November 2018 – Volume 22, Number 3
Tohoku Gakuin University
This paper reports on results of empirical research investigating written summaries of first-year Japanese university students, their task representations of summary writing, and the relationships between their representations and task performance. Data for this study were obtained by asking 108 students to read an English text and write a summary of it. Afterwards, the students answered survey questions which examined their task representations of English summary writing. The results reveal the following: 1) High copy rates with an average copy rate of 72.2 % (SD 20.5%) were found, 2) Students did not seem to think that writing in their own words is required, 3) Students tended to look at the source text and use expressions from it in their summaries, 4) Positive correlations between copy rate and students’ task representations regarding the use of expressions from the source were found, and 5) Negative correlations between copy rate and task representations regarding the use of their own words were also found. Though various sources of student plagiarism have been found and discussed in previous research, this paper attempts to propose task representations of summary writing as another possible source of excessive copying, which is regarded as plagiarism in English-speaking countries.
In a conversation between this writer and instructors of Japanese to international students, one of the instructors remarked that she tells students to write a summary by combining key words and sentences from the source text rather than by using their own words. This conversation made this writer realize that there could be a difference in how summarizing is represented between Japan and majority of English-speaking countries. While writing in the writer’s own words is taken for granted in academic communities in English-speaking countries, it is rarely taught in Japanese education. The current paper examines if Japanese university students’ task representations, that is, how they perceive a given task of summary writing, may contribute to their tendency toward excessive copying from the source text, which was termed “language plagiarism” by Pennycook (1996, p. 223).
Over the past decades, various reasons for ESL students’ plagiarism have been identified and discussed, such as different concepts of learning across cultures (Ballad & Clanchy, 1991; Pennycook, 1996), ESL students’ lower language proficiency (Keck, 2006), and the different educational backgrounds of ESL students (Shi, 2006, 2008). In these settings, Chinese students were often used as participants in empirical studies. Based on such studies, it is often said that copying is perceived as a legitimate means of learning and citation in Asia. Wheeler (2009), however, protested against this perception and argued that lack of knowledge of what comprises plagiarism, rather than culture, is the main factor behind Japanese university students’ forgiving attitudes toward copying. The current paper proposes another possible source of plagiarism: that is, some students’ task representations of summary writing. Wheeler’s point is expanded, and it is proposed that copying may occur in Japanese university students’ papers, especially in introductions and literature review sections because of the students’ task representations of summary writing. In introductions and literature review sections, summarization is often used as a subskill for integrating other writers’ texts. According to Flower, Stein, Ackerman, Kantz, McCormick, and Peck (1990), “task representation” is how an individual interprets a given task, and it may include how he or she represents the rhetorical context, what goals he or she attempts to accomplish, and what strategies he or she employs (p. 38). Although task representations that an individual generates influence the whole process of doing the task, they differ from individual to individual and possibly from culture to culture.
This paper presents the results of empirical research investigating copy rates of Japanese students’ English summary samples, their task representations of summary writing, and the relations between the summaries and task representations. The results of this research are significant because little empirical research has been conducted on the relationships between Japanese university students’ copying behavior and their task representations of summary writing, and more importantly, because they may shed light on the causes of some Japanese university students’ occasional plagiarism.
Review of Literature on Student Plagiarism
Student plagiarism has been a major issue in academic communities in the majority of English-speaking countries for many decades, and a common assumption has been that it occurred as a result of students’ lack of ethics, or their ignorance of citation rules. Since 1990s, however, an increasing number of research articles have been published on plagiarism, and now the issue is viewed as a complex problem with various causes, reasons, and interpretations. Plagiarism can be categorized into different types in accordance with the degree of textual borrowing and the intention to deceive. For example, Howard (1995) classifies it into the following three types: “cheating” which means handing in someone else’s work as one’s own, “non-attribution” which means that some part of the work is copied from someone else’s text without acknowledgement or quotation marks, and “patchwriting” which indicates copying parts of someone else’s text with only minor changes (p. 799). Various sources of student plagiarism have been found, such as cultural differences in values, different perceptions of learning between cultures, students’ insufficient language proficiency, inherent difficulties in academic writing, different expectations between secondary and post-secondary education, and different individual interpretations regarding appropriate citations.
Plagiarism cases which involve Asian students have been attributed mainly to Asian cultural values, which, some believe, accept or even emphasize memorization and copying. For example, Pennycook (1996) notices that memorization and copying are legitimate learning strategies in China though they might be viewed as plagiarism in Western cultures. Ballad and Clanchy (1991) contrasted Asian and Western approaches to learning and argued that memorization and imitation were the major learning strategies in Asia. Rinnert and Kobayashi (2005) conducted surveys in Japan and America, and showed that Japanese students had a more welcoming attitude toward copying than American students.
Wheeler (2009), however, questioned the common belief that copying was acceptable in Japanese culture, and demonstrated that Japanese students do have a negative view of it. In his study, Japanese university students’ reactions to copied and paraphrased texts are examined. Students’ evaluation scores of the copied text went lower after they learned that the text had been copied. Similarly, they gave higher evaluations of the paraphrased text when they believed the text had been written in the writer’s own words. Based on the students’ reactions, Wheeler concluded that Japanese university students do have negative attitudes toward copying and thus he argued against the view that plagiarism was a cultural matter.
Review of Literature on Summarizing and Paraphrasing in English-speaking Countries
Task demands and cognitive processes involved in summary writing have been investigated, and important findings have been reported in English-speaking countries. The findings include the underlying cognitive operations, the subjectivity to various factors, and the developmental trend in skills and strategies students employ. For example, van Dijk and Kintsch (1977) proposed a summary model which involves deletion (omission of trivial and redundant information), generalization (abstraction of lower level concepts), and construction (integration of details and creation of macrostructure). According to Hidi and Anderson (1986), summarizing requires evaluating the text information, selecting important ideas, and combining them coherently. In some cases, they argue, the writer should substitute a higher level concept for those at lower levels, and create an overall structure of his or her own comprehension of the text.
Hidi and Anderson (1986) also suggested that characteristics of the source text, the task procedures, and type of the summary the writer attempts to generate greatly influence the cognitive processes involved and the finished products. Characteristics of a source text include length, genre, and complexity. One of the most important procedural aspects of the task may be the presence or absence of the source text while summarizing. The former reduces the memory load on the writer because he or she does not have to keep information in their working memory while conducting the summary task, but it may encourage direct copying. On the other hand, the latter places increased memory load because of handling both processes simultaneously, but may encourage more active processing of the information. The types of the summary a writer attempts to create can be classified into “a writer-based summary” or “a reader-based summary”. (p. 479) “A writer-based summary” is a summary for helping the writer monitor and facilitate his understanding of the source text. “A reader-based summary” is written for outside reading, and consequently it should be more refined and should require attention to various constraints, making this type more difficult to write than the writer-based summary.
Brown and Day (1983) demonstrated that summary strategies develop as students mature. In their study, while immature students adopted “copy-delete” or “knowledge telling” (p. 13) strategies, older students tended to use more sophisticated strategies such as more effective condensation and paraphrasing. As a result, less verbatim copying and more drastic formal changes were found in summaries of the older students. “Copy-delete” strategy means deleting unimportant parts and copying the remaining parts word for word. “Knowledge-telling” strategy indicates that the writer does not consider the rhetorical purpose, and just relates what he or she knows about the topic. In their research, however, they also found “copy-delete” or “knowledge-telling” in some older students’ summaries and maintained that these immature strategies may persist unless they are clearly rejected. Sarig (1993) also contended that explicit instruction will be necessary for the improvement of summary skills.
Other lines of research which are relevant to the current discussion are studies examining effects of paraphrasing on text comprehension and research on academic discourse. Empirical research on effects of paraphrasing has shown that writing a summary in the writer’s own words deepens the writer’s understanding of the source text (e.g., Karbalaei & Amoli, 2011; Katims & Harris, 1997; McNamara, 2004), and therefore paraphrasing ability in summarization is believed to show the summary writer’s profound understanding of the source text (e.g., Axelrod, Cooper, & Warriner, 2008; Hirvela & Du, 2013). In addition, research on academic discourse has suggested that the definitions of university-level summarizing and paraphrasing are not what are commonly assumed from the terms. Basham, Ray, and Whalley (1993) state that in university-level reading-to-write tasks, students are required to “move beyond summarization to what Flower et al. refer to as ‘translation’” (p. 304) which exhibits the student’s critical reading and analysis of the source text. Yamada (2003) also suggests that expectations for university level paraphrase tasks are not “a faithful reproduction of the ideas in source text,” but “one that expresses the paraphrase writer’s ingenuity” (p. 252). Thus, in summaries and paraphrases, which are used as subskills in integrating sources, the writer’s interpretation of the source text should be reflected. Hirvela and Du (2013) wrote as follows: “An important marker of a scholar’s (or student’s) understanding of a source text, especially a complicated statement by the original author(s) of the source text, is the ability to find a new way to capture the gist of what was stated in the original passage” (p. 88). Such paraphrasing ability is required in academic writing, and failure to exhibit such an ability may result in accusations of plagiarism.
While the importance of writing in one’s own words may be conveyed to university students in English-speaking countries, it is a difficult concept to get across to some students. Therefore, some textbooks propose strategies to facilitate understanding of the concept. For example, Howard (2010) recommends that students write a summary “without looking at the source” to avoid copying (p. 269). Kennedy and Smith (2006) encourage students to “create a graphic overview” because it helps the writer avoid staying too close to the text (p. 55). Hunter urges students to adjust their level of abstraction in accordance with their rhetorical purpose (Yamada, 2003, p. 254). In general, because formal language, such as academic English, uses many abstract terms, students need to unpack them to understand the meaning, and then they have to repack the meaning in another expression to write a text at an appropriate level of abstraction (McGowan, 2005). Furthermore, grammatical analysis of paraphrasing strategies suggests that grammatical form change (e.g., from verb to noun, from noun to adjective) and meaning addition inferred from the source text may help a writer avoid copying and subsequent accusations of plagiarism (Keck, 2010, pp. 207-208).
Motivation for the Current Research
Wheeler (2009)’s research successfully demonstrated that Japanese university students hold a negative attitude toward illegal copying. In his research, the texts shown to the students were supposed to be opinion papers assigned as homework. Opinion papers are a type of writing task which should show the writer’s ideas, and therefore, they should be written in the writer’s own words. If, however, the task were a summary task, it would be suspected that Japanese students would not hold a similar view on copying. There are several reasons for this suspicion.
First, summary writing is a unique task where reading and writing are linked, and the source text author’s and the summary text writer’s voices come together (Hirvela, 2004; Sarig, 1993). That is, half comes from the source text author, and half from the summary writer. In this situation, various task representations may be created and various choices are available: Should the summary writer focus on conveying the meaning of the source text faithfully or showing his or her own interpretations of it? Should the summary writer follow the information order of the source text or his or her own order? Should the summary writer use expressions from the source text or choose other expressions? Because of the unique feature of linking reading and writing together, diverse task representations may be generated.
Secondly, summary writing is a complicated task. As explained in the previous section, it involves various cognitive and linguistic processes, such as evaluation, selection, combination, and even substitution and creation (Hidi & Anderson, 1986). Given the difficulties of dealing with these requirements of summarizing, adding a further requirement of writing in one’s own words may become a minor consideration that is not sufficiently paid attention to by the writer. In this case, copying may occur as a side-effect of focusing on other cognitive demands of a summary task.
Next, cognitive demands of summarization may depend on the presence or absence of the source text during summarization, as is shown in Hidi and Anderson (1986: Refer to the previous section.) Muramoto’s 1992 research, which was conducted in Japan, also reveals differences of summary procedures depending on the availability of the source text while summarizing. In Muramoto’s study, Japanese university students were asked to write a summary in Japanese (L1) under the following two conditions: referring-text condition, where the summary writer writes a summary while referring to the source text; and referring-memory condition, where the writer summarizes without referring to the source. Summary samples in the referring-text condition were written by copying or modifying the source text, while summaries in the referring-memory condition tended to deviate from the text and contain unique or even mistaken information. In English-speaking countries, in order to encourage deeper processing of information, composition textbooks often ask students to write a summary without looking at the source text (e.g., Howard, 2010; Kennedy & Smith, 2006). On the other hand, Japanese students tend to write a summary while looking at the source (Machida, 2008). This Japanese students’ tendency to refer to the source text while summarizing may cause copying in summary writing.
Importantly, it is suspected that definitions of summarization in Japan may not include using the writer’s own words. Yoshimura and Adams’ (2018) review of the standard course of study in elementary, middle, and high schools, as outlined by the Ministry of Education in 1999 and MEXT in 2010, revealed the following: First, students learn summarization skills in Japanese language classes from elementary to high school as a reading aid but not as a writing skill. Second, while identifying main ideas and including them in their summary are emphasized, using the writer’s own words is not mentioned. This situation is rarely improved even after they enter university because summary skills are not explicitly defined or taught in Japanese universities (Yoshimura & Adams, 2018). This is in striking contrast to situations in English-speaking countries, where writing in one’s own words seems to be an important element of a summary definition, and students are explicitly warned against copying (e.g., Howard, 2010; Kennedy & Smith, 2006; Roig, 2015). Thus, Japanese students are not always asked to write their summaries in their own words in Japanese language classes; nor are they punished for borrowing expressions from the source in writing a summary in their L1.
Another reason is that writing in one’s own words is a difficult concept to communicate to some students. In English-speaking countries, therefore, various approaches are suggested in writing textbooks for university students. For example, writing a summary without looking at the source text (Howard, 2010), making a graph (Kennedy & Smith, 2006), or adjusting levels of abstraction according to the rhetorical purpose (Yamada, 2003). (Refer to the previous section.)
Writing in the writer’s own words is, even if the concept is understood by students, difficult for some to implement, for it requires paraphrasing ability. Paraphrasing is especially difficult for ESL students because it involves sophisticated language processing skills, such as rewording, sentence rearrangement, and grammatical manipulation. Many ESL students may fail to paraphrase successfully, which is shown in Keck’s 2006 research that examined paraphrases of English native speakers and ESL students. Some students may be reluctant to paraphrase because they want to avoid the risk of making grammatical mistakes (Yoshimura & Adams, 2018) or distorting the original meaning of the source text (Roig, 2001).
Furthermore, summarization is an inherently difficult task where even skilled writers are challenged to write in their own words. Roig (2001) reported that summary samples from college professors contained long copied word strings. He explained this as an effort to avoid conveying inaccurate information. Even writers who write their ideas in their own words in an opinion paper may use the source text author’s expressions in a summary task. Shi’s 2004 study also suggests that copying occurs more in summaries than in opinion essays. When reasons for textual borrowing are discussed, task types should be taken into account.
In sum, there are several reasons to suspect that Japanese university students’ task representations of summary writing may not include using their own words and that they may not have a negative attitude toward copying in summaries. It is suspected that this attitude may cause copying when they write a summary. The current research was conducted to examine this suspicion. For this purpose, Japanese university students’ English summary performance, their task representations of English summary writing, and the relations between them were analyzed.
The following three research questions were investigated in the present work:
- How do Japanese university students perform a summarization task in English? Do they borrow expressions from the source text? If language borrowing occurs, what are the copy rates?
- What are Japanese university students’ task representations of English summary writing?
- What effects do various task representations of English summary writing have on the copy rate?
One hundred and eight first-year undergraduate students in the English department of a private Japanese university participated in this research. They were all native speakers of Japanese, studying English as a foreign language (EFL) in Japan. Their English proficiency was approximately at A2-B1 level in CEFL (Common European Framework of Reference for Languages), which was inferred from the placement test using TOEIC®Bridge. They were chosen as participants for the current research as a representative student group who are in transition from secondary to post-secondary education. A summary writing task was given to the students as an in-class activity in an applied linguistics class in December 2016, which was eight months after they had entered university.
Task and Procedure
The students in the current research were given an independent (one source text) summarization task rather than a synthesized summary of two or more source texts. This was done because a synthesized summary might require knowledge of citation rules that the students had not learned. Participants were first asked to read a 137-word English expository text on “reasons for the popularity of dogs as a pet” (Refer to the appendix for the source text) and then asked to write a 40-word English summary in 20 minutes. Then, after the students’ summaries were collected, copies of a survey containing 12 descriptions of task representations about English summary writing were distributed (Refer to table 3 for the English translation of the descriptions). The descriptions were presented in Japanese (L1), and the students were asked to say how strongly they agree or disagree with each task representation by giving a score on 5-point Likert scale. On the scale, 1 indicates “strongly disagree” and 5 indicates “strongly agree”.
In the survey, task representation descriptions 1 & 2, 5 & 6, 7 & 8, 9 & 10, and 11 & 12, form pairs, respectively. These descriptions were created by including characteristics of good summaries from textbooks on academic writing in English-speaking countries (Yoshimura & Adams, 2018) and asked to analyze first-year Japanese university students’ task representations of English summary writing. Task Representation descriptions 1 and 2, that is, TR1 & TR2, were included to elicit students’ perceptions regarding whether the source text author’s or the summary writer’s voice should be prominent because both voices can be reflected in a summary. TR3 was included to investigate their perceptions regarding the addition of clarification expressions in their summary (Keck, 2010, pp. 207-208) and TR4 was included to elicit their perceptions regarding adjusting their level of abstraction (Yamada, 2003, p. 254). TR5 and TR6 were included to examine their perceptions about including important information, and excluding unimportant information when writing a summary. TR7 and TR8 were included to examine their perceptions regarding change of the information order. TR9 and TR10 were included to examine their perceptions regarding whose expressions should be used in writing a summary, the source text author’s or the summary writer’s. Finally, TR11 and TR12 were included to elicit the perceptions about their own behaviors when writing a summary, whether they look at the source text and use expressions from it or they avoid referring to the source text and use their own words.
Data Collection and Analysis Procedure
The data for this research consist of first-year Japanese university students’ English summary samples and their answers to a survey investigating their task representations of English summary writing. The collected summaries and the survey answers were assigned ID numbers so that the data could be handled anonymously. Then the summary samples were analyzed for copy rate and whether or not paraphrasing was attempted using lexical criteria in Keck’s (2006, 2014) taxonomy of paraphrase types. Copy rate was calculated by dividing shared words by total words in each summary sample, and then the mean and standard deviation for all the summary samples were computed. In addition, sentences in each sample were put into one of the following six categories: Exact Copy, Near Copy, Minimal Revision, Moderate Revision, Substantial Revision, and Invented or Gist. These categories come from Keck’s taxonomy of attempted paraphrase types (Refer to table 1). Students’ scores for each task representation description were calculated for the mean and the standard deviation. Finally, the students’ scores for the task representation descriptions and the copy rate were correlated (Refer to table 4 for the correlation matrix).
Table 1. Keck’s (2006, 2014) Taxonomy of Paraphrase Types Based on Lexical Criteria
|Exact Copy||100 % words contained within unique links|
|Near Copy||50 % or more words contained within unique links|
|Minimal Revision||20-49% words contained within unique links|
|Moderate Revision||1-19% words contained within unique links|
|Substantial Revision||Paraphrased but contained no unique links|
|Invented or Gist||Not paraphrased but invented by the author|
Note. Unique links mean “individual lexical words (i.e., nouns, verbs, adjectives, or adverbs), or exactly copied strings of words used in the paraphrase that (a) also occurred in the original excerpt but, (b) occurred in no other place in the original text” (Keck, 2006, p. 266)
Results of Text Analysis
When students’ summary samples were analyzed for copy rate, high copy rates were found. The average copy rate of all the summary samples was 72.2% (SD 20. 5%), a minimum of 23% and a maximum of 100%. Table 2 shows the categorization of the sentences using lexical criteria in Keck’s (2006, 2014) taxonomy. Most of the sentences fall into Near Copy (62.8%), followed by Exact Copy (15.8%) and Minimal Revision (11.9%). The total of these three types accounts for most (90.5%) of the sentence types found in the Japanese students’ summary samples.
Table 2. The Number of Cases and Percentages of Different Sentence Types According to Keck’s (2006, 2014) Taxonomy
|Sentence type||The number of cases||Occurrence (%)|
|Invented or Gist||22||4.50%|
Table 3 provides the means and standard deviations of the students’ scores for task representation descriptions. Participants valued including important information most (TR5, M=4.72) and using their own words least (TR12, M=2.32). They respected conveying the meaning of the source text faithfully (TR1, M=3.68) more than expressing their own interpretations (TR2, M=3.01). Though they tended to agree to adjusting their level of abstraction according to their rhetorical purpose (TR4, M=3.52), they tended not to agree to adding information to make the text meaning clearer (TR3, M=2.41). Though they unanimously agreed to including important information (TR5, SD=.75), there was wide variation whether or not to exclude unimportant information in their responses to this item (TR6, SD=2.36). The means for TR7 and TR8 (3.20 & 3.17, respectively) were very close to each other around the middle point of the range. This indicates that students tended to approve of both following the order of the source text and changing it when necessary. The means for TR9 and TR10 (2.96 & 3.08, respectively) were very close, having only 0.14 difference, which seems to suggest that the students neither strongly agreed nor disagreed about using expressions from the source text and using their own words. Thus, students did not show strong attitudes regarding what expressions should be used. However, a wide difference between the means for TR 11 and TR12 (4.05 & 2.32, respectively) indicates that Japanese students had a strong tendency to write summaries while looking at the source text and using expressions from it. When the mean difference between TR 11 and TR12 (1.73 point difference) was compared with the difference between TR9 and TR10 (only 0.14 difference), the difference between TR11 and TR12 was much wider, indicating an inconsistency between their beliefs and behaviors.
Table 3. Means and Standard Deviations of Japanese Students’ Scores for 12 Task Representation Descriptions
|1. A summary should reflect the meaning of the source text
|2. A summary should reflect how the summary writer
understands the source text
|3. You can add information to clarify the meaning inferred
from the source
|4. You should adjust your level of abstraction according to
your rhetorical purpose
|5. You should include important information from the source
text in your summary
|6. You should exclude unimportant information when writing
|7. A summary should reflect the order of the information in
|8. The order of information can be changed if necessary||3.17||1.05|
|9. You should try to use expressions from the source text||2.96||1.04|
|10. You should try to use your own words||3.08||0.92|
|11. I try to write a summary while looking at the source
and using words and sentences from it
|12. I try to write a summary without looking at the source
|my own expressions|
Note. The number 1 indicates “strongly disagree” and the number 5 indicates “strongly agree”
The Relationships between Task Representations and Copy Rate
In order to examine the relationships between various task representations of summary writing and copy rate, students’ scores for task representation descriptions 1 to 12 and the copy rate were correlated. Table 4 is the correlation matrix for different types of task representations and copy rate. The correlations between TRs 9 to 12 and copy rate showed statistical significance (r = .222*, -.200*, .199*, & -.218*, respectively). While the belief in and tendency toward using expressions from the source (TR9 & TR11) were positively correlated with copy rate (r = .222*, & .199*, respectively), the belief in and tendency toward using their own words (TR10 & TR12) were negatively correlated with copy rate (r = -.200*, & -.218*, respectively). In addition, in the group of task representation descriptions 9 to 12, which showed significant correlations with copy rate, while all the TR9’s correlations with TR10, TR11, and TR12 showed statistical significance (r = -.378**, .357**, & -.449**, respectively), TR10’s correlations with TR9 and TR12 showed significance (r = -.378**, & .272**, respectively), but the correlation was not significant with TR 11 (r = -.185).
Table 4. Correlation Matrix for Copy Rate and Different Types of Task Representations
Note. * indicates correlation is significant at the 0.05 level (two-tailed) and ** indicates correlation is significant at the 0.01 level (two-tailed).
In this section, results from the current research according to the research questions (1)-(3) are discussed first and then the implications are considered.
In order to investigate research question (1), whether and to what degree students borrow expressions from the source text, the text analysis of students’ summary samples was used. High copy rates found in Japanese students’ summaries (M = 72.2%) suggest that students borrow expressions from the source rather heavily in writing a summary. Analysis of students’ sentence types using Keck’s (2006, 2014) lexical criteria of paraphrasing taxonomy further supports this tendency, as most sentences written by the students in this study are Near Copy, Exact Copy, or Minimal Revision. According to Keck (2010), “the paraphrase type of Moderate Revision is the first type along the continuum that, for the most part, succeeds in avoiding the use of copied strings of three or more words” (p. 213). “Three words in a row” is often used as the heuristic for a case of plagiarism (Drum, 1986, p. 242). If this is a valid measurement, then it can be said that most of the sentences written by these Japanese university students may be judged, at the very least, as insufficient paraphrases, or as clear plagiarism.
In order to investigate research question (2) on the Japanese university students’ task representations of English summary writing, their scores for task representation descriptions 1-12 were used. As table 3 shows, some task representations found in their answers conform to those of instructors from English-speaking countries and some do not. This contrast can be seen regarding whose expressions students think they should use in a summary, where the students’ attitudes were neutral toward the statements that they should use expressions from the source (M = 2.96 for TR9), and that they should use their own words (M = 3.08 for TR10). Though this may seem contradictory to instructors from English-speaking countries, it may not be so in the eyes of Japanese university freshmen, who may think that writing in their own words can be achieved even if they borrow expressions from the source as building blocks, as long as their summaries show their own way of condensing the source text. At least it can be learned from these answers that Japanese first-year students do not feel a strong need to write their summaries in their own words. On the other hand, when students were asked about their perceptions of their own behaviors, students tended to state that they write their summaries while looking at the source text and that they use expressions from it (TR11, M = 4.05). Though both of these practices are discouraged in English-speaking countries, Japanese students showed a strong tendency to rely on them in their summary writing. This tendency is in accordance with findings from previous research conducted in Japan (e.g., Machida, 2008).
Research question (3), on the effects of Japanese university students’ task representations of summary writing on copy rate, was examined using the correlations between different kinds of task representations and copy rate, and further, between different types of representations which show statistical significance in their relationships with copy rate (TRs 9 to 12). As predicted, students’ belief that they should use expressions from the source (TR9) and their tendency toward looking at the source and using expressions from it (TR11) significantly contributed to the increase of copy rate. On the other hand, accepting that they should use their own words (TR10) and avoiding looking at the source while writing a summary (TR12) significantly contributed to the decrease of copy rate. From the additional examination of the correlations between different types of task representations (TRs 9-12), the following four points emerged. First, students’ belief that they should use expressions from the source (TR9) greatly promoted their tendency to refer to the source and use expressions from it (TR11). Second, their belief that they should use expressions from the source (TR 9) made it unnecessary for them to avoid looking at the source text and using their own words (TR12). Third, the belief that they should write in their own words (TR10) encouraged students to minimize referring to the source text and using their own expressions (TR12). Fourth, the belief that they should write in their own words (TR10) did not have a statistically significant effect on preventing students from looking at the source text and using expressions from it (TR11).
In sum, the above research results seem to suggest that Japanese university students tend to borrow expressions from the source text rather heavily. Furthermore, some of the copying behavior of Japanese university students may be caused by their task representations of summary writing which differ from those in academic communities in English-speaking countries, and by their tendency to refer to the source text while summarizing.
In English-speaking countries, writing in one’s own words is considered extremely important for students to be accepted as a member of academic communities (e.g., Hirvela & Du, 2013). However, this requirement is not necessarily conveyed to Japanese students. Therefore, Japanese English academic writing textbooks and writing instructors at Japanese universities must carefully explain the characteristics of summary texts that their university students are expected to produce.
Considering the uniqueness of summarization tasks where reading and writing meet and various task representations may be generated, explicit instructions should be given so that students can create task representations which are appropriate in their target academic community. It should be emphasized that a summary written in one’s own words is not an option but a requirement. Furthermore, merely warning students not to violate this principle is not sufficient. Instructors must provide clear definitions and guidelines for implementation. Effective strategies proposed in some composition textbooks for Anglosphere students, such as writing a summary without referring to the source text (Howard, 2010), generating a graphic representation (Kennedy & Smith, 2006), and changing their abstraction level according to their rhetorical purpose (Yamada,2006), may also help Japanese university students.
Given the complexity of summary writing which involves various cognitive and linguistic demands, instructors should break the task down into smaller steps and carefully guide students through the process so that the students can complete this difficult task successfully. For example, Kirkland and Saunders (1991), based on their analysis of the cognitive skills involved in summary writing, proposed a concrete model for writing an independent summary. It included a statement of purpose, a description of the format and steps in the process, a checklist and questions for monitoring cognitive load. They emphasized the importance of being consciously aware and in control of cognitive processes and of creating “superordination” (pp. 110-111), that is, a cognitive operation in constructing a holistic understanding of a text. Their model may be beneficial especially for Japanese students, who are used to “bottom-up processing” in English text reading (Kirkland & Saunders, 1991, p. 111).
The current research was conducted to examine Japanese university students’ English summary performance, their task representations of English summary writing, and the relationships between them. The results show 1) that Japanese first-year students tended to borrow expressions from the source text rather excessively; 2) that they did not seem to think that writing an English summary in their own words is a requirement; 3) that they tended to write their summaries while looking at the source text and using expressions from it; 4) that the belief that they should use expressions from the source may have encouraged some students to refer to the source text for the expressions, and this may have led to copying in the finished text; and 5) that while the belief that they should use their own words may have encouraged some students to minimize referring to the source text and using expressions from it, it did not necessarily prevent students from doing so and as a result they ended up copying. These results seem to support the hypothesis that some copying behavior of Japanese university students may come from their task representations of summary writing.
The results from this research should be discussed critically by considering the following limitations of the study. First, participants in this study were first-year university undergraduates who may be regarded as representatives for first-year Japanese students, but not as representatives for Japanese undergraduate students as a whole. In Keck’s 2014 experiment, both L1 and L2 students depended rather heavily on the source text in writing a summary when they were first-year students but that both groups of students became more independent and used their own words in their second or third years of study. Research conducted by Yoshimura and Adams (2018), which examined Japanese third-year university students’ summary performance, revealed that approximately a third of the students somehow managed to write an English summary by avoiding copying when they were explicitly told to do so. Therefore, copying behavior may be discarded as students get accustomed to the academic culture and acquire its values. Secondly, this research used a single group design and there was no comparison group. For example, there was no comparison group from English-speaking countries to offer insights regarding differences of task representations across cultures. However, this limitation can be partially overcome in other cases. For example, interpretations of the average copy rate of 72.2% can be informed when the figure is compared with available data from an equivalent student group in English-speaking countries. The third consideration is that cognitive operations involved in summary writing depend on the qualities of the text to be summarized, task procedures, and the type of summary to be produced (Hidi & Anderson,1986). (Refer to the literature review section.) In the current research, the source text was a short exposition with a clear thesis and main points. This may be a reason why some students did not use their own words. Students may have thought that some sentences or parts of sentences were good enough as they were to be used in their summaries. Further research using texts with different text characteristics, different task procedures, and different types of the summary to be generated should be conducted to explore effects of these factors on language borrowing behaviors.
Despite these limitations, the results of this research are valuable because very few empirical studies have been conducted regarding Japanese students’ English summary performance, their task representations of English summary writing, and the relations between them and because the insights gained from this study may provide another possible reason to explain the copying behavior of students in their academic writing.
About the Author
Fumiko Yoshimura is a professor at Tohoku Gakuin University, Japan. Her research interests and previous publications include EFL writing and instructions.
Axelrod, R. B., Cooper, C. R., & Warriner, A. M. (2008). Reading critically writing well: A reader and guide (8th ed.). Boston, MA: Bedfort/St. Martin’s.
Ballad, B., & Clanchy, J. (1991). Assessment by misconception: Cultural influences and intellectual traditions. In L. Hamp-Lyons (Ed.), Assessing second language writing in academic contexts (pp. 19-35). Norwood, NJ: Ablex.
Basham, C., Ray, R., & Whalley, E. (1993). Cross-cultural perspectives on task representation in reading to write. In Carson, J. G., & Leki, I. (Eds.), Reading in the composition classroom: Second language perspective. Boston, MA: Heinle & Heinle.
Brown, A. L., & Day, J. D. (1983). Macrorules for summarizing texts: The development of expertise. Journal of Verbal Learning and Verbal Behavior, 22, 1-14.
Drum, A. (1986). Responding to plagiarism. College Composition and Communication, 37(2), 241-243.
Flower, L., Stein, V., Ackerman, J., Kantz, M. J., McCormick, K., & Peck, W. C. (1990). Reading-to-write: Exploring a cognitive and social process. New York, NY: Oxford University Press.
Hidi, S., & Anderson, V. (1986). Producing written summaries: Task demands, cognitive operations, and implications for instruction. Review of Educational Research, 56(4), 473-493.
Hirvela, A. (2004). Connecting reading and writing in second language writing instruction. Ann Arbor: The University of Michigan Press.
Hirvela, A., & Du, Q. (2013). “Why am I paraphrasing?”: Undergraduate ESL writers’ engagement with source-based academic writing and reading. Journal of English for Academic Purposes, 12, 87-98.
Howard, R. M. (1995). Plagiarism, authorships, and the academic death penalty. College English, 57(7), 788-806.
Howard, R. M. (2010). Writing matters: A handbook for writing and research. New York, NY: McGraw –Hill.
Karbalaei, A., & Amoli, F. A. (2011). The effect of paraphrasing strategy training on the reading comprehension of college students at the undergraduate level. Asian EFL Journal, 13(3), 229-244.
Katims, D. S., & Harris, S. (1997). Improving the reading comprehension of middle school students in inclusive classrooms. Journal of Adolescent & Adult Literacy, 41(2), 116-123.
Keck, C. (2006). The use of paraphrase in summary writing: A comparison of L1 and L2 writers. Journal of Second Language Writing, 15, 261-278.
Keck, C. (2010). How do university students attempt to avoid plagiarism?: A grammatical analysis of undergraduate paraphrasing strategies. Writing & Pedagogy, 2(2), 193-222.
Keck, C. (2014). Copying, paraphrasing, and academic writing development: A re-examination of L1 and L2 summarization practice. Journal of Second Language Writing, 25, 4-22.
Kennedy, M. L., & Smith, H. M. (2006). Reading and writing in the academic community (3rd ed.). New Jersey, NJ: Pearson Prentice Hall.
Kirkland, M. R., & Saunders, M. A. (1991). Maximizing student performance in summary writing: Managing cognitive load. TESOL Quarterly, 25(1), 105-121.
Machida, Y. (2008). Examination of tendency in university student’s summary production. Annual Bulletin: Graduate School of Education: Tohoku University, 57(1), 241-252.
McGowan, U. (2005). Does educational integrity mean teaching students NOT to ‘use their own words’?, International Journal for Educational Integrity, 1. Retrieved from https://www.ojs.unisa.edu.au/index.php/IJEI/article/view/16
McNamara, D. S. (2004). SERT: Self-explanation reading training. Discourse Processes, 38,1-30.
Muramoto, T. (1992). The variety of summaries: Strategies of summary production and the goodness of summaries. Japanese Journal of Educational Psychology, 40, 213-223.
Pennycook, A. (1996). Borrowing others’ words: Text, ownership, memory, and plagiarism. TESOL Quarterly, 30, 201-230.
Rinnert, C., & Kobayashi, H. (2005). Borrowing words and ideas: Insights from Japanese L1 writers. Journal of Asian Pacific Communication, 15(1), 31-56.
Roig, M. (2001). Plagiarism and paraphrasing criteria of college and university professors. Ethics & Behavior, 11(3), 307-323
Roig, M. (2015). Avoiding plagiarism, self-plagiarism, and other questionable writing practices: A guide to ethical writing. Retrieved from https://ori.hhs.gov/sites/default/files/plagiarism.pdf
Sarig, G. (1993). Composing a study-summary: A reading/ writing encounter. In J. G. Carson & I. Leki (Eds.), Reading in the composition classroom: Second language perspective (pp. 161-182). Boston, MA: Heinle and Heinle.
Shi, L. (2004). Textual borrowing in second-language writing. Written Communication, 21, 171-200.
Shi, L. (2006). Cultural backgrounds and textual appropriation. Language Awareness, 15, 264-282.
Shi, L. (2008). Textual appropriation and citing behaviors of university undergraduates. Applied Linguistics, 31, 1-24.
Yamada, K. (2003). What prevents ESL/EFL writers from avoiding plagiarism?: Analyses of 10 North-American college students. System, 31, 247-258.
Yoshimura, F., & Adams, K. (2018). Analyzing Japanese University Students’ English Summaries and Task Representations as a Possible Source of Plagiarism. Journal of Institute for Research in English Language and Literature (Tohoku Gakuin), 43, 53-78.
van Dijk, T.A., & Kintch, W. (1977). Cognitive psychology and discourse: Recalling and summarizing stories, In W.U. Dressler (Ed.), Current trends in textlinguistics. New York: de Gruyter.
Wheeler, G. (2009). Plagiarism in the Japanese universities: Truly a cultural matter? Journal of Second Language Writing, 18, 17-29.
The source text
Dogs have been kept as pets throughout history. There are many good reasons for their popularity. Two important reasons are that they are friendly and they are good at keeping a home safe. First of all, dogs show their love to their owners. They like to play with their owners and follow them around. Dogs may be one of the most loving animals. In addition, dogs keep a home and family safe. They bark when strangers approach, and warn the family. They will also bark if something does not seem right, for example, if a fire is starting or if a window is broken. A security alarm may protect us, but it will not cheer us up or look happy when it sees us. Now, I think you can see why dogs are so popular as pets. (Source unknown)
|Copyright rests with authors. Please cite TESL-EJ appropriately.
Editor’s Note: The HTML version contains no page numbers. Please use the PDF version of this article for citations.