Is EFL students’ academic writing becoming more informal?

1. Introduction

An abstract is an essential part of a research article, thesis, or dissertation. It provides a concise summary of the research, consisting of 100–300 words that allow the readers to review the entire content of the research immediately. Lores (2004, 281) explains that an abstract is the gateway of a research article which could lead the readers to either continue or stop reading the article. An abstract is concise and not time-consuming to read, therefore, an abstract can gain the readers’ attention quickly. Cooley and Lewkowicz (2003) assert that an abstract is a key to attract the readers’ attention toward an article. For journal editors, an abstract assists them to make a quick evaluation to either accept or reject the article (Wallwork 2011, 177). In addition, Pho (2008, 231) mentions that an abstract functions as the selling point of an article. Hence, the students and academics attempting to write articles for academic journals or final assignments must pay good attention to their abstracts.

Loutayf (2017, 18) argues that an abstract is intended to a specific target audience with its own requirements. One of the requirements is the use of formal tone of writing. Regarding the formality in writing, Hyland and Jiang (2017, 41) state that the degree of an author’s academic writing highly depends on its formal manner in presenting his/her writing. The formal tone and manner assist the specific readers to perceive the correct information about the content of the writing.

The demand to write an academic writing in a highly formal manner is unquestionable. Nonetheless, there were several studies suggested that academic writing contained informality features. Unlike other writing genres, e.g., creative writing, academic writing must be written in a formal tone which is an inflexible and conservative style (Hyland and Jiang 2017, 41). Hyland and Jiang (2017, 41–42) further explain that formality is a significant characteristic in an academic writing because it avoids the writing content from vagueness, misapprehension, and personal opinion, which can lead the writing to become less objective. Thus, the absence of formality in an academic writing can become disadvantageous to the author of academic writing and the content of the writing itself.

There have been several studies which indicated that there was a trend of spoken-styled scientific English in academic writing. Biber and Finegan (1989) found that there was a drift from written to spoken languages in academic writing. Hundt and Mair (1999) asserted that academic writing was affected by informality; Seoane and Loureiro-Porto (2005) showed that there was colloquialization in academic writing; and Hyland and Jiang (2017) reported that there was a small increase of features of informality in academic writing. Those studies indicated that the formality in academic writing has decreased. In other words, academic writing, particularly among native English speakers, has become less formal from time to time.

Based on the survey on style manuals and writing guidebooks, Chang and Swales (1999) noted that there are 10 informality features, i.e., singular and plural first-person pronouns I and we that refer to the author; broad reference/unattended anaphoric pronouns such as this, that, these, those, it, and which; split infinitives which occurred when an infinitive has adverb between to and the verb; beginning a sentence with a conjunction or a conjunctive adverb; ending a sentence with a preposition; run-on sentences and expressions; sentence fragments; contractions; direct questions; and, exclamations. In a later study, Hyland and Jiang (2017) adapted those informality features and suggested the following features of informality; 1. first-person pronouns; 2. unattended anaphoric pronouns; 3. split infinitives; 4. sentence initial conjunctions or conjunctive adverbs; 5. sentence final prepositions, 6. listing expressions; 7. second person pronouns/determiners that refer to the readers; 8. contractions; 9. direct questions; and 10. exclamations.

The studies on informality have also been conducted among ESL and EFL students’ academic writings. Tang and John (1999) who conducted a study on Singaporean students’ essays found that the first-person pronoun I was used by the students to construct authorial identities in their essays. The students developed six types of authorial identities of I in their essays, they are: I as the representative, I as the guide, I as the architect, I as the recounter, I as the opinion holder, and I as the originator. McCrostie (2008) conducted a study of Japanese students who major in English. The study showed that the first-year students wrote redundantly spoken variety, plural and singular first-person pronouns, and ambiguous words in their essays. Nonetheless, the study also revealed that in the second year, the students’ redundancy in using spoken variety, first-person pronouns and incoherent words reduced. This decrease was the effect the students’ experience in writing and the university writing curriculum. Leedham and Fernandez-Parra (2017) conducted a study on Chinese, Greek, and British undergraduate and graduate students majoring in Engineering from five different universities in the UK where the data were obtained from the BAWE corpus. The study indicated that Chinese and Greek students preferred the use of the first-person pronoun we, whereas the English students preferred the use of the first-person pronoun I in their writings. The study also revealed that the use of the first-person pronouns I and we in academic writing was influenced by the students’ social and cultural background.

The previous studies suggested that academic writing in the corpora, students’ essays, and reputable journals has shifted to be more informal. Nevertheless, there has never been any study focusing on the academic writing, particularly thesis abstracts of EFL learners studying in an EFL country. This study aims at analyzing the informality features used in the thesis abstracts of Indonesian EFL undergraduate students who major in the English language study program at Airlangga University. This study also determines which informality features have increased or decreased overtime in their use in academic writing.

2. Method

In this study, the writers employed the quantitative method. Dörnyei (2007, 24) mentioned that the result of data collection techniques in a quantitative study was mainly in the form of numerical data and it was mainly examined by statistical analysis. The writers of the study used correlation as the statistical analysis for the data. The data of this study were the abstracts written by Indonesian EFL undergraduate students of the English Study Program, Faculty of Humanities, Airlangga University. Airlangga University is one of the top five state-owned universities in Indonesia. The English Study Program is one of the fast-growing study programs in this university. The continuous improvement of its education process has enabled this study program to obtain the Asean University Network quality assurance certification in 2016.

The abstracts of the theses were selected from the theses published in the years of 1992, 1996, 2000, 2004, 2008, 2012, and 2016 at the English Study Program. The selected abstracts from the years of 1992, 1996, and 2000 must earn a grade of at least B, whereas the selected abstracts from the year of 2004, 2008, 2012, and 2016 must earn a grade of at least AB. The different scores are due to the change in the grading system from five scales into seven scales in 2002. Therefore, before 2002, the top two scores were A and B, while after 2002 the top two scores were A and AB. Based on these criteria, we had a total of 114 abstracts.

The writers checked the availability of the abstracts from the first graduates of this study program, i.e., 1992 until 2016 to the librarian of the Faculty of Humanities, Airlangga University. After confirming the availability of the data, the writers obtained this inventory of Indonesian EFL students’ undergraduate theses. The inventory consisted of a number of important information. The information was the dates when the theses were published, the theses’ codes which function for indexing the students’ names, students’ numbers, thesis titles, the students’ department, the students’ grades, the years of students’ admission, and the years of graduation. The inventory information of Indonesian EFL students’ undergraduate theses guided the writers in collecting suitable data.

The writers accessed the abstracts from two online sources. The online sources were selected because the abstracts in Airlangga University’s theses were placed online. Thus, the online sources helped the writers to collect the data in an efficient time. The first source was Airlangga University Library repository, which had the abstracts collections ranging from the year of 1996 until 2016. Since some of the abstracts were not available from this first online source, the writers used the online repository of the English Department, Faculty of Humanities, Airlangga University as the second online source. The writers found that the abstracts from the year of 1992 were not available online. Eventually, the writers took the hard copies from the library of the Faculty of Humanities of Airlangga University and scanned them.

In analyzing the data, the writers processed the abstracts using the Constituent Likelihood Automatic Word-tagging System, henceforth Claws, in order to determine the part of speech for each word. In this study, the writers used Claws tagger that is available online from the website http://ucrel.lancs.ac.uk/claws/trial.html. The writers copied the abstracts manually, one at a time, into the Claws tagger. The website offered two types of tag sets which are called Claws5 or C5 and Claws7 or C7, and the writers chose the second one as it is more recent version.

After tagging the data with the parts of speech, the writers processed the abstracts using AntConc software (Anthony 2014) in order to identify the occurrence of the informal features. In this step, the writers input the selected Claws7 tag sets to the Search Term column of AntConc software. The results were retrieved through AntConc concordance menu. The writers also managed the data manually to identify informality features which cannot be retrieved using AntConc software. After retrieving the data, the writers normalized the data frequency using the number of words per thousand. The normalized data per thousand words were required because the total tokens of the abstracts were different from 1 year to another. The normalized numbers were obtained from the number of hits of informal features in each year which was divided by the number of tokens and then multiplied by one thousand.

The writers applied correlation analysis to find the relationship between two variables (X, Y). In this study, the variables were the year of the abstracts and the normalized number. Mertler (2016) stated that a correlation is a value between −1 and + 1. A value between 0 and +1 is a positive correlation which means that as X increases, Y will also increase. A value between 0 and −1 is a negative correlation which means that as X increases, Y will decrease. A value of 0 indicates that there was no relationship between X and Y. The writers employed Microsoft Excel in conducting the correlation analysis. The writers calculated the number of hits of informality features and the normalized numbers. The variables of the correlation were the year of the abstract (X) and the normalized numbers (Y). Finally, the writers interpreted the results of the data analysis. At this point, the writers explained the data based on the informality theory in academic writing.

3. Results and discussion

In this study, the writers found that there were eight types of informality features written by Indonesian EFL undergraduate students of Airlangga University. The tabulation is shown in Table 1. The informality features found are the first-person pronouns, second-person pronouns, sentence-initial conjunctions/conjunctive adverbs, sentence-final prepositions, run-on sentences/expressions, sentence fragments, contractions, and direct questions. From those eight types of informality features, only four of them constantly appeared overtime in the abstracts of the theses. They were the first-person pronouns, sentence-initial conjunctions/conjunctive adverbs, run-on sentences/expressions, and sentence fragments.

This study found that Indonesian EFL undergraduate students developed three types of authorial identities in their thesis abstracts. Hyland (2001, 215–219) explained that when the author of an academic writing used the first-person pronouns, they developed an authorial identity which attempted to emphasize on his/her contribution in the article. Indonesian EFL undergraduate students mostly developed their authorial identity of I as the recounter such as I observe (e.g., “A year ago before I was going to observe this case.”); I misunderstood (e.g., “I still misunderstood about hearing-impairment children.”); I do (e.g., “Based on maturity input theory, I do a research on kindergarteners’ ability to define words.”); I choose (e.g., “I choose four out of six categories.”); I use (e.g., “I use corpus linguistics as the approach to my analysis.”); and I determine (e.g., “To accomplish this research, I determine the typical words in female tweets then determine the typical words in male tweets before comparing the features of tweets of both genders.”). Then, it was followed by I the architect, e.g., I named. Finally, it was followed by I as the opinion holder, i.e., I conclude and I thought. The first-person pronoun I was found in the introduction, aim, result/conclusion of the Indonesian EFL the abstracts.

Indonesian EFL undergraduate students also developed their authorial identity we in their thesis abstracts. It was we as the representative that became the most revealing result, such as we know, we read, we understand, we master, we obliged, we succeed, we perform, we involved, we speak or write, we design, and we communicate. It was followed by we as the guide, i.e., we see. Finally, it was we as the opinion holder, such as we conclude, we find, we learn, we infer, and we meet. The first-person pronoun we was found in the introduction and result/conclusion of the Indonesian EFL the abstracts.

Indonesian EFL undergraduate students did not use the first-person pronoun I as the originator to claim on a new knowledge or idea as his/her own in the abstracts. It might be the influence of the teacher of academic writing in Indonesia who usually taught the students to avoid using the first-person pronoun I and replaced it with the writers or the researcher. The first-person pronoun we received a higher frequency than the first-person pronoun I. This might be influenced by Indonesian collective social culture which influenced the students to use more of the word we compared to I. This result is similar to that conducted by Leedham and Fernandez-Parra (2017) who found that the first-person pronoun we received higher frequency than the first-person pronoun I.

Different from the result of the study made by Tang and John (1999) who discovered that Singaporean undergraduate students’ essays contained all six types of authorial identity including the strongest identity I and we as the originator, in this study, the Indonesian EFL undergraduate students only used the least authoritative identities of I and we to construct authorial identity. The first-person pronoun we is the originator the strongest identity and it was employed by the author of an academic writing to claim the new knowledge or idea. Nonetheless, Indonesian EFL students did not use the first-person pronoun we to claim on a new knowledge or idea as their own in the abstracts. In this study, only the least authoritative identities of we were used by the students to construct authorial identity.

The first-person pronoun started to appear in Indonesian EFL undergraduate the abstracts in 1992 with a normalized number of 2.51. In the following 4 years, the normalized number decreased to 1.86 in 1996. Nonetheless, the normalized number increased again in 2000 to 3.12 and it reached the highest appearance in 2004 with normalized number of 3.49. The first-person pronouns were absent in 2008 and made an appearance again in 2012 with a normalized number of 1.41. In the latest one, in 2016, the first-person pronoun is not found. The writers calculated the correlation to identify whether in general the use of the first-person pronouns has increased or decreased. The correlation of the use of the first-person pronouns was −0.64, indicating a moderate negative correlation between the year of the abstracts and the normalized number of the first-person pronouns. This moderate negative correlation means that the use of the first-person pronouns decreased in the later years.

The next type of informality features used by Indonesian EFL undergraduate students was the second-person pronoun. Hyland and Jiang (2017, 44) mentioned that the second-person pronoun you was intrusive and considered taboo in academic writing because it indicated the authors’ intention to construct direct involvement with their readers. In addition, the second-person pronoun was prohibited in academic writing because it sounded like the author is offering an instruction and advice instead of information to the readers (44). The second-person pronoun you was only found one time in the introduction of Indonesian EFL thesis abstract. This means that it is very rarely used in the abstracts. Consequently, a statistical calculation cannot be made for the correlation between the year of thesis abstracts and the normalized number of the second-person pronouns.

Indonesian EFL undergraduate student theses present some examples of sentence-initial conjunctions/conjunctive adverbs. In this study, the writers found that Indonesian EFL undergraduate students employed five types of sentence-initial conjunctions in their abstracts. The types of sentence-initial conjunctions were coordinating conjunctions, adversative conjunctions, subordinating conjunctions, as as a conjunction (e.g., “As yielded by Stanford Binet that word definition is one item of a verbal intelligence test.”), and whether as a conjunction (e.g., “Whether she maintained the values of feminism or not and to know her point of view toward the new regime of China.”). The study revealed that Indonesian EFL undergraduate students used coordinating conjunctions and, adversative conjunction but, and subordinating conjunctions such as because, since, if, though, while, although, when, after, and whereas.

Indonesian EFL undergraduate students also quite often employed conjunctive adverbs at the beginning of the sentences. Conjunctive adverbs, which are called linking adverbials by Biber et al. (1999, 875), are the connectors between passages in the text which functioned to create stylistic cohesion through it five out of six types of semantic relations such as enumeration and addition, apposition, result or interference, contrast or concession, and transition. In this study, the conjunctive adverbs were additionally, also, besides, finally, first, for example, further, furthermore, hence, however, in addition, in contrast, instead, lastly, meanwhile, moreover, nevertheless, next, on the other hand, second, secondly so, then, therefore, third, thirdly, thus, and yet. Sentence-initial conjunctions/conjunctive adverbs were found in the introduction, aim, method, and result/conclusion of Indonesian EFL the abstracts.

Sentence-initial conjunctions/conjunctive adverbs began to appear in Indonesian EFL student thesis abstracts in 1992 with the normalized number of 10.05. The study revealed that in the following 4 years the normalized number slightly decreased to 9.94 in 1996. The normalized numbers significantly decreased further in 2000 to 6.02. The sentence-initial conjunctions/conjunctive adverbs reached the highest frequency in 2004 with the normalized number of 10.48. Later in the next 4 years, the normalized number decreased to 8.73 in 2008 and reached its lowest frequency in 2012 with the normalized number of 5.39. The sentence-initial conjunctions/conjunctive adverbs frequency increased again in 2016 with the normalized number of 7.57. The correlation was −0.85 which indicated a strong negative correlation between the year of the abstracts and the normalized number of the sentence-initial conjunctions/conjunctive adverbs. This strong negative correlation revealed that in the later years, the use of sentence-initial conjunctions/conjunctive adverbs was decreasing. It signified that in the later years, Indonesian EFL undergraduate students’ academic writing was getting more formal or more conservative.

The next type of informality features used by Indonesian EFL undergraduate students was the sentence-final prepositions. In this study, the sentence-final prepositions only appeared one time, i.e., “Questions often arise to categorized sentences or phrase into which categories they belong to.” It was found in the method part of the Indonesian EFL thesis abstract. Chang and Swales (1999, 167) explained that it was influenced by the editor control toward the scientific paper that minimized the appearance of the sentence-final preposition. Thus, the sentence-final preposition frequency also did not appear much in the study conducted by Hyland and Jiang (2017, 45) who asserted that the sentence-final prepositions were statistically among the lowest informal features in academic writing. Ending a sentence with a preposition to appeared in 2000 with normalized number 0.22. There was no correlation between the year of theses abstract and the normalized number of the sentence-final preposition. It indicated that these thesis abstracts very seldom use the sentence-final preposition.

The thesis abstracts from the Indonesian EFL undergraduate students show some examples of run-on sentences/expressions. Chang and Swales (1999, 149) described run-on as the nature of spoken language; therefore, it was prohibited in academic writing. Oshima and Hogue (2007, 87) described run-on sentences as sentence errors or fragments or incomplete sentences, which did not convey a complete thought. Run-on sentences occurred when the author combined two independent clauses without a comma or a coordinating conjunction (87). In this study, the Indonesian EFL undergraduate students tend to write punctuations and conjunctions incorrectly; thus they lead to run-on sentences. The run-on sentences were found in the introduction, method, and result/conclusion of Indonesian EFL thesis abstracts. Nonetheless, the result indicated that the number of run-on sentences tend to decrease in the later years.

Chang and Swales (1999, 154) asserted that run-on expressions in an academic writing indicated incomplete thought on the rest of the list. In this study, the case can be seen in the use of the abbreviation etc., which appeared seven times in the thesis abstracts and the full form et cetera which appeared one time. The run-on expressions were found in the introduction and result/conclusion of Indonesian EFL the abstracts. In general, however, Indonesian EFL undergraduate students’ academic writing was getting more formal or more conservative because the number of run-on sentences/expressions committed by the students rarely appeared overtime.

Based on the statistics, run-on sentences/expressions started to appear in the Indonesian EFL student thesis abstracts in 1992 with the normalized number of 5.03. The run-on sentences/expressions reached the highest frequency in 1992. In the following 4 years, the normalized number significantly decreased to 1.86 in 1996. The run-on sentences/expressions normalized number slightly decreased to 0.89 in 2000 and decreased further to 0.50 in 2004. The run-on sentences/expressions normalized number slightly increased to 0.62 in 2008, but it decreased again in 2012 to 0.51. Finally, the normalized number increased again to 0.72 in 2016. The correlation is −0.74 which indicated a strong negative correlation between the year of the abstracts and the normalized number of the run-on sentences/expressions. This strong negative correlation means that in the later years, the use of run-on sentences/expressions has decreased. This could mean that the Indonesian EFL undergraduate students’ academic writing tend to become more formal because the use of run-on sentences/expressions in their thesis abstracts has decreased over time.

The next type of informality features committed by Indonesian EFL undergraduate students was sentence fragments. Chang and Swales (1999, 148) defined sentence fragments as the sentences which did not have a subject, a verb, or an object or a dependent clause. In this study, sentence fragments mostly occurred because the sentence was with lack of the main verbs. Besides, sentence fragments also occurred because the student committed sentence-initial conjunctions which turned the sentence to a dependent clause e.g., “And the study which handle language in use is Discourse Analysis.” In this study, sentence fragments were found in the introduction, method, and result/conclusion of Indonesian EFL the abstracts.

Based on the statistical calculation, the sentence fragments started to appear in Indonesian EFL student thesis abstracts in 1992 with the frequency of normalized number of 3.77. The sentence fragments also reached the highest frequency in 1992. In the following 4 years, the normalized number significantly decreased to 1.86 in 1996, which decreased further to 1.34 in 2000. The decrease continues to the following years of 2004, 2008, and 2012, with the normalized numbers of 1.00, 0.83, and 0.39, respectively. The lowest normalized number was in 2016, i.e., 0.34. The correlation between the year of the abstracts and the normalized number of sentence fragments was −0.89, which indicated a strong negative correlation. This result may indicate that the academic writing of the students have become more formal. In addition, it may also indicate an improvement in the academic writing skills of the students. The English study program in this university has continuously increased its teaching and learning process, and it has obtained the Asean University Network Quality Assurance in 2016.

The next type of informality features made by Indonesian EFL undergraduate students was the contractions. Chang and Swales (1999, 149) argue that contraction was normally found in spoken language; therefore, it is prohibited to be used in academic writing. Biber et al. (1999, 165) mention that the first type of contraction was not-contraction e.g., aren’t, isn’t, can’t and won’t. The second type of contraction was operator contraction e.g., it’s not and we’re not. Biber et al. (1999, 166) also explained that only full form words are suitable for academic writing. In this study, the use of the not-contraction is only found once. The contraction was found in the result/discussion of the thesis abstract. The reduced form of does not, i.e., doesn’t, appeared in 1992 and the normalized number was 1.26. This means that there is no correlation between the year of the abstracts and normalized number contraction. This means that the Indonesian EFL undergraduate students are aware that contractions may not be used in academic writing.

The last type of informality features committed by Indonesian EFL undergraduate students was the direct questions. Chang and Swales (1999, 149) explained that direct questions naturally occurred in a conversation because questions needed participants to give a response; besides, it also increased the degree of the author’s personal concern. Hyland (2002, 3–4) asserted that direct questions in academic writing functioned as interactional and persuasive purposes to involve the readers in the writing. There was only one appearance of the direct question in the Indonesian EFL undergraduate student thesis abstracts. It was found in the introduction of the abstract. The direct question is as follows: “If God is All-good, why are the pains inserted toward the life of the innocent people?” This direct question appeared in the 1996’s thesis abstract and the normalized number was 0.62. As with the previous cases, where the occurrence is only once, the statistical correlation calculation cannot be made. The rare use of direct questions in the abstracts can be related to the fact that teachers generally do not recommend the use of direct questions in the thesis abstracts. From 2000 until 2016, no thesis abstracts use direct questions.

4. Conclusion

Based on the findings and the analysis of the data in this study, it can be concluded that there are eight types of informality features found in the Indonesian EFL undergraduate student theses abstracts. These eight types of informality features are as follows: (1) first-person pronouns, (2) second-person pronouns, (3) sentence-initial conjunctions/conjunctive adverbs, (4) sentence-final prepositions, (5) run-on sentences/expressions, (6) sentence fragments, (7) contractions, and (8) direct questions. From those eight informality features, only four of them have constantly been used by these EFL students in the thesis abstracts. These four informality features are the first-person pronouns, the sentence-initial conjunctions/conjunctive adverbs, the run-on sentences/expressions, and the sentence fragments. The correlation calculation shows that there is a negative correlation between the use of these four informality features and the years when the thesis abstracts were published. This suggests that the use of informality features has decreased overtime among these EFL students. This can be due to the improvement in the teaching and learning process in this university and the tendency for the EFL learners to become more conservative in their academic writing.