Political scientists commonly stress the importance of good writing. Editors of journals such as PS: Political Science and Politics state that they seek clear, compelling writing in reviewing manuscripts.Footnote 1 Reviewer guidelines routinely ask referees to assess articles in terms of the quality of writing. Editors may reject manuscripts if the presentation is perceived to be poor, even if the topic is important and the findings interesting.
We also tend to emphasize high-quality writing to our students. Indeed, there are indications that political scientists may be taking on more responsibility in this area as a component of university writing across the curriculum efforts, responding in part to faculty frustration at the quality of writing they encounter (Sherman and Waismel-Manor Reference Sherman and Waismel-Manor2003).
However, it is unclear whether political scientists and political science students share many of the same norms about what constitutes good writing. As we began this research project we speculated that one potential area of conflict concerned the use of the active and passive voice, most especially the acceptability of the pronoun “I” in academic work. The lead author was motivated in large part by reviewing scores of master's theses. In their initial drafts, master's students commonly went out of their way to avoid using “I.” This resulted in extensive use of passive voice—not a logical necessity (one can write in the active voice without using the personal pronoun) but true in practice. More important, refraining from using “I” often led to lack of clarity about the students' own hypotheses and findings, as opposed to those in the literature. This prompted a number of questions. Did colleagues share the lead author's views about the appropriateness of “I” in at least some circumstances? Were students generally as averse to personal pronouns as we perceived? If there was a divergence, why did it exist? And what were the larger implications for how to communicate with students about writing? This article attempts to answer these questions as part of the authors' ongoing work related to voice in academic disciplines. It draws on a review of political science articles over a long period, a survey of students at one campus, and interviews with faculty at that campus.
The article is organized as follows. First, we consider what the small body of literature on voice in academic writing suggests about the acceptability of using “I.” Next, we move to more systematic evidence of what political scientists consider acceptable. Third, we draw from our student survey to gain insight into student views. Fourth, we draw from interviews and discussions with faculty that help us identify the subtleties that underlie the other data. Last, we draw implications for what should be communicated to students.
REVIEW OF THE LITERATURE
Although only a handful of studies have focused on the use of the personal pronoun in academic writing, researchers who have studied pronoun usage have pointed to the importance of this work in understanding how academic writers construct their identities. As Hyland (Reference Hyland2002) argues in “Options of Identity in Academic Writing,” “a writer's identity is created by, and revealed through, the use or absence of the ‘I’ pronoun” (352). Tang and John (Reference Tang and John1999) feel that the value of focusing on first-person pronouns is that they are “arguably the most visible manifestation of a writer's presence in a text” (S23). The importance of investigating the use of first person in academic writing is further supported by recent research that suggests a growing trend away from traditional conceptions of academic writing as impersonal (Clark Reference Clark and Fairclough1992; Ivanic Reference Ivanic1994; Ivanic and Simpson Reference Ivanic, Simpson and Fairclough1992; Lillis Reference Lillis1997).
Researchers who have investigated the use of active voice in academic writing have used three primary research methods: corpus studies of student or professional texts, qualitative research such as interviews and case studies, or a mix of these two approaches. Harwood (Reference Harwood2006) argues that although corpus studies
effectively highlight disciplinary similarities and differences, they can give the reader the impression that (a) there is a consensus within each discipline concerning the (non)acceptability of I and we and (b) such practices in each discipline are stable, when neither of these is necessarily the case.
(425)Harwood recommends the use of tools such as surveys or interviews to investigate the writer's perspective and provide a complement to corpus studies. Little research, of any type, has investigated the use of personal pronouns in the social sciences, and Harwood's (Reference Harwood2007) study of political scientists at a British university is the only study of the use of personal pronouns in political science that we were able to locate.
Researchers have commented on students' reluctance to use “I” in academic writing and their belief that the inclusion of personal pronouns and their own opinions is inappropriate (Hyland Reference Hyland2002). This reluctance is true of graduate students as well. For example, Swales (Reference Swales1990) found that the graduate students across disciplines in his study did not like to use “I” and felt that use of personal pronouns is too informal and should be reserved for senior scholars.
In contrast to students' perceptions of the inappropriateness of personal pronouns, Hyland (Reference Hyland2002) found that subject teachers in the social sciences “commented on the absence of any real voice or presence behind the papers they marked, and expressed frustration at students' general reluctance to commit themselves” (354). Hyland argues that by avoiding personal pronouns, student writers may fail to create an “authorial identity” and a successful academic argument (354). In an earlier study, Hyland (Reference Hyland2001) conducted a corpus analysis of eight disciplines, collecting and analyzing three scholarly articles from 10 leading journals in each field, with sociology and applied linguistics representing the social science fields in his research. Hyland followed up with interviews with the authors of the articles. In the sociology articles, Hyland found an average of 29.4 uses of “I” per paper, and 32.3 uses of “I” per paper in the applied linguistics articles (213). In interviews, authors mentioned authorial persona and personal engagement as two primary purposes for using personal pronouns.
Harwood's (Reference Harwood2007) qualitative study of five political scientists also demonstrated frequent use of personal pronouns. Harwood found that all of the informants used “I” in their writing. The political scientists in Harwood's study believed that their use of “I” made readers feel included in their arguments, heightened the rhetorical effects of their arguments, got readers to support the writer's position, made empirical research easier to read, and helped emphasize the scholarly contribution of the research.
Again, studies have shown that unlike many scholars, students are commonly reluctant to use the personal pronoun. Some students also mistakenly believe that writing in the first person is prohibited by common style guides, such as those used by the American Psychological Association (McAdoo Reference McAdoo2009). This disconnect between the beliefs of apprentice scholars in a discipline and the actual expectations of disciplinary instructors and practice of expert writers has been noted by several researchers who have studied student writers making the transition to upper-division coursework (Caroll Reference Caroll2002; Herrington and Curtis Reference Herrington and Curtis2000; Sternglass Reference Sternglass1997). The research shows two major factors in students' struggles to enter the discourse community of their discipline: (1) lack of knowledge of the norms of the discipline and (2) lack of explicit discussion of disciplinary conventions by instructors. In the case of using personal pronouns in political science, this issue appears to be especially complex considering students' prior beliefs about what is appropriate and inappropriate for novice political scientists as well as the evolving nature of the use of pronouns in the professional discourse.
MORE EVIDENCE OF WHAT IS ACCEPTABLE TO POLITICAL SCIENTISTS
Based on our review of the literature and impressions from many years of reading relevant journals, we began with the assumption that political scientists increasingly found it acceptable to use the pronoun “I” in academic writing. To more systematically assess whether this impression was accurate, we reviewed single-authored American Political Science Review (APSR) abstracts from 1980 through 2011. We chose to examine the APSR because it is an official publication of the American Political Science Association, is generally considered the most prestigious political science journal, and regularly tops or nearly tops the list of journals having the most citation impact using a variety of methods (Garand and Giles Reference Garand and Giles2003; Giles and Garand Reference Giles and Garand2007). We focused on abstracts because they could be assessed quickly.
Figure 1 summarizes our key findings. This figure shows the proportion over time of single-authored APSR abstracts within a particular volume that use “I” at least once. The key point is that the portion of such abstracts using the first person pronoun rose steadily through the 1980s and has remained relatively high since that decade. Although the pattern has varied considerably from year to year, in general more than half of the single-authored abstracts include “I” and, in some years, more than two thirds do so. This evidence understates the acceptability of the first-person pronoun because we found instances in which published articles used “I” in the body of the text but not in the abstract.
Figure 1 Percent of Single Authored APSR Abstracts Using “I”
The evidence, therefore, indicates that use of “I” is both common and acceptable in the most prestigious disciplinary journal. Moreover, while we did not engage in systematic analysis of other journals, we have no reason to believe the APSR is unique in sanctioning the first-person pronoun. Nor do recent APSR instructions to contributors offer specific guidance with respect to writing style, much less whether it is permissible to write in the first person, which suggests that authors are not writing in the first person simply as a result of cues from editors. All of this indicates a norm is evolving that it is fine to make at least limited use of “I” (which is not the same as contending that frequent use of the personal pronoun is desirable; anecdotal evidence suggests that many writers are sparing in such use). Note that the volumes we reviewed covered multiple APSR editorial regimes.
We offer a few words of caution about these findings. The fact that political scientists generally find use of “I” acceptable in their own published work does not necessarily imply that they find it acceptable in all student work—a point to which we will return. Additionally, nothing about our review of APSR abstracts suggests that the norms political scientists follow in their own writing are effectively communicated to students.
WHAT IS ACCEPTABLE TO STUDENTS?
To gain a more precise understanding of how undergraduate political science students assess choices about voice, and why, we conducted a written survey of a set of 2010 summer session students at California State University, Sacramento (CSUS), a large, urban, comprehensive university. The students were enrolled in six upper-division political science classes (or “government” courses as they are labeled at CSUS), such as “International Politics” and “State and Local Government,” taught by six full-time faculty members. We surveyed the students at the beginning of a single class period after a brief explanation of our interest in students' views about good writing. A total of 131 students responded to the written questionnaire.
The survey presented students with two versions of an abstract of an article appearing in a recent volume of the APSR (see Appendix). One version (Abstract B in the Appendix) was the actual abstract used by Zolton Hajnal (Reference Hajnal2009) for his article “Who Loses in American Democracy? A Count of Votes Demonstrates the Limited Representation of African Americans.” That abstract was written in the active voice and included a few instances of the first-person pronoun; we chose it partly for those reasons and partly because we thought the subject matter of the article might be of interest to students.Footnote 2 The other version (Abstract A in the Appendix) was one we rewrote in the passive voice, avoiding use of the first-person pronoun. Sometimes Abstract A appeared at the top of the survey and Abstract B appeared after that, while other times the order was reversed.Footnote 3 At the bottom of the survey we asked respondents which abstract they believed was more effectively written and why. We made no mention of which version appeared in the published article. Our survey is appended to the end of this article.
The students were quite divided in terms of which abstract they considered more effectively written, but they leaned toward the passive voice. More than 50% preferred the abstract avoiding the use of the first person, about 39% preferred the version that included “I”, and 11% believed the two versions were equally effective.
We identified some patterns with respect to the reasons for student choices. Students who preferred the version without “I” often cited reasons involving the inappropriateness of the first person. They tended to emphasize that the use of “I” made the abstract overly personal, distracting, or biased. Students who preferred the version using the personal pronoun tended to stress such considerations as its being more direct.
Verbatim survey comments underscore these themes. The majority of respondents who preferred the passive voice version without “I” wrote statements including the following:
-
• Abstract A [passive voice version] is professionally written while Abstract B sounds like a high school essay.
-
• The author shouldn't personalize by saying “I” all the time in Abstract B [active voice version].
-
• The use of I in Abstract B [active voice version] makes me question the credibility of the writer.
-
• Abstract A [active voice version] uses “I.” I am not used to reading first person abstracts. While it sounds like findings for a research or a study it sounds less academic than Abstract B.
-
• Abstracts shouldn't be in the first person.
-
• 3rd person presentation is more professional.
By contrast, respondents who believed the abstract that included personal pronouns was more effective wrote these statements:
-
• It's easier to understand when written in first person.
-
• Too many unnecessary words in Abstract A [passive voice version].
-
• Abstract B [active voice version] talks about the information in the first person. It is more direct.
-
• [The active voice version] is more simple, easy to digest and the important thing is clear for the reader to avoid misinterpretation [emphasis in original].
Although some respondents had well-developed views about the abstracts, many responses were lacking in depth. The strongest such evidence is the difference in responses to the varying versions of the survey. If respondents had well-considered views of writing standards we would not expect significant divergence in ratings based simply on the order in which the versions were presented. Yet that was the case. Excluding the small portion of respondents who rated both abstracts equally, 46% rated the version without personal pronouns as more effective when presented first in the survey while 78% did so when it was presented second; the difference is statistically significant at the .01 level using a chi-square test.
In short, respondents were divided in their views of the abstracts. Although many responses lacked depth, overall the students' views leaned away from the reasons for supporting use of first person expressed by several political scientists. This is provocative and worthy of further study.
A FACULTY PERSPECTIVE ON STUDENT USE OF VOICE
If the students in our study tended to see the use of “I” as unprofessional and inappropriate for academic writing in political science, those beliefs may be often reinforced by the instruction they are receiving from their political science professors. Most of the CSUS faculty we interviewed expected students to write in a formal, objective style for most of the writing assigned in their courses. For some faculty this is a reflection of what they believe are the conventions of the discourse of the discipline. One faculty member said of writing in political science, “To use the personal pronoun is generally discouraged,” and this was emphasized again by another who said, “I haven't seen a lot of political science writing that uses ‘I.’” Another faculty member said it “stood out when people use ‘I’ or ‘we’ in scholarly writing.” One faculty member confessed that the conventions of political science writing are “empirical, objective, boring, dry, painful to read.” The valuing of avoiding “I” among these political science faculty members, however, is only partially a result of a perception that the use of the personal pronoun is not common in academic writing in the discipline.
Perhaps a more important reason these faculty discourage the use “I” has to do with problems students have distinguishing between analysis and opinion. As one faculty member says, “In my classes I tell them the papers they are doing are not opinion pieces, they are analytical pieces.” This faculty member makes “a sharp distinction between personal views and analysis.” In a research methods course, another faculty member tells students, “You need to be objective, this is not about your personal opinion.” When the lead author presented the results of our research at a department colloquium, faculty emphasized that students tend to be too personal in their writing, and instructors want to move them away from that by discouraging use of the first person, even if they think the first person is appropriate in some contexts.
Despite the emphasis on objective analysis in most of their courses, a few faculty members acknowledged the changing nature of the scholarly voice in political science, as reflected in our analysis of APSR abstracts. One faculty member commented, “Within the literature that I've seen, ‘I’ is much more common.” Another said, “Increasingly I'm seeing people using the personal pronoun ‘I’ compared to 10 years ago. The trend is to be more relaxed in terms of the academic standards related to the use of ‘I.’” This faculty member said that he often uses “I” when writing for the general public. Another faculty member emphasizes storytelling and making personal connections in students' writing.. This faculty member said, “I find it much more interesting when people can connect and give me stories.” He said that he got into academia because he “felt [he] had a story to tell.” Another faculty member felt that “political science could benefit from more use of ‘I.’” Based on our interview responses, there is little consensus among the faculty on the degree to which the use of “I” is accepted in academic writing in political science, and whether or not the traditional scholarly voice is “professional” or “boring” and “painful to read.”
Although most of the faculty members do not often ask for personal examples and storytelling from their students, all of the faculty members are aware of the role audience, purpose, and genre plays in authors' decisions about whether to use personal pronouns. Even a faculty member who was taught in his graduate training never to use “I” told us that if he was writing for a broader audience or doing case study work he would use “I.” This faculty member said, “For a class paper it's not the same as a journal article. Writing depends on who your audience is and what purpose it is.” Another faculty member echoed this sentiment when he said he would use “I” for op-ed pieces and writing for the general public but not in a scholarly journal. The importance of context in dictating writing style was a point the faculty reemphasized at the department colloquium.
These two issues—the conflicted nature of conventions of scholarly voice in political science and the importance of context—seem to be concerns that students may not be aware of, given prior research and our own limited study of students' responses to political science abstracts. However, the faculty we interviewed face several challenges in bringing students to a more sophisticated level of rhetorical awareness. As one faculty member admitted, “Students are being introduced to scholarly writing and don't know anything about it.” However, he added: “I don't really focus on style.” He was not unusual in that regard: most of the faculty did not include style as part of their grading rubrics or make it a focus of their written comments on student papers, and none of the faculty mentioned any explicit discussions in class regarding the conflicted nature of the use of “I” in scholarly writing in political science or the way the use of “I” may vary depending on the genre, audience, research method, and so forth.
We acknowledge several challenges that prevent faculty from focusing as much as they might like to on writing style. As one faculty member said, “There's a lot of time pressure in the classroom. The more time you take teaching them to write, the less material you can teach them. There is so much material to get to. You have to talk about all of American government in a semester.” Other faculty members commented on the need to focus on the quality of content and analysis in student writing, and on the time-consuming nature of responding to student writing and the need to choose your battles. One faculty member said this about students and academic writing: “There are classes where they should learn how to do that. We're not all writing professors.” This was echoed by several other faculty we interviewed.
CONCLUSIONS AND IMPLICATIONS
Our research speaks to the complexity of teaching students who may have static notions of the academic discourse conventions of political science when those conventions are in a state of flux. It appears that the students' views contrast with journal norms as well as the preferences for active voice expressed by the political scientists in Harwood's (Reference Harwood2007) study and by many of our faculty interviewees who emphasized the importance of purpose, audience, and genre. Our research points to a possible disconnect between students' and instructors' beliefs about the appropriateness of the use of the personal pronoun in political science writing—a disconnect that likely exists at many academic institutions, in light of Hyland's (Reference Hyland2002) and Swales' (Reference Swales1990) findings that undergraduate and graduate students alike believe “I” is inappropriate in academic writing. Whether it was incorrect genre “rules” (“Abstracts shouldn't be in the first person”), outdated conceptions of authorial persuasiveness (“The use of the pronoun ‘I’ weakens the statement”), or mistaken notions about what counts as professionalism (“third-person presentation is more professional”), a majority of the students in our research lacked an understanding of what appear to be political science conventions when it comes to the use of personal pronouns, at least in scholarly work.
Given the conflicting views of the faculty on the use of first person, students at the institution in our study may receive conflicting messages about the use of first person depending on the instructor's background and beliefs about what counts as good writing in political science. These conflicting messages may be a natural result of a field that is undergoing a change in discourse conventions. Without an understanding of the nature of this change, and its implications for the discipline's conception of authorial identity and how knowledge is produced, students may rely on prior, stereotypical beliefs about academic writing or get the mistaken impression that good writing in political science is simply whatever the individual instructor believes it to be. Our research has several implications for the teaching of writing in political science. Political science instructors' knowledge of writing conventions, such as the appropriateness of the use of the personal pronouns, has been learned over years of reading and writing in the field. Students entering a field, however, are likely to lack this awareness of rhetorical conventions, as prior research has shown (Caroll Reference Caroll2002; Herrington and Curtis Reference Herrington and Curtis2000; Sternglass Reference Sternglass1997). In the case of a convention like the use of the personal pronoun, which has undergone a shift in the discipline, students may be especially uninformed, relying on generalizations about academic writing learned in high school or in college composition courses.
Political science instructors can demystify disciplinary conventions for their students by making these conventions explicit in their writing assignments, rubrics, and responses to student writing. In the case of the personal pronoun, political science instructors can work to make students aware of the shift in the field and perhaps the differing opinions about it, as well as the reasoning behind this shift. Instructors can also ask students to write in a variety of genres and a range of styles, and invite students to reflect on the different rhetorical choices they make. The difference between making one's own perspective readily apparent is more than just a stylistic preference. Underlying this shift is a deeper consideration of the role of the writer's identity and a rethinking of the relationship between writer and audience—a topic we continue to explore in our own scholarly work. It is understandable that political science instructors feel there is simply not enough time to work closely with students on their writing, whether that means working on appropriate voice or other writing issues. However, political science instructors (and instructors in any academic discipline) cannot assume that students will come to them with the level of rhetorical awareness necessary to succeed in writing for their specific academic discipline. Required general education composition courses may introduce students to broad conventions of academic writing, but no single composition course can delve into the complexities of disciplinary writing issues such as the one our study explores.
Despite the many constraints on their time, if political science instructors wish to initiate students to the ways of thinking and communicating in their discipline, they would be well served to demystify these conventions. Without explicit instruction in what counts as “good writing” in political science—and explicit talk about how what counts as good writing changes over time and in various contexts—students operate on myths and guesswork rather than a growing understanding of the ways political scientists communicate in their discipline.
ACKNOWLEDGMENT
We are grateful to CSUS department of government faculty for administering the survey, participating in interviews, and offering helpful feedback regarding our preliminary findings.
APPENDIX
Note : In the survey as administered we randomly reversed the order of the following two abstracts.
Student Survey Regarding Wording of Political Science Abstracts
Please review the following two abstracts (i.e., summaries) of research published recently in The American Political Science Review:
Abstract A. It has long been feared by critics that America's winner-take-all electoral system would undermine the interests of minorities. Unfortunately, few available tests broadly assess how well minorities fare in a democracy. To gauge winners and losers in the American case, in the following study a new measure of representation was introduced. For any election, the number of voters from each demographic group who vote for a candidate that loses was counted. After a comparison of this new measure to its alternatives, data from the entire series of Voter News Service exit polls and a sample of mayoral elections were used to determine which kinds of voters end up losers. It was found that across the range of American elections, African Americans are consistently more likely than other groups to end up losers, raising questions about equity in American democracy. The one exception to the pattern of black failure-congressional House elections-suggests ways to better incorporate minority interests.
Abstract B. Critics have long feared that America's winner-take-all electoral system would undermine the interests of minorities. Unfortunately, few available tests broadly assess how well minorities fare in a democracy. To gauge winners and losers in the American case, I introduce a new measure of representation. For any election, I count the number of voters from each demographic group who vote for a candidate that loses. After comparing this new measure to its alternatives, I use data from the entire series of Voter News Service exit polls and a sample of mayoral elections to determine which kinds of voters end up losers. I find that across the range of American elections, African Americans are consistently more likely than other groups to end up losers, raising questions about equity in American democracy. The one exception to the pattern of black failure-congressional House elections-suggests ways to better incorporate minority interests.
Which abstract do you think is more effectively worded? (Please check the appropriate box)
[ ] Abstract A is more effectively worded.
[ ] Abstract B is more effectively worded.
[ ] The abstracts are equally well worded.
Please explain why you think so.
