher), you must tell (student) where to put his/her pieces, without touching or pointing, so that the puzzle looks exactly like yours. (Student), you need to remember that you can ask questions, but you can not look at (teacher)'s puzzle designs. Are you ready? I will tell you when to start. Continue working until I say to stop. 155 Appendix H Definitions and Examples of Dore's (1978) Conversational Act Categories Request Category: Any request for information, A. action, or acknowledgment. Yes-No Request 1. Definition Any question seeking a true-false judgment. 2. Example "Like this?" Wh-Request 1. Definition Any either-or or fill-in—the-blank question seeking factual information. 2. Example "What color is at the top?" Clarification Request 1. Definition Any question seeking information about a prior utterance. 2. Example "Will you say that again?" Action Request 1. Definition Any remark requesting that the listener do something. 156 2. Example "Put the blue triangle above the red square." E. Permission Request 1. Definition Any question seeking permission to do some- thing. 2. Example "Can I play with those?" F. Rhetorical Question Request 1. Definition Any question seeking acknowledgment from the listener to allow the speaker to continue. 2. Example "Ya know what?" II. Response Category: Any response to a request. A. Yes/No Response 1. Definition Any answer supplying a true-false judgment. 2. Example "Yeah, like that." B. thResponse 1. Definition Any answer supplying requested factual infor- mation. 2. Example "On the right hand side." 157 Clarification Response 1. Definition Any remark supplying a relevant repetition. 2. Example "Turn it over." Compliance Response 1. Definition Any remark expressing acceptance, denial, or acknowledgment of a prior action or permis- sion request. 2. Example "All right." Qualification Response 1. Definition m Any remark supplying an unexpected response to a soliciting question. 2. Example "But he isn't there." Repetion Response 1. Definition Any remark repeating part or all of a prior utterance. 2. Example "I said, 'The red one goes there.'" 158 III. Description Category: Any description of verifiable past or present facts. Identification Description 1. Definition Any remark labeling objects, events, etc. Example "This is the bottom." Event Description 1. Property description 1. Definition Any remark describing events, actions, processes, etc. Example "I'm setting up the next design." Definition Any remark describing properties, traits, or conditions. Example "So it sticks out." Location Description 1. Definition Any remark expressing locations or direction. Example "It's in there." Time Description 1. Definition Any remark reporting times. 2. 159 Example "I met him this morning." IV. Statement Category: Any statement of facts, rules, attitudes, feelings, or beliefs. Rule Statement 1. Definition Any remark expressing rules, procedures, definitions, facts, etc. Example "We're not allowed to talk." Evaluation Statement 1. Definition Any remark expressing attitudes, judgments, etc. Example "That's good." Internal Report Statement 1. Definition Any remark expressing emotions, sensations, or mental events. Example "I don't know how to say it." Attribution Statement 1. Definition Any remark reporting beliefs about another person's internal states. 2. 160 Example "He doesn't feel well." E. Explanation Statement 1. Definition Any remark expressing reasons, causes, or predictions. Example "It will break." V. Acknowledgment Category: Any acknowledgment that recognizes or evaluates responses or nonrequests. A. Acceptance Acknowledgment 1. Definition Any remark neutrally recognizing answers or nonrequests. Example "Mm. H B. Approval/Agreement Acknowledgment 1. Definition Any remark positively recognizing answers or nonrequests. Example "All right." C. Disapproval/Disagreement Acknowledgment 1. Definition Any remark negatively evaluating answers or nonrequests. 161 2. Example "I don't agree." D. Return Acknowledgment 1. Definition Any remark returning the floor to the speaker by acknowledging a rhetorical question or a nonrequest. 2. Example "Oh." VI. Organizational Device Category: Any remark that regulates contact and conversation. A. Boundary Marker Device 1. Definition _ - Any remark indicating an opening, closing, or significant point in the conversation. 2. Example "That's all." B. Calling Device 1. Definition Any remark soliciting attention. 2. Example "Mark." C. Speaker Selection Device 1. Definition Any remark specifically identifying the next speaker. 162 2. Example "Mark." D. Politeness Marker Device 1. Definition Any remark indicating politeness. 2. Example "Thank you." E. Accompaniment Device 1. Definition Any remark maintaining verbal contact by con- veying information redundant with the con- text. 2. Example "Here we go." _ VII. Performative Category: Any remark that accomplishes a fact by being said. A. Protest Performative 1. Definition Any remark registering a complaint about a listener's behavior. 2. Example "Wait a second." B. Joke Performative 1. Definition Any remark aimed at achieving a humorous effect. 163 2. Example "I'll just stick it there." ("Student" imi- tation of ambiguous "teacher" instruction.) C. Claim Performative 1. Definition Any remark establishing rights by being said. 2. Example ’ "I get the first one." D. Warning Performative 1. Definition Any remark alerting the listener of impending harm. 2. Example "Look!" E. Teasing Performative 1. Definition Any remark annoying, taunting, or playfully provoking a listener. 2. Example "You can't catch me." VIII. Miscellaneous Category: Any remark that is not de- scribed by one of the above categories. A. No Answer-Miscellaneous 1. Definition Two or more second silence from the listener following a question. 2. 16” Example (Silence) B. Uninterpretable-Miscellaneous 1. Definition Any remark that is unintelligible, incom— plete, or anomalous. Example "80 that it..." C. Exclamation-Miscellaneous 1. Definition Any remark expressing an emotional reaction or other nonpropositional information. Example "Oh, wow!" 165 - - OO OO - - HO OpHHHnHOcomOOm 9o< :099009::EEoo 9:9000003m OOH OO OOH OO OOH OOH OO OHHH9OHOcoOOOm 96< COH #MOflGDEEOU HDM m meUDmCD OO OO OO OO OO OO OO OOOH9OHOH> OH9OHOOOHH OO OO OO OO OO OO OO OcoHpmHOHO OHpmeOmgm OO OO OO OO OO OO OO OmHaommpmo 90< 90:0990090>:oo OOH OOH OOH OOH OO OOH OO Opo< eoHHOO9aOesoo mo OcH9mm 9090000000:D\959000003m OO :O OO OO OO OO OO O96< OOH9OOHOOEEOO mo :0990:9E9090Q OO-m HO-m z-m Om-o HO-O z-O O00»: OHmmmma< oHpmewmam mmO90< 9:05009w< 0m030-909:H 90 w 00>Q 90:09>90:H :00m 909 0:0 000>Q wmo9o< m0m>90:< 0990Ew09m :o 9:0E009m< 0w030n909:H mo m0w09:0090m H 390:0am< 166 Appendix J Number of Conversational Acts for Each Utterance Function Category # of Conversational Acts Utterance Function Category Control Experimental '8 Conversational Acts Dyads Dyads Request Category Action Requests 157 - 1H7 Yes/No Requests 65 71 Wh-Requests H ’2 Clarification Requests 3 6 Total 229 226 Response Category Clarification Responses 18% - 133 ~ Yes/No Responses ' 150 117 Compliance Responses 15 I 31 Repetition Responses 1n 7 Wh-Responses 3 0 Total 366 288 Description Category Identification Descriptions 11 1 Location Descriptions 12 ~ 0 Property Descriptions 2 l Total 25 2 167 # of Conversational Acts Utterance Function Category Control Experimental 8 Conversational Acts Dyads Dyads Statement Category Internal Report Statements 10 2 Evaluation Statements 6 l Explanation Statements 2 0 Rule Statements 0 1 Total 18 u Acknowledgment Category Approval/Agreement Acknowledgments 0 3 Acceptance Acknowledgments 0 2 Total 0 5 Organizational Device Category Q Boundary Marker Devices 86 61 Politeness Marker Devices 2 0 Total 88 61 Performative Category Protest Performatives l l Joke Performatives 0 2 Total 1 3 Miscellaneous Category Uninterpretable-Miscellaneous 39 37 Exclamation-Miscellaneous 5 3 Total nu #0 Total # of Conversational Acts 776 629 168 Appendix K Type and Number of Conversational Acts Within Each Utterance Function Category for the Experimental and Control "Teachers" and "Students" "Teachers" "Students" Utterance Function Category 8 Conversational Acts C E C E Request Category Action Requests 157 1H7 I 0 0 Yes/No Requests 0 0 65 g 71 Wh-Requests 0 l u 1 Clarification Requests 0 0 3 6 Total 157 lu8 72 78 Response Category Clarification Responses 18” 133 0 0 Yes/No Responses lu9 116 l 1 Compliance Responses 0 l 15 30 Repetition Responses 13 6 l l Wh-Responses 3 0 0 0 Total 399 256 17 32 Description Category Identification Descriptions 6 1 5 0 Location Descriptions 8 0 u 0 Property Descriptions 1 0 1 l Total 15 1 10 l 169 "Teachers" "Students" Utterance Function Category 8 Conversational Acts C E C E Statement Category Internal Report Statements 5 2 5 0 Evaluation Statements 6 l 0 0 Explanation Statements 2 0 0 0 Rule Statements 0 l 0 0 Total 13 u 5 0 Acknowledgment Category Approval/Agreement Acknowledgments 2 0 l 0 Acceptance Acknowledgments 0 0 2 0 Total 2 0 3 0 Organizational Device Category _ Boundary Marker Devices 79 92 7 l9 Politeness Marker Devices 1 0 1 0 Total 80 M2 8 19 Performative Category Protest Performatives 0 l l 0 Joke Performatives 0 0 0 2 Total 0 l l 2 Miscellaneous Category Uninterpretable-Miscellaneous 35 33 H u Exclamation-Miscellaneous 0 3 5 0 Total 35 36 9 u Total # of Conversational Acts 651 R88 125 136 170 30 0> 99 O \, 2”. HH O\, O . \, 2U 30HHO,$\, 9H 330% 99 O \, 2U v\, 90on0...» 99 H \, O\, . O8 0 O \, 2U V\, 009 99 m \, 2U v\/ 009 99 H x 0 s . 009 99 O \/ HQ v\’ AD 009 3309:00 . .U 009 900093 Q0090 0:0 m \, 0 \’ 00 99 H \, 3 H£\, 900 ~100m 9:00:09.» :090 0 09900000 090030: :0909>0m 09:09090m 9:090w0m 0:002 . ow 090m? 0 . 9:00090 09909009990 >0 00>0m 009090000 09900005., 90 9:000W09002 0>90009w0m >0 00>0m >0 00>0m 090: 90 03.9. :09009:00 .oz 0.9 0 90m 90 0 80% mSHCOO 589$ 00: :00:o9m 09900005» 0. :10 9000009. 90 090>90:< < 0H3. 0 x9883 171 9:090909 00000 >99:0009 9006 00900900 :9:00090: .00000 >990090> 90>0: 003 999090.909 9009900 0 . 9090090909 099000090 009009099 0000095900 :9 :0999090950 0 mm OO 9 o o 9 m9 m9 9090.9. 9 \, 9Q 0090v\, 99 O \ 30:03, 99 900090 H \, 90 858K. 99 O 30an0\’ 99 m \, A<0090v\, 99 O \, 333K, 99 m \/ 9<0090v\’ 99 O \, 3030K, 99 H \, AdsoHH03\, 99 90090900 0:900000 090x902 :09 09>0m 0090090999 9:090 90m 0:002 900 .000 :00:09m 09 090x90: 0.9:0009m 09909009990 >0 00>0m 0909000: 09900005~ 90 958% 900: 0>90009w0m >0 00>0m >0 00>0m 0902 90 039. :09009:00 oz 09 090 900 90 .9. $9.90...” WS-HCOO 589$ 6009988 < 030.9 172 9009:090009 0990000990 00900905 000009:0900 :9 :0990590u9:90 .OH mm OO 9 o o 9 m 0 9090.9. 9 \, AG 0090 v\/ 99 H \, 2U onH03\, 905. 9030990> .D 30990>V H \/ \, 99 O \, a 0090\/ 9009 9 \/ AG 009 V\’ 99 9 \’ AD 0090\’ 9009 90090-900 0:90 0000 090090: :0909>00 0090090900 90090900 0:002 900 .9000 :00:090 8. 0909909 0.98030 0990920990 ,3 8200 039800 09.92005 90 988m 900: 0>9000ch00 >0 00>0m >0 00>0m 090: 90 039. :09009:00 02 09 090900 90 9 0:09 00.9:00 :00:o90 00: :00:090 099000059 0.9.010 9000000. .90 090>90:< 0 0900.9. 173 . 39590909 09900005 00900929 0000095900 :9 :099290950 o 009 99 o o o . o 0 0 9090.9. 2 \, g99mm§ 99 m \/ RU 30990> v\/ 99 N \ 2U 309999 V\, 99 9 \, a”. 309903\, 99 9 \, 3309903\, 99 9 \, AG 0095\/ 99 9:09090m w:90000w 0900902 :0909>0m 009:09090m 9:09090m 0:002 900 9000 :00:09n9 09 0900990: 0 . 9:00090 099090wc90 >0 00>0m 09090000 099000059. .90 9:000m 90oz 0>90009m0m >0 00>0m >0 00>0m 0902 90 039. :0900u9:oo oz 09 090 90% 90 m 0:09009:00 :00:09m 00D :00:0c.£ 099000059 0.90:0 9000008 90 090>90:< 0 099908 174 2999 AEPCOO «D 89 900093 .0309903 0 \’ \, 5009 9 \, 200963 99 0098 .089903 . m \z \/ 0:0 9089 n99.309993 I x x a 99x09:00 Amu009 90009>V .nu30990>v m x x a.” 9 \> Amu009y\, 99 N \, 903099033 99 9 \, 5.69903\, 99 9:090900 lw:90000w 090x902 :0909>0m 009:090900 9:090900 0:002 9009000 :00:090 09 0909902 0 . 9:00:90 0990950990 3 00290 09.990000 0992020 90 900000 900: 0>90009w0m >0 00>0m >0 00>0m 0902 90 039 :09009:00 02 09 090900 “90 0 0:09 00 9:00 :00:090 00: :00:o90 09900005~ 0.21m 9000000. 90 090>90:< Q 09000. 175 m . mm m . mm 99 z 9 9 0 O9 o9 9090.9. 9 \, 9Q 30990> V\» 99 A 0000 AD 30990> :0990:09:9v .G 0095 H x x 9 m \, U 309903\, 99 990909000 AD 30990> 99.5093 6309903 m \9 0 \9 0:0 9 \, AU B90\, 99 0 \z A 00090 v\/ 99 m \/ A G 0090 v\' 99 m \, AG 0090 V\z 99 9:090 0900 w9990000w 0900902 :09 09>0m 009:090900 9:090 .900 0:002 900 9000 50:80 09 09099902 0 . 9:00090 099090m090 >0 00>0m 09090000 0990000:< “90 9:080 9002 0>90009w0m >0 00>0w >0 00>0m 0902 90 03.9. :0900.9:oo 02 09 090 .900 90 9 0:090099900 :00:090 9902990980 0 999999 176 .9:090909 00000 >99:0009 9006 00900900 =9:0009m= .00000 >990090> 90>0: 00309:090909 90099000 .A0v9:090909 0990000:0 0090090:9 000009:0900 :9 :09905909090 9:090900 m:90000w 0900902 :0909>00 009:090900 9:090900 00002 9009000 000:090 09 0900902 0.9:0009m 099090m:90 >0 00>0m 09090000 0990000:< 90 9:0000 9002 0>90009m00 >0 00>0m >0 00>0m 0902 90 036 :09009:oo 02 09 090900 90 9 wCOH mam—”COO £309.90er 9829983 0 09909 177 . 9030090909 0990000.:0 0090090:9 0000095900 :9 :099099909090 o 959 99. o o o o 9. 9. 9.0900. 9 \, AWUF009V\, 99 N \, AG. 30990> v\, 0:0 9 \, 3309903\, 096 9 \, 8 89$ 98 . 9008 0% 9 \9 U 30990> V\’ 8009 9 \, an... 309903\, 99 9 \, 90 .< 0:98\, 9009 9:090 900 M:90000w 0900902 :0909>00 009:090900 9:090900 0:002 900 W000 9900:0990 . O9 0.90990: 0.95030 09990509999 E 00200 09090000 09999008 90 958m 900: 0>90009w00 >0 00>0m >0 00>0m . 0902 90 039. :09009000 02 09 090900 90 9 0:09009:00 :00:090 00: :00:090 099000059 0.90:0 9000000. 90 090>90:< 0 0900.9. 178 . 90 9590909 099000000 0090090:9 0000095900 999 :099099990950 o . 89 o o o N m 9 \ 9000903 9 \ 2.899033 9:090900 w:90000o 0900902 :0909>00 009:090900 9590900 :002 9009000 09 0900902 0950090 099090wc90 >0 00>0m 09090000 099000059 90 9:0000 9002 0>90009w00 >0 00>0m >0 0900 0900 90 03.9. :09 009:00 02 09 090900 90 9 0:09 009:00 :00:090 00: :00:Q90 099000059 0.310 9000000. 90 090>90:< .0 0990.9. REFERENCES 179 Reference Notes 1. Deal, L.V. Personal communication, Fall 1982. 180 References Austin, J.L. How to do things with words. Cambridge, Massachusetts: Harvard University Press, 1962. Baker, H.J., 8 Leland, B. Detroit Tests of Learning_Apti- tude. Indianapolis: The Bobbs - Merrill Company, Inc., 1959. Bates, E. Language and context: The acquisition of_prag- matics. New York: Academic Press, 1976. Blank, M., Gessner, M., S Esposito, A. Language without communication: A case study. Journal of Child Language, 1979, 8, 329-352. Bloom, L., 8 Lahey, M. Language development and language disorders. New York: John Wiley 8 Sons, 1978. Boyce, N.L., 8 Larson, V.L. Adolescent's Communication: Development and disorders. Eau Claire, Wisconsin: Thinking Ink Publications, 1983. Brinton, B., 8 Pujiki, M. A comparison of request-response sequences in the discourse of normal and language- disordered children. Journal of Speech and Hearing Dis- orders, 1982, 31, 57-62. Bruininks, R.H., Rynders, J.E., 8 Gross, J.C. Social acceptance of mildly retarded pupils in resource rooms and regular classes. American Journal of Mental Defi- Ciency, 197M, 18, 377-383. Bruininks, V.L. Peer status and personality characteristics of learning disabled and nondisabled students. Journal of Learning Disabilities, 1978, ll, nan-use. Bryan, T. Peer popularity of learning disabled children. Journal of Learning Disabilities, 1974, 1, 621-625. (a) Bryan, T. An observational analysis of classroom behaviors of children with learning disabilities. Journal of Learning Disabilities, 197u, 1, 26-34. (bTE Bryan, T. Peer popularity of learning disabled children: A replication. Journal of Learning Disabilities, 1976, 9, 307—311. Bryan, T. Social relationships and verbal interactions of learning disabled children. Journal of Learning Dis- abilities, 1978, 1, 107-115. 181 Bryan, T., Wheeler, R., Felcan, J., 8 Henek, T. "Come on, Dummy": An observational study of children's communica- tions. Journal of Learninngisabilities, 1976, g (10), 661-669. Chapman, R.S. Exploring children's communicative intents. In J.F. Miller (Ed.), Assessing language production in children: Experimental_procedures. Baltimore: Univer- sity Park Press, 1981. Coggins, T., 8 Carpenter, R. Categories for coding pre- speech intentional communication. Unpublished manuscript, University of Washington, 1978. Coopersmith, S.A. Method for determining types of self- esteem. Journal of Abnormal and Social Psychology, 1959, £9, 87-9”. Denckla, M.B., 8 Rudel, R.G. Naming of object-drawings by dyslexic and other learning disabled children. Brain and Language, 1976, 3, l-15. Dore, J. Holophrases, speech acts and language universals. Journal of Child Language, 1975, 2, 21-80. Dore, J. Requestive systems in nursery school conversa- tions: Analysis of talk in its social context. In R- Campbell 8 P. Smith (Eds.), Recent advances in the psychology of language: Language development and mother-child interaction. New York: Plenem Press, 1978. Dore, J. Conversational acts and the acquisition of lan- guage. In E. Ochs 8 B. Schieffelin (Eds.), Develop- mental pragmatics. New York: Academic Press, 1979. Dunn, L.M. Peabody Picture Vocabulary Test. Circle Pines, Minnesota: American Guidance Center, 1981. Fordyce, W.G., Yauck, W.A., 8 Raths, L. A manual for the Ohio Guidance Tests for the Elementary Grades. Columbus, Chic: Ohio State Department of Education, 19H6. Gallagher, T.M. Contingent query sequences and their rela- tions in discourse. Journal of Child Language, 1981, 8, 51-62. Gallagher, T.M., 8 Darnton, B.A. Conversational aspects of the speech of language-disordered children: Revision behaviors. Journal of Speech and Hearing Research, 1978, 21, 118-135. 182 Garry, R. Guidance techniques for elementary teachers: Guess Who Sociometric Technique. Columbus, Charles E. Merrill, 1963. Garvey, C. The Ohio: Requests and responses in children's speech. Journal of Child Language, 1975, 2, Hl-63. Glucksberg, 8., 8 Krauss, R. have learned how to talk? What do people say after they referential communication. Studies of the development of Merrill-Palmer Quarterly Behavior and Development, 1967, 13, 309-316. Goldman, R.L., 8 Hardin, V.B. The social perception of learning disabled and non-learning disabled children. The Exceptional Child, 1982, 29 (I), 57-63. Gough, H. Manual for the California Psychological Inven- tory Revised. Palo Alto, California: Consulting PsychologiSts Press, 1975. Grice, H.P. Logic and conversation. In P. C01e 8 J.L. Morgan (Eds.), Syntax and semantics, Vol. 3° acts. New York: Academic Press, . Speech 1975. Halliday, M.A.K. Explorations in the functions of lan- guage. London: Edward Arnold, 1973. Learning how to mean: Explorations in- the development of language. New York: Elsevier - North Holland Publishing Company, 1975. Haniff, M.B., 8 Siegel, G.M. The effect of context on ver- bal elicited imitation. Journal of Speech and Hearing Disorders, 1981, 56, 27-30. Halliday, M.A.K. Hartup, W.W. Peer interaction and social organization. P.H. Mussen (Ed.), Carmichael's manual of child psy- chology. New York: Wiley, 1970. Hartup, W.W., Glazer, J.A., 8 Charlesworth, R. forcement and sociometric status. 1967, 38, 1017-102u. Hymes, D. In Peer rein- Child Development, Competence and performance in linguistic theory. In R. Huxley 8 E. Ingram (Eds.), Language acquisition: Models and methods. New York: Academic Press, 1971. James, S. Effect of listener age and situation on the politeness of children's directives. Journal of Psy- cholinguistic Research, 1978, 7, 307-317. Johnson, D.J., 8 Myklebust, H.R. Learning disabilities: Educationalgprinciples and_practices. New York: Grune 8 Stratton, 1967. 183 Karp, 3., a Konstadt, N. Children's Embedded Figures Test. Palo Alto: Consulting Psychologists Press, 1971. Lakoff, R. The logic of politeness; or minding your p's and q's. In C. Corum, T. Smith-Clark, 8 A. Weiser (Eds.), Papers from the Ninth Regional Meeting of the Chicago Linguistic Socieny. Chicago: UniVersity of Chicago, Department of Linguistics, 1973. Lasky, E.Z., 8 Klopp, K. Parent-child interactions in normal and language-disordered children. Journal of Speech and Hearing Disorders, 1982, 31, 7-18. Loban, W. Language development: Kindergarten through grade twelve. Urbana, Illinois: National Council of Teachers of English, 1976. Longhurst, T.M., 8 Reichle, J.E. The applied communication game: A comment on Muma's "Communication Game: Dump and Play." Journal of Speech and HearingADisorders, 1975, 3Q, 315-319. Lucas, E.V. Semantic andnpragmatic language disorders: Assessment and remediation. Rockville, Maryland: Aspen Systems Corporati0n, 1980. Miller, L. Pragmatics and early childhood language disor- ders: Communicative interactions in a half-hour sample. Journal of Speech and Hearing Disorders, 1978, 33, u19-u36. Muma, J. The communication game: Dump and play. Journal of Speech and Hearing Disorders, 1975, 18, 296-309. Muma, J. Language handbook: Concepts, assessment, inter- vention. Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1978. Nowicki, S. 8 Strickland, B. A locus of control scale for children. Journal of Consulting and Clinical Psychology, 1973, fig, 148-15u. Piaget, J. (The language and thought of the child) (M. Gabain, Trans.). Cleveland: Meridian Books, 1955. Piers, E. Manual for the Piers-Harris Children's Self Con- cept Scale. Nashville: Counselor Recordings and Tests, 1969. Prutting, C. Pragmatics as social competence. Journal of Speech and Hearing Disorders, 1982, 31, 123-134. 18H Rees, N.S. Pragmatics of language: Application to normal and disordered language development. In R.L. Schiefel- busch (Ed.), Bases of language intervention. Baltimore: University Park Press, 1978. Retherford, K., Schwartz, B., 8 Chapman, R. Semantic roles in mother and child speech: Who tunes into whom? Journal of Child Language, 1981, 8(3), 583-608. Rom, A., 8 Bliss, L.S. A comparison of verbal communica- tive skills of language impaired and normal speaking children. Journal of Communication Disorders, 1981, 33, 133-1H0. . Rucker, C.N. Acceptance of mentally retarded junior high school children in academic and non-academic classes. Unpublished doctoral diesertation, University of Iowa, 1967. Schachter, P.F., Kirshner, K., Klips, B., Briedricks, M., 8 Sanders, K. Everyday preschool interpersonal socio- linguistic speech usage: Methodological, developmental, and sociolinguistic studies. Monographs of the Society for Research in Child Development, 197%, 33, (Serial No. 156). Schutz, W.C. The PIRO Scales Manual. Palo Alto, Califor- nia: Consulting PsychologiSts Press, 1967. Scranton, T.R., 8 Ryckman, D.B. Sociometric status of learning disabled children in an integrative program. Journal of Learning Disabilities. 1979, 33, (6), #02- ”07. Searle, J. Speech acts. Cambridge, Massachusetts: Cam- bridge University Press, 1969. Searle, J. A classification of illocutionary acts. Lan- guage and Society, 1976, 3, 1-23. Shipley, K.G., 8 McFarlane, S.C. Facilitating reading development with speech- and language-impaired children. Language, Speech, and Hearing Services in Schools, 1981, igC23,100-106. Soenksen, P.A., Flagg, C.L., 8 Schmeits, D.W. Social communication in learning disabled students: A pragmatic analysis. Journal of Learning Disabilities, 1981, 33(5), 283-286. Spekman, N.J. Dyadic verbal communication abilities of learning disabled and normally achieving fourth- and fifth-grade boys. Learning Disability Quarterly, 1981, 3, 139-151. 18.5 Tough, J. The development of meaning: A study of chil- dren's use of language. New York: John Wiley 8 Sons, Inc., 1977. U.S. Office of Education, Federal register, Part II. Washington, D.C.: U.S. Department of Education, 1977, 65083. Watzlawick, P., Beavin, J.H., 8 Jackson, D.D. Pregmatics of human communication: A study of interactional pat- terns, pathologies, and_paradoxes. New York: W.W. Norton 8 Company, Inc., 1967. Wiig, E.H., 8 Semel, E.M. Languege assessment and inter- vention for the learning disabled. Columbus, Ohio: Bell 8 Howell Company, 1980. Wiig, E.H., Semel, E.M., 8 Nystrom, L.A. Comparison of rapid naming abilities in language-learning-disabled and academically achieving eight-year-olds. LanguageL Speech, and Hearing Services in Schools, 1982, 33(1), 11-23.