621 Studies in Second Language Learning and Teaching Department of English Studies, Faculty of Pedagogy and Fine Arts, Adam Mickiewicz University, Kalisz SSLLT 8 (3). 2018. 621-642 http://dx.doi.org/10.14746/ssllt.2018.8.3.5 http://pressto.amu.edu.pl/index.php/ssllt Expressing authorial self in research articles written by Polish and English native-speaker writers: A corpus-based study Katarzyna Hryniuk University of Warsaw, Poland k.hryniuk@uw.edu.pl Abstract This cross-linguistic and cross-cultural, corpus-based study explores the no- tion of writer identity expressed through self-reference. The study examines how writers from two cultural regions – Polish and Anglo-American – con- struct a credible representation of themselves in writing. That is, it investi- gates the differences and similarities in the frequency of use, and the role of first person pronouns and determiners, in the corpora of 40 research articles in the area of applied linguistics – 20 written by Polish authors in English, pub- lished in Polish institutions, and 20 by native English speakers, published in Anglophone journals. Additionally, the frequency of use and the role of nom- inal lexical items referring to the writers, such as the author(s) and the re- searcher(s), are explored. The location of pronouns, determiners and the lex- ical items in the IMRD structure (Introduction-Method-Results-Discussion) is also researched, as certain types of pronouns and determiners were expected to occur in the given sections, depending on their functions. The results clearly show that there is a striking difference between the use of pronouns and determiners in the texts written by the two groups of writers. The findings carry important implications for formulating clearer instructions and develop- ing appropriate writing strategies by novices writing for publication in EFL. Keywords: writer identity; authorial self; research article; academic writing; first person pronoun Katarzyna Hryniuk 622 1. Introduction Academic writing, regarded as a very complex skill, develops through a writer’s socialization into new discourse communities and their practices (e.g., Duff, 2007, 2010; Flowerdew, 2013). In this process, not only must novice writers acquire new genres with all their characteristics, but they also try to construct new academic writer identities (Ivanič, 1994, 1998). In Polish higher education institutions, such learners, when writing in English as a foreign language, learn to follow Anglo- American conventions of writing, but at the same time, they are under a strong in- fluence of the local, socio-cultural context in which they write. Therefore, the term discourse communities is used in plural in this introduction to indicate that even in one discipline there is no one monolithic community (Ivanič, 1998). Novice writers have many options to choose from in terms of content, wording and text organi- zation when creating academic discourse. However, they most often follow the patterns present in the texts they are exposed to, that is, academic publications, and frequently the choices they make are subconscious (Ivanič, 1994). Therefore, in this article I emphasize the importance, and explore the means of raising critical awareness of academic writers’ linguistic choices, since “every word a writer writes contributes to the impression he is creating of herself to a reader” (Ivanič, 1994, p. 5). This impression of a writer which emerges from the text, based on characteristics ascribed by a reader, creates writer identity. Following Ivanič (1998), in this article the term identity will be used to denote self. Ivanič and Camps (2001) also discuss the issue using the term voice, which is understood as a representation of identity in a given text. To clarify the meaning of the term with reference to writing, they compare it to speech, stating “that the lexical, syntactic, organizational, and even the material aspects of writing con- struct identity just as much as do the phonetic and prosodic aspects of speech, and thus writing always conveys a representation of the self of the writer” (Ivanič & Camps, 2001, p. 3). The authors adopt the Bakhtinian, social-interactionist view of language development, in which people write by drawing on voices they have encountered in the past when participating in discourses. These voices locate their users historically and culturally. But at the same time they select, recombine and use them in new, unique ways, establishing their identity in this way. In this view, writing is a social, interpersonal process (Ivanič, 1998). It takes place in en- counters with other people and texts (Ivanič & Camps, 2001). In other words, in Bakhtin’s framework, a learner establishes the self, and “the source of conscious- ness is embodied in the voices surrounding” (Vitanova, 2005, p. 167). Expressing authorial self in research articles written by Polish and English native-speaker writers. . . 623 2. Theoretical framework In the study described in this article, the framework developed by Ivanič (1998) will be used. She distinguished the following aspects of writer identity: · autobiographical self – associated with the writer’s history, roots, usually not deducible from a text; it is socially constructed and continuously developing; · discoursal self – the impressions of writers, which they project in the texts; it is reflected in discourse characteristics (relating to beliefs, val- ues, and power relations in a given context); · authorial self – the degree to which they establish presence in their texts, relating to their opinions, beliefs, position; it can be also described as “authoritativeness writers are prepared to invest, to personally get behind statements” (Hyland, 2011, p. 166). It is this last aspect which I focus on in my study – that is, writer identity expressed by the use of personal pronouns and determiners. In particular, the appearance of the pronoun I in academic texts is worth exploring, because, by using it, authors enhance the persuasive effect of their arguments (Szymańska, 2011). As Ivanič (1998) wrote, some writers attribute ideas more to other au- thorities in their writing, presenting the content as objective truth, and avoid using first person pronouns, others take more responsibility and intrude into a text, which changes readers’ perception. In the present study the notion of an authorial self will be explored rather than voice, because the latter has a broader meaning. It encompasses both authorial self and discoursal self (Matsuda, 2001). It must be also emphasized that the three aspects listed above are not completely separate. They interrelate as authorial self can be regarded as an as- pect of discoursal self, and it is also likely a product of autobiographical self. These aspects present just different perspectives on writer identity (Ivanič, 1998). The fourth aspect of identity distinguished by Ivanič (1998, p. 27) refers to the so-called “available possibilities for self-hood.” These are social identities, not belonging to particular individuals, but prototypical possibilities. This aspect shapes all the three previously discussed. The authorial self is influenced or, one can say, socially constructed by these “possibilities,” as the conventions for whether and how to establish authorial presence differ from one social context to another. Writers try to predict values, beliefs and expectations of readers, depending on the context, and in accordance with them either consciously or subconsciously shape the self (Ivanič, Aitchinson, & Weldon, 1996). A wide ranging discussion on writer identity started at the turn of the cen- tury, resulting in many publications on the topic, some of which will be outlined in the next part of this article, but it seems inconclusive. Still, since non-native- speakers’ rhetorical identities are shaped by different literacy traditions, the Katarzyna Hryniuk 624 conventions of personality remain a dilemma for both novice and experienced writers (Hyland, 2011). Many writers, academic writing instructors, and thesis advisors are hesitant when they are asked the question of whether and how to use personal pronouns and determiners in academic texts. They often either follow traditional conventions (what they were taught) or refer to style manuals and handbooks, which often give contradictory information. My suggestion is to use an inductive approach to text exploration, specifically corpus-based data driven learning (Boulton, 2012; Hryniuk, 2015; Johns, 1991), such as the one in the present study, to increase both learners’ and instructors’ awareness of the possible ways of expressing authorial self in various discourses and contexts. Therefore, the aim of the present cross-linguistic and cross-cultural, cor- pus-based study is to explore the notion of writer identity expressed through self-reference. The study examines how writers from two cultural regions con- struct a credible representation of themselves in writing. That is, it investigates the differences and similarities in the frequency of use, and the role of first per- son pronouns and determiners in the corpora of 40 research articles in the area of applied linguistics – 20 written by Polish authors in English, and 20 by native English speakers. Additionally, the frequency of use and the role of nominal lex- ical items referring to the writer, such as the author(s) and the researcher(s) is explored. The location of pronouns, determiners and the lexical items in the IMRD structure (Introduction-Method-Results-Discussion) is also researched. In this structure, typical of articles in experimental sciences, each section performs a different communicative function (see Hryniuk, 2017). The results of the study carry important implications for teaching academic writing for publication in EFL settings by enabling the formulation of clearer instructions for novice writers. However, before presenting the results, the main findings from previous re- search will be outlined. 3. Previous research The use of pronouns, determiners and other means to express authorial self in academic writing has not received enough attention in publications by foreign, and even less by Polish scholars so far. However, the most prominent studies which will be outlined below brought about significant results. The largest num- ber of studies focused on disciplinary differences in the use of pronouns by writ- ers writing in their mother tongues. There were also studies in which a compar- ison was made between expert and novice writers writing in English as a sec- ond/official language (e.g., in Hong Kong). Only a few researchers, however, con- centrated on the comparison between pronoun use in different languages by L1 writers, and between the use of pronouns and determiners by English native Expressing authorial self in research articles written by Polish and English native-speaker writers. . . 625 speakers and non-native speakers writing in English as a foreign language, in the same discipline. These studies will be reviewed in the abovementioned order. 3.1. Studies on expressing authorial self carried out outside Poland Many studies concentrate on the use of pronouns and determiners in English texts in specific academic disciplines written by native speakers. For example, Kuo (1999) explored the use of personal pronouns in scientific journal articles in the area of computer science, physics and electronic engineering. The author found that, although none of the journal guidelines stated the editorial policy concerning the use of pronouns, most frequently appearing were first person plural pronouns and determiners (we, us, our), even in single-authored articles. They were used to distance the writer from the text. There were no occurrences of first person singular pronouns at all. The author also analyzed different mean- ings and functions of the first person plural pronouns. She found that writers most often used the so-called exclusive we, which refers to speaker-writer only (e.g., “. . . we propose to use a statistical . . .” Kuo, 1999, p. 132). This pronoun emphasizes writers’ role in research. It highlights their contribution. The second most frequently used first person plural pronoun was the so-called inclusive we, which refers to both speaker-writer and hearer-reader (e.g., “We easily see that . . .” Kuo, 1999, p. 132) or to the whole discipline (“. . . we [researchers in the field] deal with unstructured targets . . .” Kuo, 1999, p. 134). It is most often used to express shared knowledge, beliefs, goals, and the like. It decreases the dis- tance between the writer and the reader. This division into exclusive and inclu- sive we will be also considered in the study part of the present article. A few studies by Hyland (2001, 2003, 2011) also focused on the use of first person pronouns and determiners in English texts from various disciplines. In the study published in 2001, Hyland explored their use by experienced native-speaker writers in the corpus of 240 research articles in eight disciplines: mechanical en- gineering, electrical engineering, marketing, sociology, philosophy, applied lin- guistics, physics and microbiology. In another study (Hyland, 2003), besides the articles, he also analyzed 800 article abstracts in the same disciplines to explore the use of self mentions in them. He found out that a large majority (75%) of the pronouns and determiners was in the so-called “soft” disciplines (i.e., the human- ities and social sciences), on average over twice more per article than in science and engineering (Hyland, 2001, 2003). More first person singular pronouns were used in the humanities and social sciences, in philosophy in particular, while in sciences the plural form was preferred, which was explained not only with the fact that the writers attempted to make their texts less personal, but it also pointed to more collaborative nature of research in sciences. The author also found important Katarzyna Hryniuk 626 disciplinary differences in the functions of first person pronouns used in the article abstracts. In those where quantitative research was presented, that is, mainly in sciences, the first person pronouns were more often used for describing proce- dures, in other ones – for elaboration of arguments. Also, in the humanities and social sciences, personal pronouns were more frequently used in abstracts to state the author’s viewpoint (Hyland, 2003). Two studies carried out by Harwood (2005b, 2005c) also focused on disci- plinary differences in the use of the personal pronoun I, and inclusive and exclu- sive we by English native-speaker writers. In the first one, he quantitatively and qualitatively explored research articles from four academic disciplines: business and management, computing science, economics and physics. The results showed that in the “soft” disciplines (here: business and management, and eco- nomics) the pronoun I was preferred, while in computing science and in physics, exclusive we was used instead. In the “soft” disciplines, only one case of exclusive we was found. The use of I in the “hard” disciplines was very rare. Inclusive we was used in both the “soft” and “hard” disciplines, but in computing science in- clusive we comprised only one third of the cases, and in the physics articles only 10%, while in the “soft” disciplines all but one of the pronouns were inclusive (Harwood, 2005c). The second study by Harwood (2005b) was a qualitative anal- ysis of the use of first person pronouns in research articles in the same disciplines. Harwood (2005a) also focused on differences between the use of pronouns in the methodology part of works written by native-speaker students of computing and by expert writers in the same discipline, using a taxonomy by Tang and John (1999) and other scholars, as well as a category distinguished by himself, that is, the methodological I (used for step by step descriptions of procedures). He found that the methodological I comprised 80% of all pronouns used in the works by novice writers, while only 3% in the works by expert writers. The pronoun that expert writers chose most frequently was we (in 90% of instances) and most of the in- stances were exclusive we – referring to the writer alone. Harwood (2005a) also carried out a qualitative functional analysis of the methodological I use in the cor- pora to show how the writers achieve different effects. In two other studies Har- wood (2006, 2007) investigated the views of writers in political science on appro- priate and inappropriate use and functional-pragmatic effects of personal pro- nouns in their own and other scientists’ writing. These were interview-based ac- counts which the author strongly recommends conducting for the purposes of instruction in English for specific purposes. Harwood and Hadley (2004) also sug- gest a pedagogical approach and activities based on such investigations, to which I will return in the final part of this article. In the 2011 study, Hyland explored the use of first person pronouns and de- terminers in 64 reports written in L2 by undergraduates at a Hong Kong university Expressing authorial self in research articles written by Polish and English native-speaker writers. . . 627 and compared it with their use in a parallel corpus of 240 articles written by expert native-speaker writers. The texts written by students were from the fol- lowing disciplines: biology, mechanical engineering, information systems, busi- ness studies, teaching English as a second language, economics, public admin- istration and social sciences. The main results showed that expert writers used four times more pronouns and determiners than novice writers. The novice writ- ers used I most frequently (it comprised 60% of all occurrences of pronouns), and we was the second most often used pronoun, even in single-authored texts. The pronouns were more often used to state goals and report on the method- ology in the students’ research rather than to make strong claims, unlike in arti- cles written by expert writers. In other words, the novice writers were reluctant to firmly stand behind their claims and downplayed their role in interpreting results. Overall, the most often appearing functions of pronouns in students’ reports were the ones which made their statements less risky and less face- threatening. Their aim was mainly to guide readers through the discourse. A very often cited study of first person pronoun use in many languages is a qualitative exploration by Vassileva (1998). She compared preferences of L1 ac- ademic writers from five countries (English, German, French, Russian and Bulgar- ian) with regard to the use of the pronouns I and we in linguistics research articles. She attributed the differences which she found to native language cultures of the writers and to their pragmatic choices rather than to the linguistic systems. She observed large differences, in particular between the use of the pronoun I in the five languages. In English, it appeared twice as often as in German and even more often when compared with any other of the languages explored. This pronoun was very frequently used in English for expressing the authors’ views, in the dis- cussion section in particular, and for showing commitment to the aims and pro- cedures of their studies. To the contrary, in Slavic languages (Bulgarian and Rus- sian) the pronoun we dominated as most often used for expressing personal views. In English and in German, unlike in the other three languages, inclusive we was commonly used when referring to other people’s work, evaluating previous research results and claims. Exclusive we, on the other hand, was more often used for describing methodology, procedures, and analyses in the Slavic languages, while the writers of German and English origin preferred to use I in such contexts (the French were on the borderline between the two). Vassileva (1998) concludes that the differences can be attributed to native language cultures. Thus, the degree of self-promotion through self-reference was the greatest in English, next in German and then in Slavic languages. The greatest difference, however, was between English and German writers, who used mostly I as a sign of individuality, and Russian and Bulgarian ones, who used mainly we as a sign of community as a whole. Vassileva (1998) claims that there Katarzyna Hryniuk 628 is a strong influence of English in terms of expressing individuality, through the use of the pronoun I, on German and French academic writing. Although English has become dominant also in Eastern Europe, Slavic languages seem to resist this trend. In Bulgarian and Russian academic culture, writers still prefer to em- ploy the collective we instead. She also explains that the fact that Russian and Bulgarian discourses favor a “collective approach” is the sign of the long-lasting influence of the powerful communist ideology in these countries. By avoiding first person singular pronoun use, Vassileva (1998) claims, writers from Slavic countries try to preserve their cultural identity. Hyland (2011) expresses the same view on collective identity of writers in Asian cultures. However, Matsuda (2001) offered a different explanation of the fact that Japanese writers’ texts seem to lack voice, strongly associated with individuality in the English native-speakers’ eyes. Matsuda (2001) showed how Japanese writers, who also belong to the so-called “collectivist cultures,” constructed their voice in electronic discourse (web diaries) through the discursive means which are language-specific. He claims that the English language lacks the coun- terparts of these discursive features, and that is what makes voice construction in English difficult for the Japanese, as it may be for any other second language writers. He argues that the notion of voice does not have to be tied to the ideology of individualism and claims that what the Japanese need is to get better acquainted with the proper discourse features and strategies for con- structing voice in English writing. A very relevant study on expressing identity in academic writing in English as a foreign language was carried out by Martinez (2005). She focused on the differences between the use of first person pronouns in writing by native and non-native English speakers whose mother tongue was Spanish. She analyzed the distribution of the pronouns in different sections of the IMRD structure of research articles in the area of biology. All instances of pronouns were plural forms because all articles in the corpora were of multiple authorship. Only the pronouns which referred to the writers, that is, the occurrences of exclusive we were considered. Martinez (2005) found that native-speakers overall used twice as many pronouns than non-native speakers, six times more in the Results sec- tions, and almost twice more in the Discussion sections. These results contradict the view that first person pronouns are not used in sciences and that the Results section is used exclusively for reporting facts impersonally. The researcher also found that, as opposed to native English speakers, non-native-speakers avoided using the pronouns in the functions which would make their arguments more persuasive (e.g., for expressing intentions, decisions, or stating claims). This strategy may decrease their chances of getting published. Native English speak- ers claimed more responsibility for their findings by using pronouns, making their Expressing authorial self in research articles written by Polish and English native-speaker writers. . . 629 arguments more effective in this way. Martinez (2005) explains that avoidance of first person pronoun by non-native speakers may be the result of transfer from Spanish which is (like Polish) a pro-drop language. Thus, the first person pronoun is redundant in a sentence because it is also marked by a verb suffix. However, the author claims that raising awareness of the differences between the use of the pronoun by native and non-native speakers of English may lead to the employment of more effective writing strategies. Hence, all of the authors of the studies came to a similar conclusion (for a review of research on the use of personal pronouns and determiners see also Işik Taş, 2010). 3.2. Studies on expressing authorial self carried out by Polish researchers Although none of the early studies comparing Polish and Anglo-American academic discourse (for an overview see Hryniuk, 2017) concentrated on the use of pronouns and determiners, an important observation was made by Duszak (1994) that Polish academic writing, in the area of language studies in particular, is strongly influenced by German writing conventions. After Clyne (1981), she claimed that it follows the so-called Teutonic intellectual style. Thus, in line with the traditional, conservative trends set by the post-war German academia, writing is supposed to be objective, impersonal, and dismissive of personal experience. In this view, the use of pro- nouns to express authorial self is not allowed. However, Mikołajczyk (2011), who compared Polish and German academic writing, noticed that this trend is chang- ing. English as a scientific lingua franca has influenced even German rhetorical conventions in this area (see also Vassileva, 1998). Slavic conventions, such as Polish, however, seem to be more resistant to the changes. An example of a more recent study which implies the abovementioned view is the one by Trepczyńska (2016). She analyzed 49 undergraduate students’ papers in the area of literature, culture, linguistics and applied linguistics, writ- ten by Polish students of English studies, to explore the number and the distri- bution of pronouns and determiners in them. She also analyzed the functions of the pronouns. A vast majority of the pronouns (75%) were in the plural form, although all papers were single-authored, and they were located in the main body of the text. Students most frequently used first person singular pronouns in the low-risk functions of signposting and stating aims, while the plural pro- nouns were preferred for the more face-threatening functions of making claims or elaborating on arguments. This tactic, as the author claims, “dilutes the au- thority of the writer” (Trepczyńska, 2016, p. 117). Similarly to the results of the studies outlined before, Trepczyńska (2016) concluded that novice writers are reluctant to express their identity through the use of singular first person pro- nouns, and they avoid direct criticism in this way. She claims that the results imply Katarzyna Hryniuk 630 a lack of assertiveness on the part of the writers, and unwillingness to assume responsibility for their claims. Summing up the previous research, it must be stated that more research is needed to get a thorough insight into the ways that authorial self is expressed in different contexts and cultures. Therefore, the study described in the follow- ing section is intended to contribute to the research on this issue with regard to Polish EFL writers. 4. The study This study quantitatively and qualitatively investigates the differences and simi- larities in the frequency and functions of first person pronouns and determiners, as well as nominal phrases referring to the writer(s), in the corpora of 40 re- search articles in the area of applied linguistics, written in English by Polish and Anglo-American writers. Its aim is also to compare the location of the lexical items in the IMRD structure of the articles. For this purpose, the following re- search questions were addressed: 1. What are the differences (and similarities) in the use of first person pro- nouns, determiners, and nominal phrases referring to the author, be- tween the texts written by Polish and Anglo-American writers? 2. Are there any significant differences between their uses in specific sec- tions of the IMRD structure in the two sub-corpora? 3. What rhetorical functions do the analyzed items perform in the articles? 4.1. Corpus and methodology In this study the same corpus was used as in my previous ones (Hryniuk, 2015, 2016, 2017). It consisted of 40 empirical research articles in the area of applied linguistics – 20 written by Polish authors in English, published by Polish institu- tions, and 20 by native English speakers, published in Anglophone journals. All of the articles were published in representative, peer-reviewed, highly reputa- ble linguistics journals in the years 2009-2013. It is important to note that arti- cles in the Anglo-American corpus were longer – on average by 2,800 words, but in such small specialized corpora full texts should be used (Bowker & Pearson, 2002; Flowerdew, 2004). The Polish sub-corpus consisted of 135,358 words and the Anglo-American one of 191,423. Therefore, the number of the pronouns’ occurrences per 1,000 words was counted as well. Moreover, only 2 articles in the Polish corpus had more than one author, in the Anglo-American one – 7 had two or three authors. Expressing authorial self in research articles written by Polish and English native-speaker writers. . . 631 All of the articles had the IMRD structure (i.e., Introduction-Method-Results- Discussion). The Conclusion section was added because it was present in most of the articles – in 95% of the articles in the Polish sub-corpus, and in 60% of the arti- cles in the Anglo-American one. In the remaining 40% of the articles from the Anglo- American sub-corpus, the Conclusion sections were absent or merged with the pre- ceding ones. Similarly, in 40% of the articles from the Polish sub-corpus, the Discus- sion sections were absent or merged with the Results sections (Hryniuk, 2017). The following pronouns: I, we, us, and determiners: my, our were searched for. The examples of pronouns and determiners which were used in the utterances expressed by the writers’ informants, rather than by the writers themselves, were excluded (e.g., me was only found in the utterances of the writers’ informants so it was not counted). Also, only the cases of first person plural pronouns referring to the writers, known as exclusive we, were considered. Certain types of pronouns were expected to occur in the given sections of the articles depending on the function they performed in the texts. Thus, in the qualitative analysis of the pronouns’ functions, categories distinguished by Hy- land (2011) were used. They were matched with the functions of the given arti- cle sections in the following way: 1. Stating a purpose/goal, which acts like signposting – expected in the In- troductions. 2. Explaining a procedure, also named methodological I by Harwood (2005a) – expected in the Methodology sections. 3. Stating results/claims – expected in the Results sections. 4. Elaborating an argument – expected in the Discussion sections. 5. Expressing self-benefits – expected in the Conclusion sections. Three of the functions listed above, that is Stating a purpose, Explaining a pro- cedure, and Expressing self-benefits, involve little risk, but the functions Stating results/claims and Elaborating an argument are regarded as the most face- threatening for the writer. Additionally, the use of nominal lexical items, such as the author(s) and the researcher(s) was explored in the present study. Although they introduce greater distance to the statements expressed in texts, they also commonly mark a writer’s presence. In order to arrive at an in-depth insight into the use of the lexical items explored, and to achieve precision, a concordance program WordSmith Tools 6.0 (Scott, 2012) was used in the study. I have generated concordance lines in order to explore the context of use of all the lexical items. Katarzyna Hryniuk 632 4.2. Results The results clearly show that there is a striking difference between the use of pronouns and determiners in the two sub-corpora, as shown in Table 1. The total number of the pronouns and determiners in the Polish corpus (PC) is 52, while in the Anglo-American corpus (AC) 297 – almost six times more. If we look at the frequency of their use, in the AC, the number is 1.55 per 1,000 words, while in the PC, it is only 0.38, that is four times less. Table 1 The number and the frequency of pronouns and determiners per 1,000 words Polish corpus Anglo-American corpus 52 (0.38) 297 (1.55) As can be seen in Figures 1 and 2, the greatest difference between the cor- pora is in the use of the pronoun I. If we consider the percentages, they are four times larger in the case of the AC than in the PC. Looking at the numbers, the difference is even larger. Also, although the percentage of the use of the pro- noun we in both corpora is the same (58%), when we look at the number, it is almost six times larger in the case of the AC. Interestingly, the largest proportion of the pronoun we can be found in both corpora, although in the Polish one only 2 articles had more than one author, and in the Anglo-American – 7. Figure 1 Percentages and the numbers of pronouns and determiners in the Polish corpus 4% (2) I 58% (30) we8% (4) us 13% (7) my 17% (9) our Expressing authorial self in research articles written by Polish and English native-speaker writers. . . 633 Figure 2 Percentages and the numbers of pronouns and determiners in the An- glo-American corpus Looking at Figures 3 and 4, we can see large differences between the two sub-corpora in the distribution of the pronouns and determiners in particular article sections. The greatest number of the lexical items in the PC is in the first two sections of the articles (Introduction and Method), while in the AC, the numbers referring to the first two sections of the articles, as well as to the Dis- cussion sections, are similar. In the Results sections, the percentage of the pro- nouns and determiners is over twice smaller (13%) than in the other three sec- tions of the AC, but the number (38 instances) is much larger when compared with the number of occurrences of these items in the same sections of the PC. Only one determiner was found in the Conclusion sections of the AC, so it is not marked in the chart. Figure 3 Percentages and the numbers of pronouns and determiners in article sections of the Polish corpus 16% (47) I 58% (172) we 5% (16) us 2% (7) my 19% (55) our 42% (22) Introduction 33% (17) Method 11% (6) Results 6% (3) Discussion 8% (4) Conclusion Katarzyna Hryniuk 634 Figure 4 Percentages and the numbers of pronouns and determiners in article sections of the Anglo-American corpus When we compare the frequency of pronouns and determiners use per 1,000 words in particular sections of the articles, as Table 2 shows, the largest concentration of them can be found in the Introductions, the Method sections and the Discussions in the AC. In the PC, although the frequency number is the largest in the Method sections, it is still almost three times lower than the fre- quency number in the same sections of the AC. The numbers presenting the largest differences between the corpora are in italics in the tables below. Table 2 The numbers (n) and the frequency (f) of pronouns and determiners occurrence per 1,000 words in particular sections of the Polish (PC) and the An- glo-American corpora (AC) Article section PC AC n f n f Introduction 22 0.41 94 1.57 Method 17 0.72 82 1.94 Results 6 0.16 38 0.77 Discussion 3 0.27 82 2.29 Conclusion 4 0.39 1 0.21 Totals 52 0.38 297 1.55 A juxtaposition of the numbers of particular pronouns and determiners in the specific sections of the articles also reveals many differences between the cor- pora, as shown in Table 3. The frequency numbers of the items occurrence per 1,000 words are not given in the table because they are very low. Only in the case of the AC, the frequency of occurrence of the pronoun we is larger than 1 in the Introduction sections (i.e., 1.16) and in the Discussion sections (i.e., 1.29). In the PC the frequencies of the item occurrence are very low (the largest number is 0.46 in the Method sections). 32% (94) Introduction 27.50% (82) Method 13% (38) Results 27.50% (82) Discussion Expressing authorial self in research articles written by Polish and English native-speaker writers. . . 635 Thus, the largest differences, as shown in numbers, between the corpora can be noticed in the use of the pronoun I in the Method sections (1 in PC and 35 in AC), and the pronoun we in the Introductions (12 in PC and 69 in AC), as well as in the Discussion sections with regard to the pronoun we (0 in PC and 46 in AC) and the determiner our (0 in PC and 27 in AC). The numbers of the pro- nouns and determiners in those article parts are in italics in the table. Table 3 The numbers of particular pronouns and determiners occurrence in the specific sections of the Polish (PC) and the Anglo-American corpora (AC) Article section I we us my our PC AC PC AC PC AC PC AC PC AC Introduction 1 3 12 69 2 7 1 3 6 12 Method 1 35 11 29 0 2 4 2 1 14 Results 0 5 6 28 0 3 0 0 0 2 Discussion 0 4 0 46 1 3 2 2 0 27 Conclusion 0 0 1 0 1 1 0 0 2 0 Totals 2 47 30 172 4 16 7 7 9 55 Additionally, the numbers of lexical items, such as the author(s) and the re- searcher(s), in particular article sections of the corpora were examined, as they are commonly used in research articles with reference to the writers, as an alternative to the first person pronoun use. As Table 4 shows, the same number of the noun pharse the researcher(s) appears in the texts by Polish and Anglo-American writers. The noun pharse the author(s), however, is over twice more often used by Polish writers. It seems that Polish writers tend to identify themselves with the authors of the texts, rather than researchers, when referring to themselves in the third per- son singular or plural. The largest concentration of both of those lexical items is in the Method sections. Interestingly, in 12 cases, the phrase the present author was used in the PC, unlike in the AC – here no instances of the use of this adjective together with the noun author were found. Table 4 The numbers of the phrases the author(s) and the researcher(s) in par- ticular sections of the Polish (PC) and the Anglo-American (AC) corpora Article section The author(s) The researcher(s) PC AC PC AC Introduction 2 1 0 1 Method 12 4 7 12 Results 4 2 3 1 Discussion 2 1 2 0 Conclusion 0 0 2 0 Totals 20 8 14 14 Katarzyna Hryniuk 636 Although in most of the cases the pronouns and determiners performed the functions which were expected in the particular sections of the articles (listed in the previous part of this article), in the qualitative analysis I will focus on the largest differences between the corpora in this respect. In all of the article sections the largest number of the pronoun we was found in both corpora, so examples given below will mostly include this pronoun. If we look at the functions in which the pronoun is used in the Introductions, we can no- tice that apart from the main one, which is Stating a goal, in the AC we was also often used in the functions involving high risk, for example: (1) . . . we maintain that . . . (AC) (2) As we demonstrate in this study . . . (AC) (3) In this study, we take a different approach . . . (AC) Even if such phrases could also be found in the PC, they were often hedged through the use of modal verbs, or the writers expressed caution by stating ex- plicitly that making a claim is risky, for example: (4) . . . we may also hypothesize . . . (PC) (5) . . . we may forward the claim . . . (PC) (6) . . . we may risk stating . . . (PC) In the Method sections of the PC, there was only one occurrence of the singular pronoun, but in the AC, both plural and singular first person pronouns were used in the function of Explaining a procedure, as it was expected. The overall number of the pronoun use in this section was much smaller in the PC. Also, an example of a surprisingly highly risky statement was found there: (7) . . . We postulate that . . . (PC) In the Results sections, also differences between the corpora were found in the use of the pronoun we. While in the AC it was used in the typical function for this article section, namely Stating results and claims (e.g., “we present”), and also in the function typical of the Method sections, namely Explaining a procedure (e.g., “we compared,” “we examine,” “we used”), in the PC, again the sentences with the pronouns were hedged by the use of modal verbs or negations, weakening the strength of the claims in this way, for example: (8) . . . We may risk stating . . . (PC) (9) . . . we may conclude . . . (PC) (10) . . . on the basis of the results, we cannot say . . . (PC) Expressing authorial self in research articles written by Polish and English native-speaker writers. . . 637 In the Discussion sections, the pronoun we was found only in the AC. It most often performed the risky function of Elaborating an argument (e.g., “we argue that . . .,” “we agree with . . .,” “we stated”), but also the function of Stating results and claims, typical for the preceding section (e.g., “we noted,” “we captured,” “we ex- amined,” “we reported,” “we compared”). The latter function occurrence can be ex- plained with the writers’ need to restate the main results in the Discussion sections. Single occurrences of the first person pronouns were found in the Conclu- sion sections of the Polish corpus. The only one found in the AC, was the pronoun us, and it performed a typical for this section function of Expressing self-benefits: (11) . . . It gives us reasons to keep searching . . . On the basis of this analysis, it can be stated that Polish writers often pre- fer other means of expression, such as passive voice and other structures mak- ing their writing impersonal, instead of first person pronouns, but by avoiding them, they do not take advantage of all of the available language resources which can make their writing more effective in certain contexts. 5. Discussion The present study results have shown that Polish writers’ texts written in English as a foreign language are characterised by an infrequent use of first person pro- nouns, when compared with native English speakers’ writing in the same disci- pline. The most evident differences were found in much less frequent use of first person singular pronouns in the Methodology sections of research articles, and first person plural pronouns use in all sections of the IMRD structure. Moreover, while the distribution of the pronouns’ occurrence was very similar in the Intro- duction, the Method and the Discussion sections in the texts written by Anglo- American writers, in those written by Polish writers, larger concentration of the pronouns could only be found in the Introductions and the Method sections. Finally, the analysis of the functions performed by the pronouns, as indicated by the functions of research article sections, has shown that even if Polish writers in the majority of cases used the pronouns in the proper functions, they often weakened their claims by hedging and using negations in the same sentences. Polish writers prefer to use impersonal structures, or plural forms of pronouns, the effect of which is also distancing themselves from the ideas expressed in the texts, more typical of writing in sciences, as the previous research has shown (Kuo, 1999; Hyland, 2003; Harwood, 2005b, 2005c). It seems that using personal pronouns more often to get behind their statements, and assume responsibility for their claims, would make their writing more persuasive (Szymańska, 2011). Katarzyna Hryniuk 638 Another issue is that the differences can only be discussed here in terms of the underuse of first person pronouns by Polish writers when taking Anglo- American writers’ practices as a point of reference. However, especially with re- gard to English as the language of international publication, it is advisable to avoid privileging Anglo-American conventions of writing over any other native language writing conventions (Connor, 2011). We must also bear in mind that the writing conventions themselves are changing. In line with the prevailing practices of positivist epistemology in the past, writers used to hide their iden- tities and authority behind the shield of objectivity, and it seems that this style of writing is still most often practiced by Polish writers, both in sciences and the humanities. Now, after the shift towards constructivism, a change in writing style can be observed. As Hyland (2005, p. 66) writes, now “academics do not simply produce texts that plausibly represent an external reality, but use lan- guage to acknowledge, construct and negotiate social relations . . . Writers seek to offer a credible representation of themselves.” For example, in the United States, where much emphasis is put on developing academic writing, especially at tertiary level education, students are taught to signal the presence of the au- thor and interact with the reader, which reflects cultural ideals of the society at large (Dahl, 2004). However, it may be disputable if EFL writers who are obliged to publish in international journals also have to strictly follow these changes. It must be acknowledged that the present study results cannot be gener- alised to all academic writing by Polish and Anglo-American writers, as the cor- pus explored was not very large. Nevertheless, it may shed light on the issue of expressing authorial self, as an aspect of writer identity, by those two groups of writers. Also, following Martinez’s inquiry (2005), one may conclude that more frequent use of first person pronouns by non-native speakers of English would make their writing more persuasive. However, further research would benefit from examining reactions of readers to the use of various conventions of writ- ing, due to the fact that identity is also ascribed by them, as it has been stated in the introduction. Overall, the current study emphasizes the complexity of the relationships between personal identity, knowledge construction, and writing conventions in a given discipline, which are worth further exploration. 6. Conclusions and implications for instruction Obviously, there may be many reasons for the avoidance of pronoun use. As it was stated in the introduction, the cultural context in which academic writing in Slavic languages has been shaped is one of them (Duszak, 1994; Vassileva, 1998). Another reason can be the influence of transfer from the mother tongue. Polish is a pro-drop language, which means that apart from using a pronoun in a sentence, Expressing authorial self in research articles written by Polish and English native-speaker writers. . . 639 the person is also marked by a verb suffix. The use of subject pronoun is then redundant. When it is overused, it sounds unnatural and pompous. The reason for the omissions of personal pronouns may also be little experience of the writer, the native culture conventions of writing or the target audience. With regard to the problem, the critical-pragmatic approach can be intro- duced in EFL academic writing instruction, as suggested by Harwood and Hadley (2004). According to the authors, in order to enhance strategic writing for pub- lication, and not to move from one prescriptivism to another at the same time, novice writers can study different patterns of pronoun use present in corpus data, and be provided with alternatives which they can adopt or flout. Increas- ing awareness in this way should make tutors as well as international publica- tions gatekeepers more tolerant of different discourse practices. Raising editors’ and reviewers’ awareness of different patterns in academic writing may posi- tively influence acceptance rates and reduce rejecting and overlooking the find- ings from less known areas in the world. Katarzyna Hryniuk 640 References Boulton, A. (2012). What data for data-driven learning? The EUROCALL Review. Proceedings of the EUROCALL 2011 Conference, 20, 23-27. Bowker, L., & Pearson, J. (2002). Working with specialized language: A practical guide to using corpora. London: Routledge. Clyne, M. (1981). Culture and discourse structure. Journal of Pragmatics, 5, 61-66. Connor, U. (2011). Intercultural rhetoric in the writing classroom. Ann Arbor: The University of Michigan Press. Dahl, T. (2004). Textual metadiscourse in research articles: A marker of national culture or of academic discipline? Journal of Pragmatics, 36, 1807-1825. Duff, P. A. (2007). Problematising academic discourse socialisation. In H. Mar- riott, T. Moore & R. Spence-Brown (Eds.), Learning discourses and the dis- courses of learning (pp. 1-18). Melbourne: Monash University Press. Duff, P. A. (2010). Language socialization into academic discourse communities. Annual Review of Applied Linguistics, 30, 169-192. Duszak, A. (1994). Academic discourse and intellectual styles. Journal of Prag- matics, 21, 291-313. Flowerdew, L. (2004). The argument for using English specialized corpora to un- derstand academic and professional language. In U. Connor & T. Upton (Eds.), Discourse in the professions: Perspectives from corpus linguistics (pp. 11-33). Amsterdam: John Benjamins. Flowerdew, J. (2013). English for research publication purposes. In B. Paltridge & S. Starfield (Eds.), The handbook of English for specific purposes (pp. 301-321). Chichester: Wiley-Blackwell. Harwood, N. (2005a). “I hoped to counteract the memory problem, but I made no impact whatsoever”: Discussing methods in computing science using I. English for Specific Purposes, 24, 243-267. Harwood, N. (2005b). “Nowhere has anyone attempted . . . In this article I am to do just that”: A corpus-based study of self-promotional I and we in aca- demic writing across four disciplines. Journal of Pragmatics, 37, 1207-1231. Harwood, N. (2005c). “We do not seem to have a theory . . . The theory I present here attempts to fill this gap”: Inclusive and exclusive pronouns in aca- demic writing. Applied Linguistics, 26(3), 343-375. Harwood, N. (2006). (In)appropriate personal pronoun use in political science: A qualitative study and a proposed heuristics for future research. Written Communication, 23(4), 424-450. Harwood, N. (2007). Political scientists on the functions of personal pronouns in their writing: An interview-based study of “I” and “we”. Text & Talk, 27(1), 27-54. Expressing authorial self in research articles written by Polish and English native-speaker writers. . . 641 Harwood, N., & Hadley, G. (2004). Demystifying institutional practices: Critical pragmatism and the teaching of academic writing. English for Specific Pur- poses, 23, 355-377. Hryniuk, K. (2015). Metalinguistic expressions in research articles written by Polish and English native-speakers: A corpus-based study. In A. Turula & M. Chojnacka (Eds.), CALL for bridges in school and academia (pp. 33-47). Berno: Peter Lang. Hryniuk, K. (2016). The use of citations in research articles written by Polish and Eng- lish native-speaker writers. In H. Chodkiewicz, P. Steinbrich, & M. Krzemińska- Adamek (Eds.), Working with text and around text in foreign language en- vironments (pp. 143-157). Heidelberg: Springer. Hryniuk, K. (2017). Linguistics research articles written in English: Comparing na- tive English speakers and Polish writers. International Journal of Applied Linguistics, 27(1), 3-23. Hyland, K. (2001). Humble servants of the discipline? Self-mention in research articles. English for Specific Purposes, 20, 207-226. Hyland, K. (2003). Self-citation and self-reference: Credibility and promotion in academic publication. Journal of the American Society for Information Sci- ence and Technology, 54(3), 251-259. Hyland, K. (2005). Metadiscourse: Exploring interaction in writing. London: Continuum. Hyland, K. (2011). Authority and invisibility: Authorial identity in academic writ- ing. In L. Wei (Ed.), The Routledge applied linguistics reader (pp. 165-184). New York: Routledge. Işik Taş, E. (2010). “In this paper I will discuss . . .”: Current trends in academic writing. Procedia Social and Behavioral Sciences, 3, 121-126. Ivanič, R. (1994). I is for interprsonal: Discoursal construction of writer identities and the teaching of writing. Linguistics and Education, 6, 3-15. Ivanič, R. (1998). Writing and identity: The discoursal construction of identity in academic writing. Amsterdam: John Benjamins. Ivanič, R., Aitchison, M., & Weldon, S. (1996). Bringing ourselves into our writing. RaPAL Bulletin, 28(29), 2-8. Ivanič, R., & Camps, D. (2001). I am how I sound: Voice as self-representation in L2 writing. Journal of Second Language Writing, 10, 3-33. Johns, T. (1991). Should you be persuaded: Two examples of data-driven learning. Classroom concordancing. English Language Research Journal, 4, 1-16. Kuo, C.-H. (1999). The use of personal pronouns: Role relationships in scientific journal articles. English for Specific Purposes, 18(2), 121-138. Martinez, I. A. (2005). Native and non-native writers’ use of first person pro- nouns in the different sections of biology research articles in English. Jour- nal of Second Language Writing, 14, 174-190. Katarzyna Hryniuk 642 Matsuda, P. K. (2001). Voice in Japanese written discourse: Implications for sec- ond language writing. Journal of Second Language Writing, 10, 35-53. Mikołajczyk, B. (2011). Interpersonalność jako cecha tekstów naukowych na przykładzie przedmowy do monografii naukowej. Lingwistyka Stosowana. Przegląd, 4, 137-148. Scott, M. (2012). WordSmith Tools (version 6.0). Liverpool: Lexical Analysis Software. Szymańska, J. (2011). The interpersonal stage in academic written discourse. Opole: Wydawnictwo Uniwersytetu Opolskiego. Tang, R., & John, S. (1999). The “I” in identity: Exploring writer identity in student academic writing through the first person pronoun. English for Specific Purposes, 18, S23-S39. Trepczyńska, M. (2016). “In this paper I will prove . . .”: The challenge behind authorial self-representation in L2 undergraduate research paper writing. In H. Chodkiewicz, P. Steinbrich, & M. Krzemińska-Adamek (Eds.), Working with text and around text in foreign language environments (pp. 109-123). Heidelberg: Springer. Vassileva, I. (1998). Who am I/who are we in academic writing? A contrastive analysis of authorial presence in English, German, French, Russian and Bulgarian. International Journal of Applied Linguistics, 8(2), 163-190. Vitanova, G. (2005). Authoring the self in a non-native language: A dialogic ap- proach to agency and subjectivity. In J. K. Hall, G. Vitanova & L. Marchen- kova (Eds.), Dialogue with Bakhtin on second and foreign language learn- ing (pp. 149-169). Mahwah, NJ: Lawrence Erlbaum Associates.