Marginal Pedagogy - SAGE Journals - SAGE Publications

0 downloads 0 Views 160KB Size Report
ings can influence their perceptions of the social context of a writing-from-sources task. Over 120 students read variously annotated letters to the editor, wrote ...
WRITTEN COMMUNICATION / APRIL 2002

Historically, annotations have provided a means for discussing texts and teaching students about reading practices. This study argues that giving students annotated readings can influence their perceptions of the social context of a writing-from-sources task. Over 120 students read variously annotated letters to the editor, wrote response essays, and answered recall and attitude questionnaires. Evaluative annotations influenced students’ perceptions of the text: Passages annotated with positive evaluations were rated as more persuasive than identical passages without annotations; passages annotated with negative evaluations were perceived as less persuasive. Students’ global attitudes to the issue were unaffected. Evaluative annotations seemed to decrease student writers’ reliance on summary and encourage advanced engagement with source materials. However, some annotations appeared to have negative impacts on essays, causing students to include irrelevant information. A hypothesis that the perceived position of the annotator shapes students’conceptions of the rhetorical task is advanced and lent limited support.

Marginal Pedagogy How Annotated Texts Affect a Writing-From-Sources Task JOANNA WOLFE University of Louisville

Historically, the practice of annotation is often associated with medieval manuscript cultures. Medieval readers persistently used the margins of manuscripts to discuss, critique, and learn from the annotations left behind by earlier readers. These marginal glosses were so important to manuscript cultures that they were routinely transcribed along with the primary texts. In some copies of the Bible and the Talmud, marginal annotations eclipsed the primary text as multiple readers added new layers of commentary and responded to the annotations left behind by earlier readers. Such conversations could extend across centuries and are often an object of admiration Author’s Note: I would like to thank Davida Charney for her many thoughtful readings of this work. Financial support for this project was provided by the National Council for Teachers of English. WRITTEN COMMUNICATION, Vol. 19 No. 2, April 2002 297-333 © 2002 Sage Publications

297

298

WRITTEN COMMUNICATION / APRIL 2002

and fascination for contemporary readers. In fact, in his widely influential treatise on reading, the critic Mortimer J. Adler (1940/1967) claims he learned to read critically by observing the annotations on medieval manuscripts: I might as well confess here that I have learned much of what I know about reading from examining a medieval commentary. The rules I am going to prescribe are simply a formulation of the method I have observed in watching a medieval teacher read a book with his students. (p. 97)

The rise of print technology greatly constricted this once common practice of sharing and accumulating marginal annotations. As more people came to own individual copies of a text, it became increasingly rare for multiple readers to share and comment on the same material copy of a work. Annotation thus became primarily a private practice—the markings of an individual reader on a solitary copy of a text. However, whereas print technology impoverished annotation practices, newer electronic technologies today promise to revive them. In the past few years, a number of software applications have emerged that provide mechanisms for capturing and distributing the annotations made by readers of electronic texts. These applications include widely used commercial products such as Microsoft Word and Adobe Acrobat; collaborative writing programs such as CommonSpace; tools facilitating annotation and dialogue on the World Wide Web such as ThirdVoice, Co-Note, and Critical Tools; new technologies for displaying annotations such as Fluid Documents; and handheld devices, such as XLibris, that use electronic pens to capture the materiality associated with reading physical documents. As increasing numbers of readers take advantage of these annotation tools, they, like medieval scholars, will be leaving behind tangible traces of their mental activities as they interpret, analyze, and critique the texts they read. Many of these annotations will be housed in public (or semipublic) databases where they can be accessed by subsequent readers of a text who might use them as aids to guide their own reading and writing processes. That subsequent readers of a text often do attach value to the annotations of previous readers is evidenced through a variety of anecdotal data. In a study of annotations that included interviews with students buying used books in a campus bookstore, Marshall (1997, 1998) found that many students sought out copies of books with

Joanna Wolfe

299

useful-looking markings, generally in the form of extensive marginal notes. Like the college students Marshall describes, the early pioneers of hypertext systems also placed considerable value on the annotations readers might leave behind. In the earliest description of a hypertext-like document system, Bush (1945) describes how readers’ annotations might be used to form “trails” in a hypertext system that later readers could follow to connect individual documents to one another. Van Dam (1988), another early pioneer in hypertext, defends his interest in annotation by describing how in college he preferred to “grab the dirtiest copy of a book from the library, rather than the cleanest one, because the dirtiest ones had the most marginalia, which I found very helpful” (p. 892). Scholars and educators also have found annotations valuable for a variety of purposes. In a recent book focused entirely on annotations, Jackson (2001) argues that marginalia could be of benefit to study in the history of reading and suggests that booksellers and librarians be more aggressive in cataloguing annotations on books. Hawhee (1999) uses the annotations on contemporary handbooks to assess the goals and assumptions motivating John Hodges’s creation of the influential Harbrace College Handbook in 1941. Multiple scholars have made entire careers out of studying the annotations made by Samuel Coleridge (see Whalley & Jackson, 1980). Salvatori (1996) describes the benefits of having students photocopy their annotations on literary texts to share with classmates in order to compare the arguments that different readers construct in reaction to a text. In addition, the prevalence of annotated essays in composition textbooks implies that composition instructors also believe annotations to have some value for subsequent student readers. However, the exact nature of this value remains unclear. The differences in how various composition textbooks use annotated essays attests to this ambiguity. For instance, Wood (1995) uses an annotated excerpt to demonstrate how to “underline and annotate important ideas” (p. 92); Ramage and Bean (1995) use two different annotated versions of the same essay to model and contrast “reading as a believer” and “reading as a doubter;” Lunsford and Ruszkiewicz (1999a) use annotations didactically to point out rhetorical strategies in model essays (e.g., “an authority is cited in support of the proposal,” p. 189); Rottenberg (2000) contains one essay annotated to model critical reading strategies and 10 readings annotated didactically to point out rhetorical strategies; and in the innovative composition reader, The Presence of Others, Lunsford and Ruszkiewicz (1999b) again use

300

WRITTEN COMMUNICATION / APRIL 2002

annotated essays—this time including the annotations of multiple readers—to model reflective and critical reading strategies and to show how informed and critical readers can nevertheless disagree. The differences in how these texts employ annotated essays and the different degrees to which they rely on annotations strongly suggest that the composition community has yet to come to a consensus about the value annotated essays might have for student readers. A few studies have attempted to examine how annotations affect subsequent student readers. Unfortunately, most of this research has narrowly focused on how annotations affect studying, testing the impact of reading preannotated material on recall of the texts or performance on multiple-choice tests. These studies generally conclude that reading texts that have been highlighted by expert readers increases study time and improves recall of the emphasized items (Cashen & Leicht, 1970; Crouse & Idstein, 1972; Schumacher & Nash, 1991) but is not as effective as having readers make their own highlights (Fowler & Barker, 1974). Although the aforementioned anecdotal evidence suggests that annotations have benefits beyond the simple facilitation of recall or comprehension, little systematic research has been conducted that can help instructors understand exactly what those benefits might be. We might assume that when individuals read materials that have been annotated by a previous reader the social context of the textual interaction shifts to encompass this annotator. The annotator’s presence can influence readers’ interactions with the source text, causing readers to evaluate the source text differently in light of the annotator’s reactions to the material. Because of the annotations, readers might see connections to other works or outside influences that would otherwise go unnoticed. If planning a written response to the source text, readers might perceive the annotator as a potential audience for this topic and adjust their own writing to accommodate the questions, reactions, and biases reflected in the annotations. And if the annotator is someone a reader knows, he or she might reevaluate his or her assessment of this individual’s personality or viewpoints in light of the annotations. Finally, if the relationship between annotator and reader is that of instructor to student, the student might possibly take it on himself or herself to mimic the reading strategies reflected in the annotations, possibly by making similar annotations of his or her own. This assumption, at least, seems to be the rationale behind the sample annotated essays presented in many composition textbooks.

Joanna Wolfe

301

Because annotations seem to affect the social contexts in which writing occurs, they have the potential to influence the writing tasks that students construct for themselves—particularly if the annotations contain arguments about the source materials and are thought to have been made by an instructor. If students perceive the annotator as a potential reader of their own texts, they might envision a particular, opinionated reader already familiar with the source texts. Such a task representation could encourage students to compose arguments about the source texts rather than to summarize their contents, because they have evidence that their readers are already familiar with this content. Studies that suggest that providing students with information about the audience for their persuasive writing increases motivation, level of audience adaptation, and number of self-identified arguments (Black, 1989; Redd-Boyd & Slater, 1989), lending additional weight to the hypothesis that annotations could augment students’ use of argument. Alternatively, if students perceive argumentative annotations as reflecting critical strategies that they should imitate, their task representations might also shift them toward argumentative writing. Because students frequently misinterpret problem-based argumentative tasks as text-based report tasks, activities that encourage students to construct an argumentative context for their writing are a definite asset to composition instructors and students (Flower, 1990; Greene, 1991; Penrose, 1992). However, all of this presupposes that students will actually read the annotations. Because students are often thought to fail to read annotations instructors make on their own compositions (MacDonald, 1991), they might similarly ignore annotations made on source texts. And if students do read the annotations, there is no guarantee that their effects will be positive. Reder (1985) suggests that extraneous material that opens up interpretation rather than constrains it can interfere with comprehension and recall of a text. Another danger with annotations is that students might perceive them as the “correct” interpretations or responses to the source text. The fear that annotations might curtail students’ independent thought was expressed by many of the anonymous textbook reviewers of The Presence of Others (Lunsford & Ruszkiewicz, 1999b), a collection of readings containing several essays annotated by instructors and students (Solicited Masked Reviews, 1999). Thus, studies examining the effects of different types of glosses on student readers and writers are needed to determine which, if any, annotations will be useful to students. Because annotation systems

302

WRITTEN COMMUNICATION / APRIL 2002

that will allow users to search through and display comments made by other readers are currently under development, an understanding of how this commentary might best be utilized is crucial.The present study is an initial attempt to address this need.

OVERVIEW OF STUDY

Participants were asked to read a set of letters to the editor, write a persuasive essay based on these materials, and complete a postwriting questionnaire. The letters to the editor were either annotated with evaluative commentary, annotated with underlining only, or not annotated. To verify that students did read the annotations, the postwriting questionnaires were analyzed to determine if students’ recall of factual information or their attitudes toward the source materials were affected by the annotations. To assess whether students were inspired to mimic the reading strategies reflected in the annotations, the source texts students worked from were analyzed to see if those who received annotated materials did more annotating themselves. To examine whether annotations might have influenced how students represented a writing-from-sources task, student essays were analyzed for types of argumentative activity (e.g., summary, concession, and argumentation), length, and quality.

METHOD

Participants A total of 122 students enrolled in lower-level composition courses at a large public university participated in this study. Of these, 52 were enrolled in introductory first-year composition, and 71 were enrolled in other lower-level composition courses. Individual sections of these courses were self-selected by instructors who volunteered their classes for the study. Of the participants, 56 (46%) were freshmen, 35 (28%) were sophomores, and 31 (25%) were juniors or seniors. Materials Prompts. The prompts for reading and writing consisted of a series of letters to the editor of The New York Times responding to a 1997 con-

Joanna Wolfe

303

troversy sparked by the protest of five Orthodox Jewish students against Yale University’s policy requiring all students to live on campus for their first two years of attendance. The first letter in the series is a lengthy statement entitled “College Life vs. My Moral Code,” written by Elisha Dov Hack (see Appendix), one of the five students protesting the residence requirement. Hack’s letter describes the history of the issue and goes on to critique the lax sexual morals of dorm life, concluding that “the moral message Yale’s residences convey today is not one that our religion accepts. Nor is it a moral environment in which the five of us can spend our nights, or a moral surrounding that we can call home.” Hack’s letter is followed by six short responses, all critiquing to various degrees his position (see Appendix). The writers of these letters to the editor include a Yale student, the dean of Yale, a law professor from another university, and a rabbi. This issue was chosen because the stakes and history of the controversy were easy to grasp and because the topic was one likely to interest students. Moreover, where other issues might require students to address an audience with more expertise than themselves, most students in this study appeared comfortable with presenting themselves as authorities on college dorm life. Finally, because several instructors had successfully used these letters to the editor as discussion material in their classes and because Elisha Dov Hack’s original letter has been printed as a reading in Rottenberg’s (2000) Elements of Argument, these materials were selected as representative of texts that students might be asked to analyze in an introductory composition course. A pilot experiment asking 26 instructors to annotate the letters to the editor in preparation for writing their own arguments on the topic allowed me to identify six hot spots that at least one fifth of the instructors had annotated in some form. Using these hot spots, the following four variations of the letters to the editor were constructed: a. no annotations (control group): “clean” primary texts without annotations or other markings. b. underlining only: primary texts with the six hot spots underlined; no commentary accompanies the underlining. c. evaluative annotations +/–: primary texts with handwritten comments noted in the margins next to the six hot spots. Three of these six comments are clearly positive evaluations of the source text (e.g., “great metaphor—clear presentation of the claim”), whereas the remaining three are clearly negative evaluations of the source (e.g., “logical flaw—bad analogy and possibly racist”). Where possible, these comments were taken directly from annotations made by

304

WRITTEN COMMUNICATION / APRIL 2002

Table 1 Description of the Passages and the Annotations They Received in the Two Evaluative Annotation Conditions Description of Passages in Primary Text That Received Annotations

Annotation for Evaluative 1 (+/–)

1. By Elisha Dov Hack opposing “great metaphor— Yale’s dorm policy and clear presentation criticizing Yale’s “anything of the claim” (+) goes” residential religion

Annotation for Evaluative 2 (–/+) “unfair—distortion of Yale’s policy” (–)

2. By Jewish student describing “is this true? ouch!” (–) “great ethos! yes!” (+) his positive experience at Yale and claiming that, unlike Hack, he is glad to be here 3. By dean defending Yale’s poli- “sounds like Yale is cies and expressing willingdoing all it can— ness to accommodate works for me” (+) students

“specifics? sounds like typical administrator ‘b.s.’” (–)

4. By a law professor critiquing “wow. great point. “but Hack doesn’t argue the irony of Hack’s desire to enforced segregathat we should return return to the residential polition in the 50s. selfto this policy. Why this cies of the 1950s when Yale segregation now.” misrepresentation?” (–) had a quota limiting the (+) number of Jews admitted “good analogy” (+) “logical flaw—bad 5. By a Texan opposing Hack analogy and possiand comparing his lawsuit to bly racist” (–) suing a restaurant for serving non-Kosher food 6. By a New Haven resident suggesting that Hack circumvent Yale’s policy by maintaining two residences

“is this for real? harsh!” (–)

“I agree—he has options” (+)

instructors. The first annotation made in this set was positive. See Table 1 for a description of the six “hot spots” and the distribution of positive and negative annotations in this condition. d. evaluative annotations –/+: same as Group C, except the pattern of positive and negative comments is reversed (e.g., the positive comment “great metaphor” is replaced by “unfair—distortion of Yale’s policy,” and the negative comment “logical flaw” is replaced by “good analogy”). The first annotation in this set was negative. See Table 1 for more information about the distribution of these annotations.

Joanna Wolfe

305

The alternation of positive and negative evaluations in Groups C and D was implemented to isolate the valence (positive or negative evaluation) of the annotation from the content of the material being annotated. If student attitudes toward material in the primary text are the same regardless of the annotation valence, we might conclude that the positive or negative valence of an annotation has little effect. On the other hand, if student attitudes toward the two sets of evaluative annotations differ, we might then conclude that the valence of an annotation does affect readers’ attitudes toward the source material. The annotation valence was alternated within individual sets of letters (rather than creating one set of letters with all positive annotations and the other set with all negative annotations) to minimize differences in the overall perception of the annotators. Alternating positive and negative annotations should create an impression of a neutral and objective annotator in both sets of letters. Thus, alternating positive and negative annotations should help separate the effect of individual annotations from the cumulative effect of all the annotations. The underlining group (Group B) was included to differentiate between the presence of an annotation and the content of the annotation. In other words, does the mere presence of an annotation, regardless of what it communicates, affect student readers? It is important to look at the influence minimal annotations, such as underlining or brackets, might have, because these markings are the most common form of annotation (Marshall, 1998). In addition to the six hot spots marked in Groups B, C, and D, two contextual items were circled on these materials: The profession of the law professor was circled, and the city (Houston) was circled for the only respondent not located on the East coast. The purpose of circling this contextual information was to see if explicitly calling students’ attention to the rhetorical context would improve their overall attentiveness to argumentative context. Postwriting questionnaires. The postwriting questionnaire contained the following three sections: a section testing students’ abilities to recall information about the rhetorical context of the arguments they had read, a second section assessing their overall attitudes toward the arguments, and a third section measuring the effects of the annotations on their ability to evaluate and analyze specific claims from the readings.

306

WRITTEN COMMUNICATION / APRIL 2002

The recall questions consisted of four statements—some true and some false—about the letter writers. Students were asked to rate their confidence in these statements on a 1 to 5 Likert scale ranging from definitely false to definitely true. Two of the statements required students to recall information about the letter writers (profession and city of residence) that had been annotated in the experimental conditions, and two statements required students to recall information about the letter writers that did not receive any annotation in any of the conditions. Attitude questions consisted of eight 5-point Likert scale items asking students how strongly they agreed or disagreed with statements such as “I feel sympathy for Hack” or “The writers responding to Hack represent his argument fairly.” These questions were intended to provide a rough gauge of students’ overall perceptions of the arguments. Specific claims questions consisted of nine passages quoted from the readings. Six of these nine passages were the hot spots annotated in the experimental conditions; the other three passages were not annotated in any of the conditions. Students were asked to evaluate each claim on a 5-point Likert scale indicating how persuasive they found the passage. In addition, students were asked to write a brief explanation describing their reasons for each ranking. Demographic questionnaires. Prior to writing, students were given a brief demographic questionnaire asking their gender, age, ethnicity, the number of college credits completed, approximate GPA, approximate SAT verbal score, the number of college-level writing classes taken, native language, and the approximate size of their hometown. Procedures This research was conducted in regularly scheduled class periods of lower-level composition courses beginning in the second half of the spring 1999 semester. To minimize disruption to classes, the study was designed to be completed in a single 50-minute class period. Students were told that the purpose of the study was to collect data on student reading and writing practices. Students were informed that their participation in the study was voluntary and that the essays they produced would not be read by their instructors. Students who opted not to participate were asked to complete the essay, which would be returned to them at the end of the period. Two students opted not to participate.

Joanna Wolfe

307

Copies of the four variations of letters to the editor described earlier were randomly distributed to the class. Students were told they had 35 minutes to prepare a short, persuasive essay based on the materials and that any notes, drafts, or other materials produced during the activity would be collected. In addition, the instructions for students in the three experimental conditions contained the following statement, which was not read aloud: “the hand-written comments were made by an [introductory composition] instructor who was planning to give this assignment in class. They may or may not help you.” At the end of the 35-minute reading and writing period, all of the materials were collected, and students were then given the postwriting questionnaire on the materials. Most students took 10 to 15 minutes to complete the questionnaire. Data Analysis The following four dependent measures were analyzed: the effects of annotations on memory, attitude, process, and written products. Attitude is further divided into global attitude (whether the student agreed or disagreed with Hack) and local attitude (how persuasive the students found individual passages from the source texts). Written products are further divided into quality ratings and essay coding. To conserve space, discussion of these variables and their relevant levels takes place in the Results section.

RESULTS AND DISCUSSION

Recall If students paid attention to the annotations, then their recall of information that was annotated should be augmented. Moreover, if students perceived the annotations as indicating strategies they should be imitating, they might tend to note and recall information similar to what was annotated even if it had not been emphasized. Because undergraduates frequently fail to attend to the rhetorical context of arguments, such as the status or background of the author (Haas & Flower, 1988; Haswell et al., 1999), recall tests focusing on contextual information might suggest whether students imitated the reading strategies modeled in the annotations. Consequently, four items on the questionnaire tested students’ recall of information

308

WRITTEN COMMUNICATION / APRIL 2002

Table 2 Average Correctness Scores (and Standard Deviations) for Recall of Annotated and Nonannotated Contextual Information by Annotation Condition

Annotation Condition Control No annotation Annotation Underlining Evaluative 1 (+/–) Evaluative 2 (–/+) Marginal means Marginal means all

n

Recall of Annotated Information Identifying Writers

Recall of Nonannotated Information Identifying Writers

a

(1.82)

7.70

(2.01)

a, b

(1.97) (1.48) (2.12)

7.23 7.45 7.07

(2.03) (1.96) (2.48) 7.28 7.38

33

7.12

30 31 28 90 122

7.83 b 8.55 b 8.29 8.24 7.93

NOTE: Minimum score for each cell is 2, and maximum score is 10. Items in the same column with different superscripts are significant at the p < .05 level.

about the identities of the letter writers. Two of these items (shown in column 1 of Table 2) consisted of information about the identity of the letter writers that was circled in the three annotation conditions. The other two items (shown in the second column of Table 2) consisted of information about the identity of the writers that was not circled. Table 2 shows that annotations did influence students’ recall of emphasized information but not their recall of information similar to that emphasized. A repeated measures analysis for annotation condition by information group found a main effect for information group, F(1, 118) = 9.08, p < .01, and an interaction between annotation condition and information group, F(3, 118) = 4.64, p < .01. Contrast analyses indicated that items in the first information group were better recalled than items in the other groups at the p < .01 level. Separate betweensubjects ANOVAs for each information group then found that students receiving evaluative annotations had improved recall of emphasized information but did not recall nonemphasized information with any greater accuracy than students who did not receive annotations. The first column of Table 2 shows that students in the three conditions receiving annotated materials scored higher than the control group on recall of the items specifically marked by the annotator, F(3, 121) = 3.62, p < .05. A post hoc analysis using Duncan’s multiple range indicates that the no-annotation group significantly differed from the

Joanna Wolfe

309

two evaluative annotation groups at the p < .05 level but did not significantly differ from the underlining group. This suggests that evaluative annotations are more effective than underlining in improving student recall of emphasized information. However, the high recall of the two evaluative annotation groups on items that were emphasized did not result in these students noting similar information that was not emphasized. Column 2 indicates that students in the annotation groups were no more likely than the control group to recall nonannotated information about authors. Thus, Table 2 supports the hypothesis that students receiving annotated materials will score higher than the control group on recall tests of contextual information that was emphasized. In addition, evaluative annotations seem to have affected recall more than underlining. However, the hypothesis that students would model the reading strategies reflected in the annotations by noting similar but unemphasized contextual information did not receive support. Local Attitude The aforementioned analysis of recall suggests that students’ attention was drawn to the passages that were annotated; however, it is unclear whether the content of the annotation affected students’ attitudes toward the source materials. Therefore, nine items on the questionnaire consisted of passages from the source texts that students were asked to rank on a 5-point Likert scale for persuasiveness. These consisted of the six passages that had been annotated in the annotation conditions plus three passages that had not been annotated in any of the conditions. The nine passages were then consolidated into three groups to test how the valence (positive or negative) of the evaluative annotation affected how students perceived the passages: a. passages from letters by Hack, Brodhead, and Colb that received a positive annotation in Evaluation 1 (+/–) and a negative annotation in Evaluation 2 (–/+); b. passages from letters by Nadata, Ancelet, and Walfish that received a negative annotation in Evaluation 1 (+/–) and a positive annotation in Evaluation 2 (–/+); c. passages from letters by Zahavy, Hack, and Ancelet that received no annotations.

310

WRITTEN COMMUNICATION / APRIL 2002

Table 3 Average Persuasiveness Scores (and Standard Deviations) for the Three Passage Sets by Annotation Condition

Annotation Condition None and underlining Evaluative 1 (+/–) Evaluative 2 (–/+) Marginal means F(2, 114)

n

Passage Set 1: Passage Set 2: Positive Negative Annotations in Annotations in Evaluative 1(+/–) Evaluative 1 (+/–) Passage Set 3: and Negative and Positive No Annotations Annotations in Annotations in in Any of the Marginal Evaluative 2 (–/+) Evaluative 2 (–/+) Conditions Means a

(1.86)

b

(1.88)

c

(1.84)

58

9.78

29

10.97

28

8.34

115

9.73 14.22****

11.81

a

(2.28)

9.62 (1.99)

10.40

9.78b (2.30)

9.64 (2.25)

10.13

9.68 (1.96)

10.37

9.64 0.01

10.29

12.86

c

(1.38)

11.55 16.15****

NOTE: Minimum score is 3, and maximum score is 15. Items in the same column with different superscripts are significant at the p < .05 level. ****p < .0001.

Because no differences were found between the no-annotation group and the underlining-only group, results for these two conditions have been combined to simplify the display of the data. Only students who answered all of the questions (n = 115 out of 123 total participants) are included in the analysis. Table 3 shows that the content of the annotation influenced whether students saw passages as persuasive. A repeated measures analysis for annotation condition by passage set found a main effect for passage set and an interaction between annotation condition and passage set. This finding indicates that the valence of an annotation interacted with a passage’s base persuasiveness to influence students’ perceptions of individual passages from the source materials. Table 3 shows a main effect for passage set, F(2, 112) = 31.13, p < .0001, indicating that the different passage sets were not equivalent in their base persuasiveness. A Duncan’s post hoc analysis shows that Passage Set 2 is significantly more persuasive than the other passage sets at the p < .01 level. This difference in the base persuasiveness of the various passages was not part of the experiment design, but simply reflects the fact that some claims will be inherently more

Joanna Wolfe

311

persuasive to a particular group than other claims. A main effect for annotation condition was not found. This suggests that all four groups of students were fairly equivalent in their baseline willingness to be persuaded. Table 3 also indicates that the positive or negative valence of the annotation affected students’ assessments of different passage sets, producing a significant interaction between passage set and annotation condition, F(4, 224) = 13.72, p < .0001. This interaction indicates that the valence of the annotation interacted with the base persuasiveness of the passage to shape students’ evaluations of individual claims in the primary text. The first column of Table 3 indicates that persuasiveness scores for all three annotation conditions were significantly different at the p < .01 level (Duncan’s multiple range). Thus, students in the first evaluative annotation condition, who received positive annotations for this passage set, not only rated these passages as significantly more persuasive than the students who received negative annotations for the passage set but also rated these passages more favorably than the students receiving no annotations or underlining only. Similarly, students in the second evaluative annotation condition, who received negative annotations for this passage set, rated these passages less favorably than the other groups. The strong effect for annotation valence in this column, F(2, 113) = 13.35, p < .0001, suggests that the content of annotations raises or depresses students’ perceptions of the immediate source material in congruence with the valence of the annotation. The second column of Table 3 indicates that only the students receiving negative annotations for this passage set were significantly influenced by the annotation valence. Students in the second evaluative annotation group, who received positive annotations for this set of passages, did not rate these passages significantly more favorably than did students in the no-annotations or underliningonly conditions. These positive annotations may not have affected students because the base persuasiveness of these passages was already high. However, this finding does suggest the possibility that negative annotations may influence attitudes to a greater extent than positive annotations. Global Attitude If students’ attitudes toward individual passages were affected by the annotations, then it might also be possible that annotations could

312

WRITTEN COMMUNICATION / APRIL 2002

affect students’ global attitudes to the controversy. Because students were told the annotations were made by an instructor, they might perceive these comments as representing the “correct” readings of the text. Two items on the postwriting questionnaire measured students’ overall orientation to the controversy. The students in this study were overwhelmingly opposed to the Yale lawsuit, and there were no differences in global attitudes among the four groups. Responses to the statement “I feel Hack’s lawsuit is ridiculous” averaged an agreement score of 4.2 on a 5-point scale, whereas responses to “I feel sympathy for Hack” averaged a score of only 2.1. This strong orientation against the lawsuit is mirrored by the opinions of the Yale student body, 87% of whom opposed the lawsuit (Itano, 1997). Thus, the annotations did not appear to affect students’ overall positions on an issue, which they seem to have approached with very strong opinions. Process If students perceive the annotations as strategies that they should imitate, those who receive annotated materials might be inspired to do more annotating themselves. To determine whether simple exposure to annotated materials might influence students’ own annotation practices, I counted the number of annotations students made on the source texts, the number of words (where relevant) these annotations contained, and the number of words contained in notes students made on a separate sheet of paper. The total number of annotations (including underlines) students made ranged from 0 to 30, with almost two thirds of the students making no annotations at all on their materials. The number of words students made in marginal notes ranged from 0 to 78. Contrary to expectations, no effect was found for the number of independent annotations made. It may be that the limited space in the margins discouraged further annotation when marginal notes were already present, or this finding may simply indicate that receiving annotated texts had no influence on students’ own annotation practices. Written Products: Types of Argumentative Activity Because annotations modify the social context in which writing occurs, it was hypothesized that evaluative annotations might influence how students writing from sources perceive their tasks. Students receiving evaluative annotations might imagine that their audience

Joanna Wolfe

313

holds thoughts and biases that mirror those expressed in the annotations. Alternatively, students might view evaluative annotations as indicating strategies that they should imitate in their own writing. Both of these task representations should encourage students to perceive an argumentative context for writing. Thus, we might hypothesize that students receiving evaluative annotations would write essays with a higher proportion of argument to summary than students who did not receive annotations. Moreover, if evaluative annotations influenced students’ perceptions of their own readers, we might find variations in how the different groups used audience adaptation strategies such as conceding points to the opposition or explaining the writer’s qualifications for speaking on this topic. To examine how annotations affect students’ argumentative activities, essays were divided into t-units, and each t-unit was coded as either argument, summary, concession, ethos, or irrelevant. The categories of activity are defined as follows: argument: statement of a claim or reason, evidence to support a claim or reason, or defense of an underlying warrant (e.g., “Elisha Dov Hack should not sue Yale”); summary: purely factual summary of information found in the source materials (e.g., “Yale requires that all freshmen live in the dorms for their first two years of college”); concession: concede a point to the opposition (e.g., “I can sympathize with Hack”); ethos: establish writer’s qualifications on this topic (e.g., “I too am a freshman”); irrelevant: information or claims that appear unrelated to the argument (e.g., “You [Hack] are a nutcase because you don’t have sex”).

Each t-unit was coded as one of the aforementioned categories. There was no double coding of t-units, and every t-unit was counted. As a reliability measure, a second rater coded one third of the essays, k = .79 using Cohen’s weighted kappa, a level suggesting excellent agreement above chance (Fleiss, 1981). Figure 1 shows that the annotation condition did predict the types of argumentative activity in which students engaged, χ2(12) = 134.2, p < .0001. The annotation groups differed in the amount of summary they included, χ2(3) = 11.67, p < .01. Students in the no-annotation condition wrote essays with the most summary (8.3%) out of any of the groups, whereas students in the Evaluative 1 (+/–) condition had the lowest percentage of summary (3.8%) in their essays. All three of the

314

WRITTEN COMMUNICATION / APRIL 2002

100% 90%

Percentage of t-units

80% 70% 60% 50% 40% 30% 20% 10% 0%

Argument

Summary

Concession

Ethos

Irrelevant

No Annotation

84.4%

8.3%

3.7%

3.3%

0.4%

Underlining

87.4%

4.4%

5.5%

1.9%

0.8%

Evaluative 1 (+/-)

88.2%

3.8%

5.0%

2.4%

0.6%

Evaluative 2 (-/+)

72.1%

6.2%

5.4%

7.7%

8.5%

Figure 1. Distribution of argumentative activity by annotation condition

groups receiving annotations produced essays with less summary than the no-annotation group, thereby lending limited support to the hypothesis that annotations encourage more complex representations of the writing tasks. Several student essays suggest how students might have perceived their writing task. In the following excerpt, written by a student in the no-annotation condition, persuasive argument seems to take a back seat to reporting the facts: Yale University is one of the most prestigious universities in the United States. Along with the excellent education it provides come rules that Yale students must follow. One of these rules is that freshmen and sophomore students must live in on campus housing. . . . The controversy arises when students do not feel comfortable living in the dormitories. In the case of the articles, an Orthodox Jewish student does not feel comfortable living in an environment full of others with different moral beliefs and different everyday habits. This is understandable that young students do not like to be out of their comfort zone, but it is something that people need to learn to deal with because in society today everybody doesn’t always get what they want.

Joanna Wolfe

315

Elisha Dov Hack felt that being required to live in a dorm violates “the principles of Judaism lived according to the Torah and 3,000-yearold rabbinical teachings.” He along with a handful of his peers threatened to sue the university in order to protect their religious ways of life. Everyone should be entitled to their own opinion and should have the right to voice how they feel, but at the same time, the University has the right to establish rules.

The writer summarizes information from the source texts as if she expected readers with no prior knowledge of the topic. However, her assumption that readers will understand her casual reference to “the articles” seems to belie this representation of an uninformed audience. These conflicting signals about the writer’s audience are common in students’ school-based writing, which often asks students to demonstrate understanding of texts rather than to analyze or evaluate them. By contrast, a student in the Evaluative 1 (+/–) group seems to be self-consciously constructing an argumentative task for herself: Elisha Dov Hack argues that he should not be required to live in a Yale dormitory because his religious values will be compromised. The subsequent letters oppose Hack for different reasons. I, too, disagree with Hack’s argument and will incorporate the opposition’s strongest points to make my case. First of all, Yale’s position as stated by Dean Brodhead is convincing. Brodhead begins by illustrating his desire to be as accommodating to the students as possible. He shows sympathy for Hack, but explains why Yale requires students to live on-campus. . . . By choosing not to live on campus, Hack is choosing not to accept what Brodhead considers to be a fundamental aspect of a Yale education.

Like the student in the no-annotation group, this writer begins by summarizing the controversy. However, this succinct summary not only synthesizes some of the materials but also provides a way for the student to integrate her own arguments clearly with claims made in the source texts. Unlike her peer earlier, this writer analyzes the source texts and builds on their arguments. Some of this analysis may have been inspired by the annotations this writer saw, such as the annotation appended to the dean’s letter that read “sounds like Yale’s doing all it can—works for me.” By elaborating on the “convincing” and “accommodating” tone of the dean’s letter, this writer may be

316

WRITTEN COMMUNICATION / APRIL 2002

trying to appeal to readers who, like the annotator, find the dean’s arguments credible. Other writers similarly seemed to generate arguments in response to the annotations. For instance, a student in the Evaluative 2 (–/+) condition seems to be responding to the annotation on Dean Broadhead’s letter reading “specifics? sounds like typical administrator ‘b.s.’” when he writes, First and foremost, Mr. Brodhead, though biased by his position, explains some good reasoning. He essentially tells the reader that the first two years of college are a time of intensive learning about yourself, others and the college community. This is true. What better place to truly understand college than to live right in the middle of it? However, he does not provide any solution for Mr. Hack’s argument. Mr. Brodhead sounds like he copied a passage from the college prospectus. As an administrator, he should lay off of the crap and tell Mr. Hack how he plans to accommodate Mr. Hack’s concerns.

The writer seems to begin the paragraph by attempting to persuade an unconvinced reader to reconsider Brodhead’s claim. By the end of the paragraph, the writer apparently switches his purpose to elaborating on the criticisms expressed in the annotation, and the phrase “as an administrator, he should lay off the crap” is a rewording of the annotator’s claim “specifics? sounds like typical administrator ‘b.s.’” Although the writer has not effectively resolved these competing purposes, he does appear to be responding to the annotations and working through interpretations of the materials that might eventually lead him to a more nuanced and complex argument. The most surprising finding suggested by Figure 1 lies in the dramatically different distributions of argumentative activity in the two evaluative annotation conditions. Although the Evaluative 1 (+/–) annotation group included high percentages of argument in their essays (83.0%), the Evaluative 2 (–/+) group did not, χ2(2) = 28.6, p < .0001. Moreover, students in the Evaluative 2 (–/+) condition included much more irrelevant material in their essays than students in the other groups, χ2(3) = 89.79, p < .0001. Students in the Evaluative 2 (–/+) group were also more likely than others to write about their own backgrounds and qualifications for writing, χ2(3) = 25.85, p < .0001. However, although statements about the writer’s ethos are

Joanna Wolfe

317

frequently considered to be audience adaptation moves (see Hays & Brandt, 1992), the students in the Evaluative 2 (–/+) group did not appear to be considering their readers when they provided information about their backgrounds. The following excerpt from an essay by a student in the Evaluative 2 (–/+) group was judged as containing irrelevant material and a high proportion of statements about the writer’s own ethos: I do not agree with Elisha that because he was raised a certain way, he should be sheltered from the wrongdoings of those around him. Attending Yale means living in a dormitory and if he cannot accept this then he should not go. I am originally from a small east Texas town. There I lead a sheltered life away from such big city pulls like drugs and gangs (although I do admit they were there, they were not as prevalent). I had a support group of friends that I grew up with. We began in the same kindergarten and graduated together. The teachers did as well as they could but all the while I knew I was missing out. That is why I chose the University of Texas as my school of choice. I knew my ideas and beliefs were being limited by my small town environment. So, I chose the biggest school in Texas (I think the nation now) and I have not regretted my decision since.

Rather than supporting a persuasive argument or summarizing the source text, this writer seems to use the source materials as a springboard for discussing issues important to herself. Although the writer does seem to be preparing to draw parallels between the narrowness of her upbringing and that of Elisha Dov Hack, readers do not need this much information. The annotations this student received did not seem to make her more sensitive to readers’ needs. Overall, the students in the Evaluative 2 (–/+) group appeared to be the least interested in the persuasive task. Another writer in this group clearly indicates her disinterest by writing, I can’t think right now because my nose is running. I really need a tissue, but I can’t leave to go get one. . . . Does that pertain to this essay? Possibly, because I could just walk out and get that tissue. After reading these letters to the editor, I have come to a conclusion. . . . Elisha Dov Hack really needs to just go out and get that tissue. [Ellipses are the student’s.]

318

WRITTEN COMMUNICATION / APRIL 2002

Although such disinterestedness should not be surprising given the artificial conditions of the study, it is telling that nearly all of the irrelevant material was isolated in this one group. A follow-up study described at the end of the results suggests that the high proportion of irrelevant material and low proportion of argument in the essays written by the Evaluative 2 (–/+) group might be attributable to students’ perceptions of the annotator. Written Products: Essay Quality and Length Because the annotation condition seemed to correspond with the types of argumentative activities in which students engaged, there might also be a correlation between annotation condition and essay quality. Two instructors also rated each essay’s effectiveness using a 4point Likert scale for a variety of quality measures, including quality of argument, responsiveness to alternative perspectives, and use of source materials. The ratings for these three measures were combined for an argumentative quality score ranging from 3 to 12 points; interrater reliability for the argumentative quality score was calculated using Pearson’s correlation, r = .79. The average argumentative quality score for the Evaluative 1 (+/–) group was 7.08 (on a scale of 3 to 12), 6.70 for the underlining-only group, 6.48 for the no-annotation group, and 6.30 for the Evaluative 2 (–/+) group. Although essays in the Evaluative 2 (–/+) group had the lowest quality scores, as the earlier analysis might lead one to expect, these differences were not significant. Essay length ranged from 3 to 53 t-units and from 75 to 684 words and did not seem to be affected by annotation condition. The average essay length was 15.6 t-units and 262 words. The absence of significant findings for quality should not be surprising because as Flower (1990) points out, there is a gap between building an appropriate task representation and writers’ abilities to carry out the plans they intend. Moreover, the strict time limit allowed writers little space in which to develop their ideas, and the raters experienced difficulty reliably assessing such short and underdeveloped texts. However, at this stage of the research, it is sufficient to see that marginal annotations have some influence over written products in order to conclude that this intervention has potential to stimulate certain rhetorical activities, without necessarily requiring evidence that the intervention will influence overall writing quality.

Joanna Wolfe

319

Summary The data presented here show that students note and are influenced by annotations made on source texts when they are told those annotations are made by instructors. Students seemed to pay more attention to annotations that evaluated the text than annotations consisting of underlines. When students saw evaluative annotations, their recall of information that had been emphasized improved, and their opinions of annotated claims were swayed in the direction of the gloss’s valence (i.e., positive evaluations uplifted students’ ratings of source arguments, and negative evaluations depressed their ratings). Evaluative annotations also affected students’ written products. Students in the evaluative annotation conditions included less summary in their essays than students who had not seen evaluative annotations. This finding lends support to the hypothesis that evaluative annotations will encourage students working from sources to represent their task as a problem-based argument rather than a text-based report. However, whereas students in the Evaluative 1 (+/–) group wrote argumentative essays, students in the Evaluative 2 (–/+) group engaged in relatively little persuasive activity, substituting instead disproportionate amounts of ethos and irrelevant material. Students in the Evaluative 2 (–/+) group generally appeared to be less interested in the writing task than students in other groups. These results are surprising because both evaluative annotation groups saw the same source material annotated, and both groups saw equal numbers of positive and negative annotations. The following section attempts to account for these differences. Why Were the Two Evaluative Groups Different? The dramatic differences between the two evaluative annotation groups were completely unanticipated. Because the two groups differed primarily by the order in which they saw positive and negative annotations, differences were expected in their local attitudes to individual passages from the source materials, but not in their task representations or in the types of writing in which they engaged. However, the essays produced by the Evaluative 1 (+/–) group exhibited substantial persuasive activity, whereas the essays produced by the Evaluative 2 (–/+) group tended to disengage from the central

320

WRITTEN COMMUNICATION / APRIL 2002

argument by providing irrelevant information and relatively little evidence and argumentation. Two explanations for these differences seem feasible. First, if students were imitating the strategies they saw in the annotations, then they may have reacted to the two sets differently because of the wording of individual annotations. One of the annotations the Evaluative 2 (–/+) group read was “specifics? sounds like typical administrator ‘b.s.’” It may be that some students perceived the term b.s. as a license to engage in more playful, less academic activity. A second explanation for the differences between the two evaluative annotation groups is that the order of positive and negative annotations may have led to different impressions of the annotator. Because students in the Evaluative 1 (+/–) group first saw a positive annotation on Hack’s letter and then a negative annotation on a letter by one of Hack’s detractors, this group may have been inclined to perceive the annotator as supporting Hack and his lawsuit. By contrast, the Evaluative 2 (–/+) annotation group, which first saw a negative annotation on Hack’s letter and a positive annotation on the letter of one of Hack’s detractors, may have tended to perceive the annotator as unsympathetic toward Hack and his lawsuit. Because 90% of the students in this study wrote essays opposing Hack, the students in the Evaluative 1 (+/–) group may have been more likely than other groups to perceive their task as one that called for persuading an audience whose viewpoint differed from their own. By contrast, the students in the Evaluative 2 (–/+) group may have constructed a less complex rhetorical task for themselves, one that consisted primarily of preaching to the choir. These students may have engaged in less rhetorical activity than the other groups because their mental model of the audience was a reader who already shared their point of view. This image of the already persuaded reader might account for the tendency of this group to produce more irrelevant material and less argument in their essays. To test the hypothesis that students perceived the annotators of the two sets of materials differently, 20 students were asked to study the Evaluative 1 (+/–) or Evaluative 2 (–/+) materials and complete a questionnaire soliciting their impressions of the annotators. Table 4 shows that the two sets of evaluative annotation materials did lead students to draw different conclusions about the personality and position of the annotator. Students receiving the Evaluative 1 (+/–) annotations were more likely than students in the second evaluative

Joanna Wolfe

321

Table 4 Mean Student Ratings (and Standard Deviations) of Their Impressions of the Annotator

Annotation Condition Evaluative 1 (+/–) Evaluative 2 (–/+) Marginal means F(1, 19)

n

Annotator Supports Hack’s Cause

Annotator Is Probably a Teacher

Annotator Is Biased

Annotator Feels Sympathy for Hack

Annotator Seems Easily Persuaded

10

2.8 (.79)

4.1 (.99)

3.2

(.79)

2.8 (.79)

2.9

(.57)

10

1.6 (.84)

3.2 (.79)

2.1

(.74)

2.0 (.94)

1.8

(.63)

20

2.2 10.8***

3.7 5.03**

2.7 10.37***

2.4 4.24*

2.4 16.75***

NOTE: Minimum score is 1 (strongly disagree), and maximum is 5 (strongly agree). *p = .05; **p < .05; ***p < .01.

annotation group to perceive the individual making the comments as supporting Hack’s cause, F(1, 19) = 10.8, p < .01, and marginally more likely to perceive the annotator as sympathetic toward Hack, F(1,19) = 4.24, p = .05. These findings support the hypothesis that students would perceive the annotators in the two conditions as adopting different stances toward the issue. Students in the Evaluative 1 (+/–) annotation group were also more likely than those in the Evaluative 2 (–/+) group to believe that the annotator was a teacher, F(1, 19) = 5.03, p < .05; more likely to view the annotator as biased, F(1, 19) = 10.37, p < .01; and more likely to perceive the annotator as easily persuaded, F(1.19) = 16.85, p < .01. Even if these divergent perceptions of the annotator do not account for the differences between the two groups, Table 4 suggests that subsequent readers do make inferences about an annotator’s personality, viewpoint, and biases based on the annotation content. If students use these inferences about the annotator to construct a mental model of their readers, then it seems possible that certain annotations will significantly affect how students perceive the rhetorical situation. Further research will be needed to test whether perceptions of an annotator’s viewpoint vis-à-vis the writer’s own position can affect the quality of students’ written products.

322

WRITTEN COMMUNICATION / APRIL 2002

IMPLICATIONS

This study points to the potential of instructors’ annotations to shape students’ opinions of individual claims in source materials and to affect how students represent a writing-from-sources task. It was originally hypothesized that evaluative annotations might affect students’ task interpretations in two ways: (a) Students might perceive the annotator as the audience for their own texts and hence adapt their writing to meet the needs of this reader, and (b) students might perceive the annotations as reflecting critical strategies that they should imitate in their own writing. The analyses of students’ essays suggest that both of these task representations were at work: Students seemed both to respond to arguments the annotator had raised and to imitate some of the analytic strategies observed in the annotations. However, the differences in written products of the two evaluative annotation groups, combined with their different impressions of the annotator, may suggest that annotations might be more useful as a tool that can help students visualize concrete readers for their writing than they are as models of strategies that students can imitate. However, additional research is needed to test this hypothesis. The finding that students developed definite impressions of the personality of the annotator is consonant with how readers from earlier eras perceived annotations. As H. J. Jackson (2001) meticulously shows, prior to the introduction of mass production printing techniques in the mid-19th century, readers routinely shared annotated books with their familiars. Unlike public texts, private marginalia came to be perceived as sincere, trustworthy, and intimate impressions of the individual because they ostensibly were not written for the benefit of an audience. In her posthumous novel, Maria; or the Wrongs of Woman, Mary Wollstonecraft (1798/1975) dramatizes this association between marginalia and intimacy when Maria, unjustly confined in a madhouse, is brought a bundle of books that belong to another inmate: Dryden’s Fables, Milton’s Paradise Lost, with several modern productions, composed the collection. It was a mine of treasure. Some marginal notes, in Dryden’s Fables, caught her attention: they were written with force and taste. . . . She read them over and over again; and fancy, treacherous fancy, began to sketch a character, congenial with her own, from these shadowy outlines.—“Was he mad?” She reperused the

Joanna Wolfe

323

marginal notes, and they seemed the production of an animated, but not of a disturbed imagination. Confined to this speculation, every time she re-read them, some fresh refinement of sentiment, or acuteness of thought impressed her, which she was astonished at herself for not having before observed. (pp. 34-35)

A correspondence between the two inmates ensues, and they soon become lovers. Like Wollstonecraft’s (1798/1975) Maria, the students in this study seemed to use the “shadowy outlines” found in the marginal annotations to “sketch a character.” The students allowed the marginal annotations to influence their perceptions of the source texts, and possibly to influence how they perceived the social context for their writing. Teachers might attempt to take advantage of the potential of annotations to shape students’ perceptions of the social contexts in which writing occurs by encouraging students to share their own annotations on source texts with their classmates. Students might also benefit from having others annotate their own written products with readerly or evaluative annotations, such as those used in this study, rather than the directive annotations that teachers and peer reviewers often use now. Such sharing of annotations among peers and others will be greatly facilitated by the developments currently underway in annotation technologies. As more ergonomic interfaces for online reading are developed, annotating materials online will become easier (for discussions of handheld devices that use pen-based input to make digital annotations, see Farkas & Poltrock, 1995; Marshall, Price, Golovchinsky, & Schilit, 1999). Once such tools make it easy for readers to store their annotations in online databases, these annotations will be available for searching, filtering, and displaying by other readers. Instructors and students might work together to examine these annotations, observing how different readers responded to particular arguments in source texts. Continued exposure to a variety of readers’ annotations might help students, over time, develop better models of how readers interact with texts to construct meaning. Such an understanding of how readers in particular discourse communities approach source texts could help student writers better anticipate and accommodate active readers of their own writing (see Nystrand, 1990).

324

WRITTEN COMMUNICATION / APRIL 2002

Implications for Textbooks The findings from this study have several implications for how annotations are used in composition textbooks. Many composition textbooks advise students to annotate texts as they read, and often provide essays with model annotations for students to imitate. Although some of these model annotations show students how to critique or analyze source texts, others simply demonstrate how to “underline and annotate important ideas” (Wood, 1995, p. 92). Not only does this study suggest that exposure to annotated materials might fail to influence students’ own annotation practices, but it also indicates that by simply modeling strategies for summarizing, these textbooks may be missing an opportunity to influence how students perceive writing-from-sources tasks. Rather than modeling strategies for comprehending a text, composition textbooks might use annotated materials to show students how readers frequently critique, question, and evaluate arguments in a text. Students might additionally be reminded that readers will approach their own writing with biases and expectations similar to those reflected in the annotations. In addition, researchers might want to look at how the didactic annotations that are often published in composition textbooks and readers affect students’ writing or critical thinking activities. These annotations frequently point out rhetorical strategies in model essays (e.g., “the author addresses possible rebuttals to her argument”) or else are explicitly intended to provide models of reading and annotating that can differ from the markings actual readers make for their own purposes. The composition community needs additional research to assess the costs and benefits of these explicitly studentdirected annotations. In situations where comprehension of the source text is an important issue, authors of textbooks and other materials might note that students in this study seemed to pay more attention to annotations that evaluated the text than annotations consisting of underlines. When students saw evaluative annotations, their recall of information that had been emphasized improved more than that of students who only saw underlining. Potentially, annotations made by authors of textbooks might help students to understand main points in the source texts. The didactic annotations that many composition textbooks use to point out rhetorical strategies or organizational features in model essays (e.g., “the author addresses possible rebuttals to her

Joanna Wolfe

325

argument”) need additional study. Such annotations are usually created explicitly with a student audience in mind and can differ considerably from the markings actual readers make for their own purposes. Rather than encouraging students to consider the contexts or audiences for their writing, such annotations, by focusing on the text, may imply that the writer’s goal is to make his or her writing fit a hypothetical ideal text. Directions for Further Research Additional research should further assess how perceptions of annotators influence students’ written products. This study suggests that annotations reflecting viewpoints that differ from the student’s own position might positively affect the quality of argumentation, whereas annotations coinciding with the student’s own position might negatively affect the quality argumentation. However, this hypothesis that the relationship between the perceived position of the annotator and writer’s own viewpoint influences students’ rhetorical activities needs further testing. Future studies might additionally assess how students respond to annotations made by multiple readers, particularly if these readers seem to disagree in their reactions to the source materials. By presenting an issue as being truly divided and controversial within a community of readers, such annotations might encourage students working from sources to construct complex models of their readers and their tasks. In addition, future research might compare how students respond to annotations made by their peers rather than their instructors. Because students are often more motivated when writing for peers than writing for teachers (cf. Cohen & Riel, 1989), annotations by classmates or other readers might increase students’ interest in their writing tasks. The effects of annotated source materials on revision should be explored. A major limitation of the current study is that the very short time that students were given to write their essays was not conducive to planning or revision. However, the finding that, even under these artificial constraints, annotations seemed to influence students’ rhetorical activities is heartening. If the hypothesis that students use annotations to construct a representation of their readers is upheld with additional research, then such representations should be useful to students when revising. Several studies have suggested that providing student writers with information about their readers can improve their revisions (Rubin & O’Looney, 1990) and that such audience in-

326

WRITTEN COMMUNICATION / APRIL 2002

formation might be more important at the revision stage than at the drafting stage (Roen & Willey, 1988). The effects of exposure to annotated source texts over time should be explored. If online systems for sharing annotations become popular, the norm may once again be for readers to interact with annotated texts. Over time, students exposed to annotated materials might improve their abilities to predict the types of textual problems that readers will find confusing or objectionable (Schriver, 1992, found that exposure to think-aloud protocols of readers struggling with problematic texts improved students’ abilities to identify and to anticipate similar problems). Moreover, continued access to the annotations of multiple and various readers might help peripheral members of particular discourse communities to understand the assumptions and conventions motivating these communities. Reading as well as writing might come to be seen as a communal activity governed by particular norms and conventions. Continued exposure to annotated texts may have many potential benefits for readers and writers that researchers have yet to understand. APPENDIX The Letters to the Editor

The New York Times Company September 9, 1997, Tuesday, page 27 College Life vs. My Moral Code By Elisha Dov Hack; Elisha Dov Hack is a member of the Yale College class of ’01. Many people envy my status as a freshman at Yale College. My classmates and I made it through some fierce competition, and we are excited to have been accepted to one of the best academic and extracurricular programs in American higher education. I have an older brother who attended Yale, and I’ve heard from him what life at Yale is like. He spent all his college years living at home because our parents are New Haven residents, and Yale’s rules then did not require him to live in the dorms. But Yale’s new regulations demand that I spend my freshman and sophomore years living in the college dormitories. I, two other freshmen and two sophomores have refused to do this because life in the dorms, even on the floors Yale calls “single sex,” is contrary to the fundamental principles we have been taught as long as we can remember— the principles of Judaism lived according to the Torah and 3,000-year-old

Joanna Wolfe

327

rabbinic teachings. Unless Yale waives its residence requirement, we may have no choice but to sue the university to protect our religious way of life. Bingham Hall, on the Yale quadrangle known as the Old Campus, is one of the dorms for incoming students. When I entered it two weeks ago during an orientation tour, I literally saw the handwriting on the wall. A sign titled “Safe Sex” told me where to pick up condoms on campus. Another sign touted 100 ways to make love without having sex, like “take a nap together” and “take a steamy shower together.” That, I am told, is real life in the dorms. The “freshperson” issue of The Yale Daily News sent to entering students contained a “Yale lexicon” defining “sexile” as “banishment from your dorm room because your roommate is having more fun than you.” If you live in the dorms, you’re expected to be part of the crowd, to accept these standards as the framework for your life. Can we stand up to classmates whose sexual morality differs from ours? We’ve had years of rigorous religious teaching, and we’ve watched and learned from our parents. We can hold our own in the intellectual debate that flows naturally from exchanges during and after class. But I’m upset and hurt by this requirement that I live in the dorms. Why is Yale—an institution that professes to be so tolerant and open-minded—making it particularly hard for students like us to maintain our moral standards through difficult college years? We are not trying to impose our moral standards on our classmates or on Yale. Our parents tell us that things were very different in college dormitories in their day and that in most colleges in the 1950’s students who allowed guests of the opposite sex into their dorm rooms were subject to expulsion. We acknowledge that today’s morality is not that of the 50’s. We are asking only that Yale give us the same permission to live off campus that it gives any lower classman who is married or at least 21 years old. Yale is proud of the fact that it has no “parietal rules” and that sexual morality is a student’s own business. Maybe this is what Dean Richard H. Brodhead meant when he said that “Yale’s residential colleges carry . . . a moral meaning.” That moral meaning is, basically, “Anything goes.” This morality is Yale’s own residential religion, which it is proselytizing by force of its regulations. We cannot, in good conscience, live in a place where women are permitted to stay overnight in men’s rooms, and where visiting men can traipse through the common halls on the women’s floors—in various stages of undress—in the middle of the night. The dormitories on Yale’s Old Campus have floors designated by gender, but there is easy access through open stairwells from one floor to the next. The moral message Yale’s residences convey today is not one that our religion accepts. Nor is it a moral environment in which the five of us can spend our nights, or a moral surrounding that we can call home. Yale sent me a glossy brochure when it welcomed me as an entering student. It said, “Yale retains a deep respect for its early history and for the

328

WRITTEN COMMUNICATION / APRIL 2002

continuity that its history provides—a continuity based on constant reflection and reappraisal.” Yale ought to reflect on and reappraise a policy that compels us to compromise our religious principles.

The New York Times September 14, 1997, Sunday, page 14 Yale’s Jewish Students Keep Faith and Thrive To the Editor: While I respect Elisha Dov Hack’s choice not to be part of dorm life at Yale (Op-Ed, Sept. 9), I see no need for him to write about his disapproval of the Yale community. I’m an Orthodox Jew, too. I went through the same 12 years of training in Talmudic and rabbinical traditions that he did. I worked the same late nights that he must have worked to get accepted to Yale. Yet a crucial difference between us remains. I’m glad to be here. Mr. Hack’s dorming experience would have been the same as mine, because he was assigned to live next door to me in a six-person, single-sex suite. I wonder if he knows that three of his five prospective suite mates are Jewish, or that most of the people I’ve met on campus have confided that they have no intentions of having sex in the near future. Apparently he never took the opportunity to find these things out. A lot of Orthodox Jews have, though. There is Dan, who studies Talmudic law with me every morning, and Jesse, who organizes three prayer groups every day. There are Evan and Sara, who run the kosher kitchen, and many more people than I can count on two hands. Thanks to Yale for giving us the opportunity to grow without sacrificing our beliefs. SAUL NADATA New Haven, Sept. 10, 1997

The New York Times September 11, 1997, Thursday, page 30 Dormitory Life Is Essential to a Yale Education To the Editor: Re “College Life vs. My Moral Code” (Op-Ed, Sept. 9) by Elisha Dov Hack, who, along with four other Yale students, objects to living in a dormitory on the grounds of religious belief: Yale has profound respect for students who live with demanding convictions. Such students are found here in no small numbers, and we go to significant lengths to accommodate their concerns. At the same time, Yale College has its own rules and requirements, which we insist on because they embody our values and beliefs. One of Yale’s requirements is that undergraduates must live on campus in their freshman and sophomore years. This rule embodies our belief that what

Joanna Wolfe

329

students gain by living together is an essential aspect of their education. When students enter this community, their daily interaction becomes a continual scene of teaching and learning: a place to understand creeds and cultures different from one’s own, to appreciate the humanity of those who hold such beliefs, and to learn to work with others across lines of difference. To allow students to separate themselves from the full collegiate community would be to impoverish this aspect of the Yale education. As important, they would rob others of a chance to learn who they are and why their convictions require respect. Yale has a long history of working with students to accommodate their personal values, and we would happily explore accommodation in this case. RICHARD H. BRODHEAD Dean of Yale College New Haven, Sept. 10, 1997

The New York Times September 11, 1997, Thursday, page 30 Dormitory Life Is Essential to a Yale Education; Living in the Past To the Editor: There is a strong case to be made for exempting students with a preference for living off campus from Yale’s mandatory policy of housing first- and second-year students in the dormitories. However, Elisha Dov Hack (Op-Ed, Sept. 9) does not make that case persuasively. Having been raised an Orthodox Jew, I am aware of no prohibition against living in the same quarters as people who do not share your approach to sexual morality. Moreover, I would be surprised if other residents of Bingham Hall would care to convert Mr. Hack to their “anything goes . . . residential religion.” Apparently, according to Mr. Hack, a return to the policy of the 1950’s, during which “students who allowed guests of the opposite sex into their dorm rooms were subject to expulsion,” would demonstrate greater tolerance for Mr. Hack’s way of life. The irony of such an implicit claim is made only more stark by the fact that in the 1950’s, Yale College had a quota limiting the number of Jews it admitted and excluded women altogether. SHERRY F. COLB Newark, Sept. 9, 1997 The writer is an associate professor at Rutgers Law School.

330

WRITTEN COMMUNICATION / APRIL 2002

The New York Times September 11, 1997, Thursday, page 30 Dormitory Life Is Essential to a Yale Education; Piety as a Shield To the Editor: I understand Elisha Dov Hack’s concerns about the unbearable lightness of life in Yale’s dormitories (Op-Ed, Sept. 9); however, I am surprised that the only solution he is considering is to sue the university. If the Jewish teachings he mentions are established in his mind, the strength and rectitude of his feelings should shield him from any danger. As a last resort, he could always close his eyes. By the same token, I understand his religious beliefs may prevent him from eating certain types of food. Is he going to sue restaurants that serve unsuitable foods? You might think: He can choose not to eat in these restaurants. You then might think: He might choose not to go to Yale. LIONEL ANCELET Houston, Sept. 9, 1997

The New York Times September 11, 1997, Thursday, page 30 Dormitory Life Is Essential to a Yale Education; Laxity by Day To the Editor: Talmudically speaking, if Elisha Dov Hack (Op-Ed, Sept. 9) can manage to tolerate the moral laxity of Yale all day, why not at night? And if the depravity of the dorms makes it impossible for him to be there “in good conscience,” how can he traipse around such a “treif” (nonkosher) campus to his classes? It seems to me that a lawsuit is illogical and that he has no choice but to withdraw from Yale and enroll in a yeshiva. (Rabbi) TZVEE ZAHAVY New York, Sept. 9, 1997

The New York Times September 11, 1997, Thursday, page 30 Dormitory Life Is Essential to a Yale Education; Lawsuit Isn’t Answer To the Editor: Elisha Dov Hack (Op-Ed, Sept. 9) is simply wrong when he says that he and the other four students pressing for a waiver from Yale’s mandatory two-

Joanna Wolfe

331

year residence requirement for undergraduates “may have no choice but to sue the university to protect our religious way of life.” But his way of life is not being threatened. He has two choices. The first is to take an off-campus residence and maintain an on-campus residence on paper. While this is not a financially attractive option, should Yale have to subsidize the exceptions to its own policy? His other choice is to leave. DANIEL WALFISH New Haven, Sept. 9, 1997

REFERENCES Adler, M. J. (1967). How to read a book: The art of getting a liberal education. New York: Simon & Schuster. (Original work published 1940) Bedford/St. Martin’s. (1999). Solicited masked reviews of the second and third editions of The presence of others. Unpublished manuscript. Black, K. (1989). Audience analysis and persuasive writing at the college level. Research in the Teaching of English, 23, 231-253. Bush, V. (1945, July). As we may think. Atlantic Monthly, 176, 101-108. Cashen, V. M., & Leicht, K. L. (1970). Role of the isolation effect in a formal educational setting. Journal of Educational Psychology, 61, 484-486. Cohen, M., & Riel, M. (1989). The effect of distant audiences on students’ writing. American Educational Research Journal, 26, 143-159. Crouse, J. H., & Idstein, P. (1972). Effects of encoding cues on prose learning. Journal of Educational Psychology, 63, 309-313. Farkas, D. K., & Poltrock, S. E. (1995). Online editing, mark-up models, and the workplace lives of editors and writers. IEEE Transactions on Professional Communication, 38, 110-117. Fleiss, J. (1981). Statistical methods for rates and proportions. New York: John Wiley. Flower, L. (1990). The role of task representation in reading-to-write. In L. Flower, V. Stein, J. Ackerman, M. J. Kantz, K. McCormick, & W. C. Peck (Eds.), Reading-towrite: Exploring a cognitive and social process (pp. 35-75). New York: Oxford University Press. Fowler, R. L., & Barker, A. S. (1974). Effectiveness of highlighting for retention of text material. Journal of Applied Psychology, 59, 358-364. Greene, S. (1991). Writing from sources: Authority in text and task (Technical Report No. 55). Pittsburgh, PA: Carnegie Mellon University. Haas, C., & Flower, L. (1988). Rhetorical reading strategies and the construction of meaning. College Composition and Communication, 39, 167-184. Haswell, R. H., Briggs, T. L., Fay, J. A., Gillen, N. K., Harrill, R., Shupala, A. M., et al. (1999). Context and rhetorical reading strategies: Haas and Flower (1988) revisited. Written Communication, 16, 3-28. Hawhee, D. (1999). Composition history and the Harbrace college handbook. College Composition and Communication, 50, 504-523.

332

WRITTEN COMMUNICATION / APRIL 2002

Hays, J. N., & Brandt, K. S. (1992). Socio-cognitive development and students’ performance on audience-centered argumentative writing. In M. Secor & D. Charney (Eds.), Constructing rhetorical education (pp. 202-229). Carbondale: Southern Illinois University Press. Itano, N. (1997, October 31). Lawsuit draws little support from student body. Yale Daily News Online. Retrieved July 9, 1999, from http://www.yale.edu/ydn/paper/ 10.31.97/t-2support.html Jackson, H. J. (2001). Marginalia: Readers writing in books. New Haven, CT: Yale University Press. Lunsford, A. A., & Ruszkiewicz, J. J. (1999a). Everything’s an argument. New York: Bedford/St. Martin’s. Lunsford, A. A., & Ruszkiewicz, J. J. (1999b). The presence of others (3rd ed.). New York: St. Martin’s. MacDonald, R. (1991). Developmental students’ processing of teacher feedback in composition instruction. Review of Research in Developmental Education, 8. Abstract retrieved May 30, 2001, from ERIC database. Marshall, C. C. (1997). Annotation: From paper books to the digital library. In Digital libraries 1997 (pp. 131-140). New York: Association for Computing Machinery. Marshall, C. C. (1998). Toward an ecology of hypertext annotation. In Hypertext98 (pp. 40-49). New York: Association for Computing Machinery. Marshall, C. C., Price, M. N., Golovchinsky, G., & Schilit, B. N. (1999). Introducing a digital library reading appliance into a reading group. In Digital Libraries 99 (pp. 77-84). New York: Association for Computing Machinery. Nystrand, M. (1990). Sharing words: The effects of readers on developing writers. Written Communication, 7, 3-24. Penrose, A. M. (1992). To write or not to write: Effects of task and task interpretation on learning through writing. Written Communication, 9, 465-500. Ramage, J. D., & Bean, J. C. (1995). Writing arguments. Boston: Allyn & Bacon. Redd-Boyd, T. M., & Slater, W. H. (1989). The effects of audience specification on undergraduates’ attitudes, strategies, and writing. Research in the Teaching of English, 23, 77-108. Reder, L. M. (1985). Techniques available to author, teacher, and reader to improve retention of main ideas of a chapter. In S. Chipman, J. Segal, & R. Glazer (Eds.), Thinking and learning skills: Current research and open questions (Vol. 2, pp. 37-64). Hillsdale, NJ: Lawrence Erlbaum. Roen, D. H., & Willey, R. J. (1988). The effects of audience awareness on drafting and revising. Research in the Teaching of English, 22, 75-88. Rottenberg, A. T. (2000). Elements of argument (6th ed.). New York: Bedford/St. Martin’s. Rubin, D. L., & O’Looney, J. (1990). Facilitation of audience awareness: Revision processes of basic writers. In G. Kirsch & D. H. Roen (Eds.), A sense of audience in written communication (pp. 280-292). Newbury Park, CA: Sage. Salvatori, M. (1996). The “argument of reading” in the teaching of composition. In B. Emmel, P. Resch, & D. Tenney (Eds.), Argument revisited; argument redefined: Negotiating meaning in the composition classroom (pp. 181-197). Thousand Oaks, CA: Sage. Schriver, K. A. (1992). Teaching writers to anticipate readers’ needs: A classroomevaluated pedagogy. Written Communication, 9, 179-208. Schumacher, G. M., & Nash, J. G. (1991). Conceptualizing and measuring knowledge change due to writing. Research in the Teaching of English, 25, 67-96.

Joanna Wolfe

333

Van Dam, A. (1988). Hypertext ’87 keynote address. Communications of the ACM, 31, 887-895. Whalley, G., & Jackson, H. J. (Eds.). (1980). Marginalia (Vol. 12). Princeton, NJ: Princeton University Press. Wollstonecraft, M. (1975). Maria; or the wrongs of woman. New York: Norton. (Original work published 1798) Wood, N. V. (1995). Perspectives on argument. Englewood Cliffs, NJ: Prentice Hall.

Joanna Wolfe recently left graduate school at the University of Texas at Austin to join the faculty at the University of Louisville. The current manuscript is excerpted from her dissertation Pedagogical Uses of Annotations and Annotation Technologies. Previous work by Joanna Wolfe has appeared in Computers and Composition, Written Communication, and Journal for Business and Technical Communication.