THESIS” if“ 3w.ilfliall‘{ Mic h) i; an State University This is to certify that the thesis entitled THE USE OF SYNTACTIC CUES BY GOOD AND POOR READERS IN RATING AUTHORS presented by w Eric Charles Stemle has been accepted towards fulfillment of the requirements for .. A M degree in 2.17% ajor professor Date 2/1 7/37, 0-7639 OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MATERIALS: Place in book return to remove charge from circulation records THE USE OF SYNTACTIC CUES BY GOOD AND POOR READERS IN RATING AUTHORS BY Eric Charles Stemle A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of English 1981 I .g-I' 'I‘JP‘" G // -..5 my“... ABSTRACT THE USE OF SYNTACTIC CUES BY GOOD AND POOR READERS IN RATING AUTHORS BY Eric Charles Stemle The purpose of this study was to explore the percep- tions readers derive from features of an author's prose style. Two narratives, each written in a simple and a complex ver- sion, were constructed with syntactic complexity, specifically left branching, as the controlling variable. One hundred four high school students enrolled in "Speed Reading" and "Develop- mental Reading" read the simple version of one narrative, the complex version of the other, and then evaluated the authors' skill and background in six dimensions; age, intelligence, and level of education were the items of primary interest in this study. Findings of most importance were 1. Good readers differentiate between authors on the basis of syntactic cues. 2. Good readers rate authors using complex syntax as older, more educated, and more intelligent than authors using simple structures. 3. Good readers are more consistent in perceiving author differences than are poor readers. TABLE OF CONTENTS PAGE LIST OF TABLES . . . . . . . . . . . . . iii LIST OF FIGURES O O O O O O O O O O O O 0 iv INTRODUCTION. . . . . . . . . . . . . . 1 REVIEW OF LITERATURE I O O O O O O O O O O 6 DeveloPment of Syntactic Maturity . . . . . . 6 Syntax and Readability. . . . . . . . . . 9 Reader Perceptions and Attitudes . . . . . . l4 Listener Perceptions and Attitudes. . . . . . l7 METHODOLOGY 0 O O C O O O O O O O O O O 2 3 Subjects . . . . . . . . . . . . . . 23 Materials . . . . . . . . . . . . . . 24 Procedure . . . . . . . . . . . . . . 27 Compilation of Data. . . . . . . . . . . 38 Results and Discussion. . . . . . . . . . 28 Significance of Findings . . . . . . . . . 41 Implications . . . . . . . . . . . . . 42 Suggestions for Further Study . . . . . . . 43 Conclusion. . . . . . . . . . . . . . 44 APPENDICES Appendix A. The Narratives. . . . . . . . . . . 46 B. The Attitude Scale . . . . . . . . . 50 BIBLIOGRAPHY O O O O O O O 0 O O C O O O 5 1 ii LIST OF TABLES Table Page 1. Subject Comparison . . . . . . . . . . 29 2. Selection Comparison . . . . . . . . . 32 3. Combined Mean Scores for Main Groups . . . . 35 4. Number and Percentage of Items Marked by Main Groups. . . . . . . . . . . . 35 5. Percentage of Items Marked by Subgroups . . . 36 iii Figure l. 2. LIST OF FIGURES Subgroup Mean Scores Per Item . Main Group Mean Scores Per Item iv Page 33 34 INTRODUCTION "He read it, but he did not understand it." "He could not read his own writing." "Students aren't literate these days: they can't spell."1 Comments like the above are not uncommon these days. They merely reflect a problem that lay nonacademic and researchers alike encounter in trying to define the ambiguous term "reading." Newspaper editorials bemoan the rise of illiteracy in America; parents indict the public schools for failing to teach basic reading skills; teachers blame parents for allowing television to numb their children's desire to read. Clearly, public attitudes express a desire for children to learn to read and read well, but what does that learning involve? With his masterful work, The Psychology and Pedagogy of Reading, E. B. Huey laid the groundwork for research into the reading process. Yet, since the book's publication, studies regarding the mental aspects of reading have uncovered little more than Huey did in 1908. Reading instruction in this century has primarily treated reading as a perceptual process and focused on word processing; characteristic eye lMichael Stubbs, Language and Literacy (London: Routledge and Kegal Paul, 1980), p. 11. l movements, letter shapes, and word gestalts.2 However, is reading merely a matter of stimulus and response? When people are said to be reading, are they engaging in a process of decoding sound units from written symbols, or are they engaging in a process of understanding meaning?3 Can they accomplish one without the other? As Michael Stubbs observed, many studies have centered on the mechanical aspects of reading, but research is needed to explore the nature of the relationship between reader and author. Students are often instructed in how to decode the graphic information in a text, but are not as often instructed 4 in the derivation of meaning from their decoding. As a result, Stubbs observed, two views on the current state of reading instruction are now widely accepted: 1. Young children often have difficulty in understanding the purposes of written language, since many of these purposes are completely beyond their needs and experi- ence. 2. Young children will have particular difficulty in learning to read if they grow up in a home or cultural background with no tradition of literacy and hence no appreciation of the purposes of written language. With these views in mind, this study sought to deter- mine whether readers do indeed consider aspects of the reading process beyond graphic decoding. Specifically, the experiment's 2 3 Ibid., p. 5. Ibid., p. 11. 4 5 Ibid., p. 99. Ibid. purpose was to explore the perception of the author derived from features of his or her prose style. Before describing the design of this study, however, a few remarks will be made regarding the model of reading on which it is based. In considering the relationship between reader and author, this paper is framed by a psycholinguistic perspective in which reading is considered a cognitive process, a guessing game.6 Kenneth Goodman's three-step model of reading involves prediction, sampling, and confirmation. By using graphophonic, syntactic, and semantic cues,7 readers strive for the ultimate goal in reading: "the reduction of uncertainty."8 Because of the redundancy found in English, good readers use all infor- mation, visual (graphic) and non-visual (knowledge) to find 9 meaning. As Frank Smith observed, the more readers know about language, the author, and the subject, the better they will read.10 6Kenneth S. Goodman, Reading: A Psycholingustic Guessing Game (International Reading Association Conference Proceedings, XIV, 1972). 7Kenneth S. Goodman and Olive Niles, Reading: Process and Program (Champaign: Commission on the English Curriculum, National Council of Teachers of English, 1970), p. 41. 8Frank Smith, Understanding Reading (New York: Holt, Rinehart and Winston, 1971), p- 185. 9Frank Smith, Psycholinguistics and Reading (New York: Holt, Rinehart and Winston, 1973), p. 6. lOIbid. The reader's reliance on non-visual information relates to the purpose of this study: exploration of the reader's perception of the author. The experiment is based on the premise that an author's prose style can reflect certain aspects of his or her background and linguistic ability. The viability of this assumption will be discussed in the next section, but at this point, it suffices to say that a key question in this study is not whether the reader accurately perceives the author's persona, but whether the reader relies on particular clues to perceive anything about the author at all. To generate a working hypothesis, the study originally sought to sample twenty-six ninth grade students labeled by their previous reading teacher as "below grade level readers." Two narratives, approximately 150 words long, were constructed with syntactic complexity as the controlling variable. Each narrative was written in two versions: a simple version, con- sisting of short simple and compound sentences; and a complex version, consisting of simple, compound, and complex sentences. After reading a simple version of one narrative and a complex version of the other, students were to be instructed to rate the authors in terms of intelligence, age, and level of edu- cation. The working hypotheses at this stage: 1. Readers can perceive a difference in authors due to a contrast in syntactic complexity in the two versions. 2. Readers will rate the author of the complex narrative as older, more intelligent, and more educated than the author of the simple narrative. The study was piloted using two sets of narratives and thirty-five eighth grade students enrolled in "Develop- mental Reading." All of the students were termed "below grade level readers." Results of the pilot were inconsistent and no distinct pattern of reader perceptions was observable. The experiment was next conducted with two "above average" high school readers who confirmed the hypotheses. The working hypotheses were then revised: 1. Readers can differentiate between authors on the basis of syntactic cues. 2. Readers form attitudes regarding an author's back- ground by observing features of the structure of his or her writing. 3. Good readers are more consistent in forming these attitudes than are poor readers. These proposals were then revised into one hypothesis: Good readers, in contrast to poor readers, will, on the basis of syntactic cues, more consistently rate authors using complex syntax as older, more intelligent, and more educated than authors using simple syntax. This hypothesis was tested and the results of that experiment will be discussed later in this paper. In the next section, studies pertinent to this topic will be reviewed. REVIEW OF LITERATURE Few if any studies have explored the nature of the reader's perception of the author through the use of syntactic cues, but there has been research into related areas. These findings will be divided into the following topic areas: 1. Development of Syntactic Maturity 2. Syntax and Readability 3. Reader Attitudes and Perceptions 4. Listener Attitudes and Perceptions Development of Sygtactic Maturity Because the relative complexity of an author's syntax is the controlling variable in this study, it is important to consider what significance can be attached to a writer's complex syntactic structures. Is an intricate syntax a true measure of a writer's skill? Two studies of writing matura- tion discussed this question. Kellogg Hunt investigated differences in syntactic structures written by students who varied in chronological maturity and mental ability. When asked to rewrite a passage constructed in extremely short sentences, the students "exhibited the same general characteristics they exhibited in their own original compositions." 11 at various grade levels, Hunt concluded: As writers mature they take advantage of more and more Opportunities to consolidate sentences. . . . Older writers tend to produce writing that is affected by their synEgctic ability, not just by what they have to say . . . Hunt concluded further: As schoolchildren mature mentally, they tend to embed more of their elementary structures. Perhaps this is stylistic imitation, but a more attractive explanation is that as the mind matures, it forms more intricate sentences.l3 - Hunt's conclusion regarding writing "maturity" may significant in the setting of his experiment, a classroom, but After testing students be writing also reflects an apprOpriate purpose. A writer's use of complex syntax may indicate his maturity, but may also be a reflection of his audience. A mature author writing for an audience of elementary schoolchildren may choose a syntax more OWIl . able more than suitable to their level of competence than his or her Likewise, elementary school students, while not likely to alter their syntax to an adult level, may choose a formal structure when writing in a classroom environment they would choose outside of school. Nevertheless, Hunt's conclusion is relative to this study in that, to some and Adults (Monographs of the Society for Research in Child 11 Kellogg Hunt, Syntactic Maturity in Schoolchildren Development, XXXV, February 1970), p. 21. 12 13 Ibid., p. 53. Ibid. degree, it suggests that a writer's choice of syntax may be a clue to his or her level of chronological and mental maturity. W. L. Smith and George Mason sought to confirm Hunt's findings by instructing students in grades 4, 8, and 11 to rewrite a passage written in short sentences. Displayed below are samples of what the authors term as "typical eighth grade writing" (A) and "typical fourth grade writing" (B): A. The bee protects his family by stinging his enemies with his poisonous stinger which has hooks on it. As the bee pushes the stinger into the skin, the hooks dig into the skin, and when the bee flies away, the tightly holding hooks pull the stinger out of the bee. The little muscles in the stinger move, pushing the stinger which goes deeper into the skin. The muscles then squeeze the poison out. The stinger causes a sudden pain, and the poison causes the skin to swell. If the stinger is scratched off quickly, not much poison gets into the skin, and the pain and swelling are less. B. The bee protects his family. He uses his stinger, and he stings his enemies. The stinger has hooks on it, and it contains poison. The bee pushes the stinger into the skin, and the hooks dig into the skin and when the bee flies away, the hooks hold tight, and the stinger pulls out of the bee. Muscles are in the stinger, and they are little, and they move and push the stinger, and the stinger goes deeper into the skin, and the muscles squeeze the poison out. The sting causes pain, and the pain is sudden. The poison causes the skin to swell. The stinger should be scratched off quickly, and then not much poison gets into the skin and the pain is less, and the swelling is less. 4 Smith and Mason observed that while the eighth grade version is more sophisticated, it actually contains fewer 14W. L. Smith and Geoge E. Mason, "Syntactic Control in Writing: Better Comprehension," Journal of Reading, XV (1972), 356. words. This they attributed to a tendency in older students toward consolidation and embedding. Thus, they concluded, as did Hunt, that as students mature, their writing becomes more intricate. Syntax and Readability If, as Hunt suggested, the syntax of writers increases in complexity as they mature, the question basic to this study then arises: Do readers perceive complex syntax as a sign of maturity in an author? Studies regarding the effect of syntax on readability may provide an answer. Several experiments have suggested that transforming the syntax of a text from simple to complex increases the difficulty of the reader's task of comprehension. A. D. Marcus found that complex sentences containing an interruption of the subject-verb-object sequence by a relative clause are more difficult to comprehend than sentences in which the relative clause does not interrupt the sequence.15 W. T. Fagan concluded that sentence difficulty depends on the difficulty of its transformations, due most probably to redundancy.l6 In separate studies, R. C. O'Donnell and C. H. Rinne found a moderate correlation between awareness of structural 15Barbara Takahashi, "Comprehension of Written Syntac- tic Structures by Good Readers and Slow Readers" (unpublished Master's thesis, Rutgers University, 1975), p. 22. l6Ibid., p. 23. 10 relationship words in sentences and the reading comprehension of high school students.17 John Dawkins examined basal readers to see if syntax labeled "easy-to—read" truly was. He discovered examples like the following that claimed to be "simple," but which were really not effective in developing comprehension skills because of their unnaturalness: "See the dog. The dog is big. The dog is Susan's."l8 As Dawkins observed, this example does not reflect normal discourse. Truly easy-to—read syntax will be clear so it will not violate old rules of clarity as oversimplistic syntax clearly does. Because oversimplistic syntax violates basic principles of clear writing, it would be fair to conclude that it is actually difficult to process.19 Dawkins concluded further that if, as Frank Smith and Kenneth Goodman suggest, comprehension involves prediction and confirmation, then . . . clear writing will give the reader clues for pre— dicting correctly. Therefore, writing that violates convention is hard to read because it interferes with prediction making. W. L. Smith sought to determine if it was beneficial to provide poor readers with simplified syntax to aid their comprehension. Fourth, tenth, and eleventh grade students 17J. W. Schneyer, "Research: Syntactic Structures and Reading Comprehension," Reading Teacher, XXIII (1970), 467. 18John Dawkins, Syntax and Readability (Newark, Delaware: International Reading Association, 1971), p. 67. 19 20 Ibid., p. 71 Ibid. 11 were given samples of fourth, eighth, and eleventh grade writing to read. The author found that while eleventh graders comprehended significantly better than fourth graders at all three levels of writing, they actually comprehended eleventh grade writing best. Smith explained this result by noting that eleven grade writing, while more complex in syntax than fourth grade writing, more closely approximated the level of sentences that the older students were used to producing and encounter- ing, and therefore, was easier to process lingustically than simple sentences.21 Eunice Schmidt wrote a simple and a complex version of The Helen Keller Story and tested their readability by several methods, including cloze procedure. Her results indicated that although her simple version contained more words per T-unit, the complex version was more difficult to comprehend because its syntax was less predictable and less redundant.22 Barbara Stoodt found a positive correlation between understanding grammatical conjunctions and reading compre- hension. She defined conjunctions as "empty words whose 21W. L. Smith, "The Effect of Transformed Syntactic Structures on Reading," Paper presented at the International Reading Association Annual Conference, May 6—9, 1970, p. 9. 22Eunice L. Schmidt, "What Makes Reading Difficult: The Complexity of Structures," Paper presented at the Annual Meeting of the National Reading Conference, December 1-3, 1977, p. 7. 12 significance lies in the relationships they signal within sentences and between sentences."23 The embedding of relative clauses was another factor in readability found in a study by P. David Pearson. He cited E. B. Coleman's work which provided two sample sentences: 1. The rat that the cat killed ate the malt. 2. The cat killed the rat that ate the malt. Coleman concluded that highly embedded relative clause forms (1) were harder to read than less embedded forms (2).24 A number of studies considered the differing effects of syntactic complexity on good and poor readers. Ethel Young gave eighth and ninth grade students reading three and four years below grade level a series of syntactic structures from A Test of Sentence Meaning (ATSM). Young's results indi— cated significant differences in the group's comprehension of eight syntactic structures and led her to conclude that some syntactic structures are more difficult to comprehend than others.25 23Barbara Stoodt, "The Relationship Between Under- standing Grammatical Conjunctions and Reading Comprehension. Final Report" (Columbus: Columbus Research Foundation, September 1970), p. 54. 24P. David Pearson, "The Effects of Grammatical Com- plexity on Children's Comprehension, Recall, and Conception of Certain Semantic Relations," ReadinggResearch Quarterly, x (1974-75), 156. 25Ethel E. Young, "Comprehension of Syntactic Struc~ tures When Presented in Visual and Auditory-Visual Modes to a Selected Group of Adolescent Disabled Readers" (unpublished doctoral dissertation, Michigan State University, 1977), p. 247. 13 Barbara Takahashi tested three groups of readers (good sixth grade readers, good ninth grade readers, and poor ninth grade readers) to test W. L. Smith's conclusion that there is no apparent advantage in simplifying syntactic structures in materials for older students reading below grade level. Good sixth grade readers were chosen for the study because the mean comprehension scores for her sample of poor ninth grade readers fell at the sixth grade level. Good ninth grade readers were included in the study to provide a comparison with the poor ninth grade readers. After adminis- tering the ATSM to each group, Takahashi found that while good sixth grade readers and poor ninth grade readers performed quite similarly on a standardized reading test, they scored differently on the ATSM. She attributed this discrepancy to the possibility that poor ninth grade readers may rely more on redundancy and overall contextual clues during reading, while good sixth grade readers may rely more heavily on com- prehending each individual sentence, "a task more comparable to that performed on the ATSM."26 Takahashi found five structures that caused difficulty in comprehension: l. Interruption of the subject—verb-object sequence Pronoun reference Deletions Embedding ConjunctionsZ7 U1u¥>bJN O. 26Takahashi, "Comprehension of Written Syntactic Structures," p. 56. 27Ibid., p. 53. 14 and concluded that older students with reading problems develop strategies dependent on redundancy in the flow of language. The research cited in this section indicates that certain syntactic structures affect reading comprehension, although studies by W. L. Smith, Dawkins, and Takahashi indi- cate that older readers actually find simpler syntax harder to read because it seems unnatural. Because this study involved the construction of simple and complex versions of a narrative, and because syntactic complexity has been shown to be a factor in readability, the relative difference in syntactic complexity of the two versions should signal a recognizable difference to the reader. Whether the reader will perceive that difference as a reflection of the author's background will be a matter discussed along with this study's results. Reader Perceptions and Attitudes Although a number of studies have been conducted on reader attitudes toward reading, relatively few experiments have focused on the reader's attitude toward and relationship to characters, the author, or the author's prose style. I. A. Richards considered the reader as critic in his 28 study of reader reactions to anonymous poems. His focus 281. A. Richards, Practical Criticism (New York: Harcourt, Brace and World, 1929). 15 on the reader's role in the communication process, parti- cularly the reader's role in the communication process, particularly the reader's responsibilities to close reading and expression of feelings, predated the work of Louise Rosenblatt. Like Richards, Rosenblatt centered on the reader's performance and, calling on the work of John Dewey and Arthur F. Bentley, defined the "transactional theory of literature" as an ongoing process in which the elements or factors are . . . aspects of a total situation, each conditioned by and conditioning the other.29 Rosenblatt's reader not only experiences the work itself as an event, but may find that "the text may yield glimpses of the personality and (moral) codes of the author."30 Percep- tion of the author, Rosenblatt contended, is key to compre— hension: In order to decode the message (the reader) must also recreate from the text a speaker or a persona, and sometimes behind that an author.3 Therefore, a reader does not have to accurately infer the author's background or personality, but instead may infer the persona that the author has projected. A study by Janet Cochran examined the reader/author relationship and its various components. In Cochran's View of the reading-communication process, "writer and reader 29Louise M. Rosenblatt, The Reader, the Text, the Poem (Carbondale: Southern Illinois University Press, 1978), p. 17. 3OIbid., p. 56. 3lIbid., p. 76. 16 fictionalize each other according to inferences each has 32 Cochran concluded that choices are made based on made." common characteristics such as culture, education, syntactic repertoires, and values, and that those choices help form a convenant between the author and his or her intended audience.33 Reader attitudes toward authors was the subject of a study by Anne McKillop. Eleventh grade students were given three passages that concerned the same topic. One passage was favorable to the topic, one unfavorable, and one neutral. After reading each passage, students rated each author by using an adjective checklist. McKillop found that reader attitudes toward the authors were affected by the author's perceived treatment of the tOpic.34 A study by Robert LeBouef and Marc Matre on reader perceptions of story characters presented a model testing reader attitudes. Subjects read magazine short stories and rated characters using an attitude scale that spanned three categories: 1. Evaluation (good/bad, likeable/unlikeable) 2. Dynamism (bold/timid, active/passive) 3. Temperament (relaxed, tense, simple/complex) 32Janet F. Cochran, "Between Writer and Reader: The Relationship of the Concept of Audience to the Teaching of Composition" (unpublished doctoral dissertation, University of North Carolina at Greensboro, 1979). 33Ibid. 34Anne 8. McKillop, "The Relationship Between the Reader's Attitude and Certain Types of Reading Response" (unpublished doctoral dissertation, Columbia University, 1951). 17 LeBouef and Matre concluded that reader attitudes were measurable and that those attitudes reflected variations in the subjects' age, sex, and socio-economic status.35 Listener Perceptions and Attitudes Comparing the reactions of readers to the reactions of listeners may seem logical. Both listening and reading are receptive language processes; both require a knowledge of syntax and semantics; and both require an interpretation by the receiver. Yet, as Frank Smith has pointed out, written and spoken language exist independently,36 and therefore, any comparisons drawn between listener and reader perceptions must be considered in that light. Rather than strain to correlate the findings of this study with studies of listener perceptions, this paper will review such experiments as a guide to methodology. An important study in listener perceptions of speakers was conducted by G. Richard Tucker and Wallace E. Lambert. Black and white college students listened to samples of taped speech of representatives of six American-English dialect groups. The subjects then rated characteristics of the indi— vidual speakers by using an adjective checklist. This rating scale was specifically developed for the study's population 35Robert LeBouef and Marc Matre, "How Different Readers Perceive Magazine Stories and Characters," Journalism Quarterly, LIV (1977), 56. 36Smith, Psycholinguistics and Reading, p. 71. 18 and was based on to what degree the listeners believed that the Speaker was "friendly" and had achieved "success."37 Among the adjectives presented to be rated on a 1-8 scale: 1. Intelligent 2. Friendly 3. Educated 4. Trustworthy 5. Ambitious 6. Honest 7. Considerate38 The purpose of the study was to determine if subjects could differentiate dialect variations using only voice characteristics and speech style as cues and, if so, if patterns would emerge indicating dialect preferences. Tucker and Lambert concluded that subjects were able to differentiate dialect variations and that white and black judges showed different perspectives in evaluations of the least favorable of the dialects.39 Bruce Fraser conducted a modified version of the Tucker-Lambert study in which he sought to verify the original study's two basic questions and answer one of his own: "Does the subject's judgment of the speaker's race correlate with his overall evaluation of the speaker?" Fraser's experiment 37G. Richard Tucker and Wallace E. Lambert, "White and Negro Listeners' Reactions to Various American—English Dia- lects," Advances in the Sociology of Language Volume II, ed. Joshua Fishman (The Hague: Mouton, 1972), p. 176. 39 381bid., p. 177. Ibid., p. 183. 19 followed the same basic method as the one conducted by Tucker and Lambert, but Fraser used only nine of the original fifteen adjectives for his rating scale.40 Fraser's results showed that when the speaker was black, and when subjects misjudged the race of the speaker, subjects tended to rate the speaker lower in all categories. When the speaker was white, and his race misjudged, subjects did not tend to rate the speaker lower than average. Fraser explained that the misjudgers in his study were not rating low, but that the correct judges were rating high. He con- cluded: People will judge differentially on the basis of certain cues--in this case speech alone——because of their experi- ence and certain, albeit inaccurate stereotypes. Two other studies of dialect attitudes and stereotypes focused on teacher reactions. The first, conducted by Fred Williams, used a semantic differential scale (fast/slow, friendly/unfriendly). Williams tested teacher reactions and evaluations of children's speech and concluded that persons tend to employ stereotyped sets of attitudes as anchor points for their evaluation of whatever is presented to them as a 42 sample of a person's speech. Orlando Taylor employed the 40Bruce Fraser, "Some 'Unexpected' Reactions to Various Innerican-English Dialects," Langugge Attitudes, eds. Roger Shuy euad Roger W. Fasold (Washington: Georgetown University Press, 11973), p. 29. 4lIbid., p. 35. 42Fred Williams, "Dialect Attitudes and Stereotypes," Eéllguage Attitudes (Washington: Georgetown University Press, 1973), p. 126. 20 the Language Attitude Scale with five gradation to study teacher attitudes toward Black English. He found that Teachers do not appear to have a single, generic attitude toward dialects, but rather, differing attitudes depending on the particular aspect of dialect being discussed. Taylor noted an overall trend for most teachers to reveal positive to neutral Opinions toward certain dialect features.44 Williams collaborated with Roger Shuy on another study of dialect stereotypes. Using a semantic differential scale, the researchers characterized their ratings into four cate— gories: 1 2. Complexity (easy/difficult, simple/complex) 3. Potency (strong/weak, sharp/dull) 4. Activity (fast/slow)45 Value (good/bad, smart/dumb, smooth/rough) Subjects rated the speech characteristics of repre- sentatives of five dialect groups: . Detroit speech White Southern speech British speech Negro speech Standard speech46 UlubLONl-J O O O 0 43Orlando L. Taylor, "Teachers' Attitudes toward Black éand Nonstandard English as Measured by the Language Attitude £3ca1e," Language Attitudes (Washington: Georgetown University Press, 1973), p. 126. 44Ibid. 45Fred Williams and Roger Shuy, "Stereotyped Attitudes Of? Selected English Dialect Communities," Language Attitudes (Wkishington: Georgetown University Press, 1973), p. 88. 46Ibid., p. 86. 21 Williams and Shuy found that "higher status respondents tended to rate British speech more positively in terms of value and 47 However, lower status respondents rated Detroit 48 potency." and Negro speech more positively in those areas. The author concluded: Dialect may have an objective reality in the way pe0ple talk, but it seems quite clear that it at the same time has a subjective reality in the kinds of consistent attitudes which peOple hold towards another one's speech. An experiment by J. R. Edwards entitled, "Students' Reactions to Irish Regional Accents" is noteworthy in that the five dialects presented on tape to the listeners were all read by the same actor. Similarly, each narrative version in this study was written by this researcher. Edwards' results indicated that the subjects did not discern this device when rating the speakers.50 Other studies that considered listener attitudes include Alberto Rey's survey that revealed evidence of job discrimination due to employer prejudice towards differing dialects51 and a preliminary report prepared by Leslie Palmer which quoted a study by Lambert and Anisfield: 47Ibid., p. 91. 481bid. 491bid., p. 95. 50J. R. Edwards, "Reactions to Irish Regional Accents," Language and Speech, XX (1977), p. 281. 51Alberto Rey, "Accent and Employability: Language Attitudes," Language Sciences, XLVII (October 1977), ll. 22 Subjective evaluations of speech are systematically affected by associations made with the stereotypes held about majority and minority groups. A majority of the studies cited in this section have concerned the effect of stereotypes on listener judgments of speakers. The focus upon certain features in speech dialects and their role in the perception of the listener may possibly be correlated to the use of graphic cues by readers to form attitudes toward the author, but the fact remains that a writer's style may by no means parallel his speaking style. Rather than equate reader and listener, this paper has sought only to discuss pertinent aspects of the reviewed studies. The major finding in this review has been the use of attitude scales. Research in this section has supported the notion that listeners do indeed form measurable opinions about speakers based on features of oral language. Whether readers similarly form measurable attitudes about authors based on syntactic cues is the question investigated by the present study. 52Leslie Palmer, "A Preliminary Report on a Study of iihe Linguistic Correlate of Raters' Subjective Judgment of I“on-Native English Speech," Language Attitudes (Washington: (Eeorgetown University Press, 1973), p. 41. METHODOLOGY Subjects The study's population consisted of 104 high school students, grades 9-12, enrolled in two courses: "Developmental Reading" (DR) and "Speed Reading" (SR). The DR group, numbering 16 males and 12 females, ranged in age from 14 to 18. Of the 28 students, 26 were in ninth grade, 1 in tenth, and l in eleventh. A11 respondents in the group were white and ranged socio-economically from middle to lower class. Each student had been referred for the course by their eighth grade reading teacher, who had labeled all twenty-eight students as "below grade level readers." Reading habits among the group varied, but while a small number read books outside of class, the large majority read nothing more literary than magazines. In nearly every student's case, writing skills needed much improvement. The SR group, numbering 34 males and 42 females, ranged in age from 15 to 18. Of the 76 students, 51 were in tenth grade, 18 in eleventh, and 7 in twelfth. All respon— dents in the group were white and fell into the middle class socio—economically. "Speed Reading" is an elective course, with attainment of a B- or better grade in a high school 23 24 literature course as its only prerequisite. Not all SR students were "good" readers, but nearly all were above average in comprehension. Outside reading of books was a requirement in the course, and a large majority of the students exceeded the minimum established. Individual writing skills for nearly every student were considered average to above average for their respective grade levels. Materials Testing material consisted of two narratives, each divided into a simple and a complex version, and an attitude scale. The narratives ranged in length from 152 to 160 words and shared the subject matter of teenage problems (Appendix A). The simple versions consisted of short simple and compound sentences, with a vocabulary designed to facilitate compre- hension by all subjects. The complex versions maintained the vocabulary and approximate word order of their simple counter- parts, but introduced left branching of relative clauses and modifier phrases. The significance of this variable merits an explanation of its use. The term "left branching" derives from the tree diagram used in generative grammar in which the subject noun phrase (SNP) occupies the left branch and the verb phrase (VP) occupies the right: 25 S (sentence) /\ SNP VP l | John lives Left branching occurs when a clause is embedded on 53 study showed, this embedding the left branch. As Coleman's delays completion of the subject-verb-object sequence, and hence makes the sentence more difficult to comprehend. For this reason, left branching was chosen as the syntactic difference in the narrative selections. An example from Selections 1 and 2 (left branching element underlined): 1. Jim slammed on the brakes. The car skidded on the wet pavement and struck the child. 2. When Jim slammed on the brakes, the car skidded on the wet pavement and struck the child. In each sentence they process, readers seek to dis- cover who or what (subject) is performing what action (verb). In #1, the reader learns quickly and simply that "Jim slammed on the brakes." In the next sentence, the reader learns that "the car skidded on the wet pavement and struck the child." In both cases, there is little or no distance between the subject and its verb. The main information in the sentence is therefore presented initially to the reader. Modifiers may follow on the right branch, the verb phrase, but they merely provide details further defining the action. 53Pearson, "Effects of Grammatical Complexity," p. 156. 26 In #2, on the other hand, the subject and verb of the main clause ("the car skidded on the wet pavement and struck the child") are delayed by the introduction of a relative clause ("when Jim slammed on the brakes"). This delay causes the reader to store the information from the relative clause until he or she finds the subject and verb of the main clause, and therefore complicates the task of comprehension. Left branching, as Hunt observed, can also signal a maturity on the author's part. Consider an example from selections 3 and 4: 3. She would probably scold him. He walked up the stairs slowly. 4. Knowing she would probably scold him, he walked up the stairs slowly. The passage from #3 consists of two simple sentences. The author uses a straightforward subject-verb pattern in pre- senting the action. Yet, it is the job of the reader to connect Carl's slow walking with the knowledge of his impending scolding. The author of the passage from #4, however, shows an understanding of that relationship by placing the relative clause ("knowing she would probably scold him") before the main clause ("he walked up the stairs slowly"). The second author diSplays a style that relies more on transitions and combinations of actions than does the style of the first author, and this added complexity may be perceived by the reader as a sign of the author's linguistic skill. Because left branching creates difficulty in compre- hension, and because readers may perceive the complexity that 27 it forms as a sign of the author's maturity, then its use as the controlling variable in the syntactic difference between simple and complex selections is apprOpriate. The attitude scale used in this study consisted of six questions. Each item asked the respondent to rate the author on a 1-5 basis, with the choices on each scale designed to suit the population. Three major areas were tested by the attitude scale (Appendix B): A. Intelligence (Item 4) B. Level of education (Item 5) C. Age (Item 6) These areas were designed to elicit the readers' Opinions regarding the author's background. Three other areas were included to elicit attitudes toward the writing and to judge the consistency of the subjects' responses: D. Appropriateness of story for readers' age level (Item 1) E. Ease of comprehension (Item 2) F. Author's understanding of readers' age level (Item 3) Procedure Testing was conducted during the subjects' regular class periods. Subjects were given packets containing the simple version or one narrative, a rating scale, the complex version of the other narrative, and a second copy of the rating scale. Each group was read the following directions: This is a study designed to see what you can guess about an author's background by reading his or her writing, Please read the first story and answer the questions on the page that follows it. DO the same for the second story. 28 Guidance was not Offered during the reading of the narratives, but some subjects were assisted in understanding the format of the attitude scale. NO questions were permitted regarding the identity of the "authors." Compilation of Data The raw scores per item for each subject were compiled in a table called "Subject Comparison" (Table 1). Abbreviations in the table include "Sub" (Subject) and "Sel" (Selection). The mean scores per item for each subgroup were next cate- gorized by selection in a table called "Selection Comparison" (Table 2). Subgroups were identified by the main group to which the subject belonged and the selections that he or she read: SR (1-4), DR (1-4), SR (2-3), and DR (2-3). Data from Tables 2 and 3 were then transferred to bar graphs that illustrated subgroup mean scores per item (Figure l) and main group mean scores per item (Figure 2). The final set of tables lists the number and percentage of subjects who rated the complex author higher, equivalent to, or lower than the simple author for each item. These scores were arranged by main groups (Table 4) and by subgroups (Table 5). In Table 5, a higher score is represented by +, and equivalent score by 0, and a lower score by -. Results and Discussion An examination of Table 3 revealed that in the com- bined mean scores for simple and for complex authors, the 29 Table l.--Subject Comparison. Sub Sel l 2 3 3 4 5 6 2 Sub Sel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able 1.--Continued. 4 5 Sub Sel l 2 3 2 3 4 5 6 Sub Sel l 5254544343423232333432443433433234.223344453115433 423443553343233233323243343333332433334434323331 43444.3554433433233333355444333334533344534424351 423433454434443443444345334332334334335441424335 111211111121422223211333121211112121132113321115 432343354433433343434255323332334333434343324334 23142314231414231423142314.1423141423142314231423 3 7 l 5 7 7 77* 78* 79* 80* 81* 82* 83* 84* 85* 86* 87* 88* 89* 90* 91* 92* 93* 94* 2 4 6 7 7 7 424234445455243332335434322544433354333422233544 424234323344232332334434312343433344333422233334 53433333334334333334443444334343344333333344344.4 4.43444322244444434344324441154444433342222343444 21.21232223111111211321121211111.22le2112311212122 525233331245343333344234331142435433323333243435 231414231423141423142314231423142314231433142314 789012345678901234567890 444555555555566666666667 31 Table 1.--Continued. 2 3 4 5 6 Sub Sel l l 2 3 Sub Sel 21425222432343454532 21423122323343443344 42434433332333444343 42334422331434342243 21225531111211211252 42334332431334343152 23231423231414141414 95* 96* 97* 98* 99* 100* 101* 102* 103* 104* *DR 32 Table 2.-—Selection Comparison. Sel Sub 1 2 3 4 5 6 1 1-76 2.85 1.55 3.25 3.42 3.10 3.42 77-104* 3.50 1.68 3.37 3.43 3.06 3.43 2 1-76 3.72 1.52 3.61 3.66 3.55 3.88 77-104* 3.50 2.00 3.41 3.66 3.08 3.16 3 1-76 3.19 1.33 3.47 3.36 2.94 3.25 77-104* 2.91 1.75 3.25 2.91 2.25 2.58 4 1-76 3.50 1.65 3.77 3.62 3.57 3.70 77-104* 2.81 2.12 3.12 3.43 3.12 3.25 *DR 33 [3 Simple os.m7//VVA//////Vw///// 6 2.21 1 2.2 _ 36 71,1. // WM 1,, V 1/ WV 1, .. . _, 6 $.m7/////////////A/Cs was .1 5 36m mm.ma.//_.///V/////V/: ~e.m_fl///////////////// /4 62m _1 1 1 4 Nvfl $6 7/ ,., / .F/ / / // / / // 1:.m7/ $323??? fiy 3 1 :14 . . 1 3 1.. 3.: HS.n//////:///:,/;. u SAWE2 ( mm; .2 a .3 q. 32 SSE: m.m7//////////////// 3km. me: 1 2.35:3ZZAIZA: 1 n h n x 1 A l 1 m~.mE/VA//////VJ: 6 3.2 1 21m— 3.m7/V////r/V//,6 2.2/VV/J/S/V/V/SM5 2.: s WOoMF meomg/ / 2/ / // M / / / 21m /////////D//1/// 3.2 4 .41. 21m 1367/2 / / / / 2w /A 4/ / / m 2.2x/131/V/5/2/6. .. s m 2.5 3 R as. m < s.m_////3./1_,///// / m n 2&7/fl/jx/Cul u m 2.1/ 1 2 m. $.1 0&7 ,/ x m sifi/fiii /, 521 3 mi, 1 1 , . 1 m.m_ om.m T ,.r/ ,1 ., / ., 1,, / / /., . . _ w p b p _ 4” L. r 1 1.. a. .H. 11 mucflom mamom mtsufluud Subgroup Mean Scores Per Item. Figure 1. 34 1. ////////», @m— 86 m. m. 36 mp.~/_//////// mum ////MA////// 3.11 o~.m / / /1/V/VV / Hm.m DR 699V//////, RAF om.m as ///////// _ p _ fi _ _ 4 3 2 l mucflom mamom mosufluum C—P- Items .2 //////////1 mm.m .27////////// s.§//////////3 mm.m .ZV/////////// mm.m — mm; ////, _ ¢¢.H 3.1//////////5 No m“ Main Group Mean Scores Per Item. "‘ .1— — fib Figure 2. 35 Table 3.——Combined Mean Scores for Main Groups. 1 2 3 4 5 6 Simple Selection (1-3) DR 3.20 1.71 3.31 3.15 2.65 3.00 SR 3.02 1.44 3.36 3.39 3.02 3.33 Complex Selections (2-4) DR 3.15 2.06 3.26 3.55 3.10 3.21 3.61 1.58 3.69 3.64 3.56 3.79 Table 4.--Number and Percentage of Items Marked by Main Groups. DR SR # % # % Item 1 + 9 32 + 41 54 0 6 22 0 22 29 - 13 46 — 13 17 Item 2 + 14 50 + 25 33 0 5 18 0 38 50 - 9 32 - 13 17 Item 3 + 8 28 + 28 37 0 9 32 0 38 50 - ll 40 - 10 13 Item 4 + 9 32 + 26 34 0 14 50 0 40 53 - 5 18 - 10 13 Item 5 + 12 43 + 39 51 0 12 43 0 30 39 - 4 14 - 7 10 Item 6 + 12 43 + 36 47 0 28 O 26 34 8 - 8 28 - 14 19 36 Table 5.--Percentage of Items Marked by Subgroups. DR(l-4) DR(2-3) SR(1-4) SR(2-3) Item 1 + 19 50 58 50 0 12 33 3O 27 - 68 16 12 23 Item 2 + 56 42 32 33 O 25 42 45 56 - 19 16 23 11 Item 3 + 19 42 45 27 0 38 25 52 48 - 43 33 3 25 Item 4 + 25 42 30 39 0 50 50 58 48 - 25 8 12 13 Item 5 + 25 75 50 53 0 56 25 40 39 - l9 0 10 8 Item 6 + 38 50 4O 56 O 19 42 32 44 - 43 8 28 8 37 DR group rated the complex author as older, more educated, and more intelligent than the simple author. The SR group did the same. As expected, the DR group rated the simple versions and the complex versions as more difficult to under- stand than did the SR group. When the mean scores were examined by subgroups, however, a significant discrepancy was found (Figure 1). SR (l—4)--that is, the SR students who read selection 1 and 4-—and SR (2—3)--those who read selections 2 and 3--rated the complex author higher for all six items, although the difference in Item 2, Ease of Comprehension, was marginal for both subgroups. (This instance may be explained by the nature of the narratives' construction. Few SR students found any one of the four selections difficult to read because they were written to suit the abilities of all students in the population.) DR (2-3) followed the same pattern as the SR subgroups; in fact, the difference in mean scores for complex and simple authors in this subgroup was greater than the difference in either of the SR subgroups. The discrepancy lay in the scores for DR (1-4), which rated the simple author as older, nearly equivalent in level of education, and equivalent in intelligence. Examination of Table 4 revealed that for Item 4 (Intelligence), 87 percent of the SR group rated the complex author higher or equivalent to the simple author. This compared to 82 percent of the DR group. For Item 5 (Education), 38 the figures were 90 percent for the SR group and 86 percent for the DR group. Item 6 (Age) showed 81 percent for the SR group and 72 percent for the DR group. Again, these are combined scores for the subgroups, and do not truly reflect the impact of the DR (1-4) responses. Table 5 did reveal that effect. For Item 4, 75 per— cent Of the DR (1-4) students rated the complex author higher or equivalent to the simple author; for Item 5, 71 percent; and for Item 6, only 57 percent. These scores would seem to indicate that DR (1-4) was the exception of the four sub- groups in the rating of authors. However, if one considers the working hypothesis which stated that good readers more consistently rate complex authors higher than do poor readers, then it was actually DR (2-3) that did not fit the expected pattern. The DR (2-3) students revealed a pattern of scores that more closely reflected the performance Of the good readers. In summary, both SR subgroups supported the working hypothesis by rating the complex author higher for each item. As a group, the DR students also supported the working hypothe- sis by displaying less consistency than the SR group. DR (2-3) was consistent within itself, but DR (1—4) was not, and the two subgroups were inconsistent when considered together. These results indicate that: 1. SR students perceived differences between authors on the basis of syntactic cues. 2. SR students were more consistent in their ratings than were DR students. 39 The inconsistency of DR responses is interesting and deserves investigation with further studies. There are some possible explanations for the finding, though no firm con- clusion can be drawn. Why were SR students as a group more consistent in their ratings than DR students were? One explanation may be that SR students were more mature in language ability. As more experienced readers and writers, they may have attended to particular syntactic cues of which DR students were not as keenly aware. The inconsistency between DR subgroups may be explained by relative differences in the construction Of the selections. The sharp pattern of response for DR (2-3) contrasted to the mixed results of DR (1-4) may indicate that, to DR students, Selection 2 seemed more complex in relation to Selection 3 than Selection 4 did to Selection 1. The fact that this discrepancy was not evident in the SR subgroup responses may be explained as a reflection of the better readers' ability to detect subtle syntactic differences. Another explanation for the DR inconsistency may be test—taking ability. Although the two sets of narratives were distributed in alternate fashion to each class, the individual students in DR (1-4) may have had difficulty in understanding the testing instrument and hence their responses may not have truly reflected their feelings and ideas about the authors. Students enrolled in "Developmental Reading" 40 often display impatience with the tasks of reading and test- taking, and this lack of concentration may also have altered the results. "Speed Reading" students, on the other hand, are more accustomed to taking tests, and the results of this study are therefore more likely to be accurate in their instance. Still another result to consider is that Of the relative ratings of the authors by the two main groups. Aside from Item 2, Ease of Comprehension, SR students rated the authors higher than did DR students in every instance. Although the major purpose of this experiment was to determine whether a difference exists between readers' perceptions of simple and complex prose styles and their authors, it is interesting to note that the DR mean scores were lower than the SR scores. SR students rated complex authors closer to "smart" than "average" in intelligence (3.64), closer to a college graduate than a high school graduate (3.56), and closer to the 21—30 age class than the 16-20 class (3.79). DR students also rated complex authors closer to smart than average (3.55), but closer to a high school graduate (3.10), and closer to the 16-20 age class (3.21). This difference may be attributed to a lack of experi— ence on the part of the DR group, a naivete regarding the nature of test construction. Another possibility is that SR students may have given higher ratings because they perceived the authors as professionals writing "down" to their audience, 41 while DR students viewed the authors as contemporaries. These explanations are, of course, a matter Of conjecture, but are reasonably based on the researcher's knowledge Of the subjects. Significance of Findings Although a number of experiments have investigated the effect that syntactic complexity has on reading compre- hension, few if any studies have explored the nature of the reader's perception of the author based on the use Of syntactic cues. This study has accomplished that, and is therefore significant in several ways. First, this study shows that "Speed Reading" students can and do differentiate between pross styles on the basis Of syntactic rather than content differences. The students may not have been aware that left branching was the only signifi- cant difference in the two versions read, but their ratings indicate that they perceived some difference. Second, the study shows that "Speed Reading" students form consistent attitudes toward authors on the basis of syntactic information. Both SR subgroups rated the complex author as older, more intelligent, and more educated than the simple author. This consistency in attitudes indicates an involvement with the text, and further, as Stubbs observed, that these students are aware of the purposes of language and that they realize that differences in prose style reflect differences in authors. 42 The third major finding is that differences exist between "Speed Reading" and "Developmental Reading" students. The significant difference in the two groups lies in the inconsistency Of responses by the DR group. The failure Of DR students reading Selections 1 and 4 to consistently dis- cern differences between the two authors suggests that one difference between good and poor readers may lie in the good readers' ability to use syntactic cues to form a concept of the author; these readers seem to be more aware of the role an author plays in designing prose than are the DR readers. This may suggest that good readers, more so than poor readers, see reading as communication between author and reader. Implications The major purpose Of this experiment was the study of reader reactions, not the development of relevant teaching strategies; however, at least two suggestions can be made on the basis Of the study's findings. First, it is possible that readers' attitudes toward authors may influence their motivation and involvement in their reading. To take advantage of this interest, teachers may integrate reading and writing in their class lessons. Students should benefit from writing exercises that require them to copy an author's style. For example, a class reading The Adventures of Huckleberry Finn might be encouraged to write first person accounts in dialect. The assignment may 43 help students better see the connection between writing and reading, author and reader. Secondly, because SR students in this study read more and read for pleasure, perhaps DR students could benefit from increased classroom reading time and encouragement to read for pleasure. The experience of reading a variety of authors may help poor readers become more involved in their reading and eventually more aware of the relationship between author and reader. Suggestions for Further Study The failure of DR students to form consistent atti- tudes toward authors suggests that further studies are needed to discover the reasons for this inconsistency as well as reasons for the differences in the ratings of the DR and SR groups. Individual interviews with respondents might reveal their reasons for rating one author higher than another. Other studies could investigate the effects of variables other than syntax. Lexicon may be a possibility in that case. Still another study, patterned after Kellogg Hunt's, might sample student writings and compare the student's syntactic maturity to his or her sensitivity toward syntactic differ- ences during reading. NO matter what direction future studies take, it is clear that more must be learned about the cognitive aspects of reading. We know that there are good readers, poor readers, and even indifferent readers. Research to date has readers' 44 inferential processes seldom focused upon and their parti— cipation in reading as a communicative act, both Of which may contribute to the disparity among readers. Conclusion This study has attempted to determine whether readers form attitudes toward authors on the basis Of syntactic cues. Attitudes measured by this study were perceptions of the authors' background and ability as defined by age, intelli- gence, and level of education. Subjects read versions of two narratives that differed significantly only in the complexity Of sentence structure as controlled by the introduction of left branching and rated the authors separately. Both "Speed Reading" subgroups supported the working hypothesis by rating the complex author higher for each item. As a group, the "Developmental Reading" students also supported the working hypothesis by displaying less consistency than the SR group. DR students reading Selections 2 and 3 were consistent within their subgroup, but DR students reading Selections l and 4 were not, and the two subgroups were inconsistent when con— sidered together. The following conclusions are suggested by the data gathered: 1. Good readers differentiate between authors on the basis of syntactic cues. 2. Good readers tend to rate an author who uses complex syntactic structures as older, more educated, and 45 more intelligent than an author who uses simple structures. Good readers are more consistent in perceiving author differences than are poor readers. APPENDICES APPENDIX A THE NARRATIVES 47 2 It was raining. When Jim tried the windshield wipers, they didn't help. Because the road was still too hard to see, he slowed the car down. The rain seemed to come down even harder. Jim had wanted to stay home tonight, but his mother had asked him to pick up his younger sister at the school play. At the school corner, Jim switched on his turn signal. He didn't see the girl crossing the rainy street until it was too late. When Jim slammed on the brakes, the car skidded on the wet pavement and struck the child. Quickly getting out Of his car, Jim ran around the front of it. Seeing the girl lying on the street, Jim was scared. He knelt beside the child. He heard a voice. "Jimmy, what happened?" It was his sister's voice, and she was walking toward him. Jim was relieved; but if his sister was all right, who was the child lying next to him? 48 3 Carl couldn't find his homework, and he was upset. He had completely cleaned his room. He had searched his books and folder. His math assignment was nowhere to be found. Last week he had forgotten to do his homework twice. His teacher had warned him that he would stay after school to make up the work. Carl had to find that assignment. Finally, Carl decided to ask his mother. She would probably scold him. He walked up the stairs slowly. He thought about the lecture she would give him. She would tell him to keep things in order. Carl reached the top of the stairway and turned the corner. He hung his head. His mother stood in the middle of the hallway with Carl's crumpled math assignment in her hand. Carl was saved; his teacher wouldn't be upset. But his mother was upset. She began to lecture. Maybe Carl wasn't saved after all. 49 4 When Carl couldn't find his homework, he was upset. DeSpite a complete cleaning of his room and a search of his books and folder, his math assignment was nowhere to be found. Last week, when Carl had forgotten to do his homework twice, his teacher had warned him that the next time he would stay after school to make up the work. Carl had to find that assignment. Finally, Carl decided to ask his mother. Knowing that she would probably scold him, he walked up the stairs slowly. He thought about the lecture she would give him on keeping things in order. Reaching the tOp of the stairway, ‘he turned the corner. He hung his head. Holding his crumpled math assignement in her hand, his mother stood in the middle of the hallway. Carl was saved; his teacher wouldn't be upset. But his mother was upset, and she began to lecture. Maybe Carl wasn't saved after all. APPENDIX B THE ATTITUDE SCALE APPENDIX B THE ATTITUDE SCALE Directions: Circle one number for each question. How well does the author write for teenagers? I 2 3 4 5 Very poorly Poorly Average Well Very well How easy was the story to understand? 1 2 3 4 5 Very easy Easy Average Hard Very hard How well do you believe the author understands teenagers? l 2 3 4 5 Not at all Little Some Pretty well A lot How smart do you believe the author is? l 2 3 4 5 Very dumb Dumb Average Smart Very smart How much education do you believe the author has had? 1 2 3 4 5 K-6 7-9 10—12 4 years of college Master's Degree How old do you think the author is? 1 2 3 4 5 5-9 10-15 16-20 21-30 Over 30 50 BIBLIOGRAPHY BIBLIOGRAPHY Brag, Stuart. "Effects Of Redundancy and Complex Syntactical Elements on Comprehension and Rate Of Fast and Slow College Readers." Unpublished doctoral dissertation, University of Virginia, 1972. Carr, Joyce B. "An Exploratory Study on the Relationship Between Marginal and Successful Second Grade Readers and Their Thinking Patterns." Unpublished doctoral dissertation, American University, 1977. Cochran, Janet F. "Between Writer and Reader: The Relation- ship Of the Concept of Audience to the Teaching of Composition." Unpublished doctoral dissertation, University Of North Carolina at Greensboro, 1979. Dawkins, John. Syntax and Readability. Newark, Delaware: International Reading Association, 1971. Edwards J. R. "Reactions to Irish Regional Accents." Language and Speech, XX (1977), 280-286. Fagan, W. T. "Transformations and Comprehension." Reading Teacher, XXIV (November 1971), 169-72. Fraser, Bruce. "Some 'Unexpected' Reactions to Various American-English Dialects." In Language Attitudes: Current Trends and Prospects. Edited by Roger Shuy and Roger W. Fasold. Washington: Georgetown Univer- sity Press, 1973, pp. 28-35. Goodman, Kenneth S. "Reading: A Psycholinguistic Guessing Came." International Reading Conference Proceedings, XXIV, 1972, pp. 145-53. Goodman, Kenneth S. and Niles, Olive. Reading: Process and Program. Champaign: Commission on the English Curri- culum. National Council of Teachers of English, 1970. Goodman, Yetta R. "I Never Read Such a Long Story Before." English Journal (November 1974), 65-71. 51 52 Hunt, Kellogg. Syntactic Maturity in School Children and Adults. Mbnographs of the Society for Research in Child Development, XXV, 1970. Irwin, Ruth B. "Judgments of Vocal Quality, Speech Fluency, and Confidence of Southern Black and White Speakers." Language and Speech, XX (1977), 261-66. LeBouef, Robert A. and Matre, Marc. "How Different Readers' Perceive Magazine Stories and Characteris." Journalism Quarterly, LIV (Spring 1977), 50-57. McKiIIOp, Anne S. "The Relationship Between the Reader's Attitude and Certain Types of Reading Response." Unpublished doctoral dissertation, Columbia Uni- versity, 1951. O'Donnell, R. "Awareness Of Grammatical Structure and Reading Comprehension." High School Journal, XLV (1962), 184-88. Osgood, Charles. Measurement and Meaning. Urban: University of Illinois Press, 1957. Palmer, Leslie A. "A Prelininary Report on a Study of the Linguistic Correlate of Raters' Subjective Judgment of Non-Native English Speech." Language Attitudes: Current Trends and Prospects. Washington: Georgetown University Press, 1973, pp. 41559. Pearson, P. David. "The Effects of Grammatical Complexity on Children's Comprehension, Recall, and Conception of Certain Semantic Relations." Reading Research Quarterly, X (1974-75), 155—92. Rey, Alberto. "Accent and Employability: Language Attitudes." Language Sciences, XLVII (October 1977), 7-12. Richards, Ivor Armstrong. Practical Criticism. New York: Harcourt, Brace and World, 1929. Rosenblatt, Louis M. The Reader, the Text, the Poem. Carbondale: Southern Illinois University Press, 1978. Ruddell, Robert T. "The Effect of Oral and Written Patterns of Language Structure on Reading Comprehension." Reading Teacher, XVIII (1965), 270-75. Schmidt, Eunice L. "What Makes Reading Difficult: The Com- plexity of Structures." Paper presented at the Annual Meeting Of the National Reading Conference, December 1-3, 1977. 53 Schneyer, J. W. "Research: Syntactic Structures and Reading Comprehension." Reading Teacher, XXIII (1970), 467-69. Shuy, Roger and Fasold Roger W. Language Attitudes: Current Trends and Prospects. Washington: Georgetown Uni- versity Press, 1973. Shuy, Roger and Williams, Fred. "Stereotyped Attitudes of Selected English Dialect Communities." Language Attitudes: Current Trends and Prospects. Washington: Georgetown University Press, 1973, pp. 85-96. Smith, Frank. Psycholinguistics and Reading. New York: Holt, Rinehart and Winston, 1973. Smith, Frank. Understanding Reading. New York: Holt, Rinehart and Winston, 1971. Smith, W. L. "The Effect of Transformed Syntactic Structures on Reading." Paper presented at the International Reading Association Conference, Anaheim, May 6-9, 1970. Smith, W. L. and George Mason E. "Syntactic Control in Writing: Better Comprehension." Journal of Reading, XV (1972), 355-58. Stoodt, Barbara. "The Relationship Between Understanding Grammatical Conjunctions and Reading Comprehension: Final Report." Columbus: Columbus Research Foundation, September 1970. Stubbs, Michael. Language and Literacy. London: Routledge and Kegan Paul, 1980. Takahashi, Barbara. “Comprehension of Written Syntactic Structures by Good Readers and Slow Readers." Un- published Master's thesis, Rutgers University, 1975. Taylor, Orlando L. "Teachers' Attitudes Toward Black and Nonstandard English as Measured by the Language Attitude Scale." Language Attitudes: Current Trends and Prospects. Washington: Georgetown University Press, 1973, pp. 174-201. Thurstone, Louis L. The Measurement Of Attitude. Chicago: The University of Chicago Press, 1929. 54 Tucker, G. Richard and Lambert, wallace E. "White and Negro Listeners' Reactions to Various American-English Dialects." Advances in the Sociology of Language (Volume II). Edited by Joshua Fishman. The Hague: Mouton, 1972, pp. 175-84. Williams, Fred. "Dialect Attitudes and Stereotypes." Language Attitudes: Current Trends and Prospects. Washington: Georgetown University Press, 1973, pp. 113-28. Young, Ethel E. "Comprehension of Syntactic Structures When Presented in Visual and Auditory-Visual Modes to a Selected Group of Adolescent Readers." Unpublished doctoral dissertation, Michigan State University, 1977. £59331;qu 63919 "GONG!" 18 V8 811