Lexical Bundles in Research Articles in Chemistry: A Structural Analysis (Research Paper)

Lexical Bundles in Research Articles in Chemistry: A Structural Analysis

^[1]Javad Zare*

^[2]Leila Valipouri

   Research Paper                                                          IJEAP- 2103-1697                  DOR: 20.1001.1.24763187.2021.10.2.7.3

Received: 2021-04-16                       Accepted: 2021-07-02                              Published: 2021-07-04

Abstract

This paper reports the results of a corpus-driven study that investigated the structural frequency and diversity of lexical bundles in chemistry research articles. The investigation was based on a corpus of 1,185 chemistry research articles, totaling four million words. WordSmith was used to generate four-word lexical bundles and their concordance lines. Altogether, 223 lexical bundles were found. More than 55% of these bundles were phrasal; 35% of them were clausal. ‘Prepositional phrases + of’ bundles were the most frequently used bundles. Bundles with a passive verb, followed by a prepositional phrase fragment were the second most frequent bundles in the entire corpus. Bundles with the structure ‘noun phrase + of’ were the third and second most frequent bundles in the corpus and among phrasal bundles, respectively. In terms of diversity, bundles with ‘noun phrase + of’ and ‘passive + prepositional phrase fragment’ were the most varied and bundles with ‘pronoun/noun phrase + be’ structure were the least varied bundles in the entire corpus. Prepositional phrase bundles outnumbered noun phrase bundles in terms of frequency, whereas noun phrase bundles outnumbered prepositional phrase bundles in diversity. Altogether, the results show that frequency and diversity correlate with the type of phrasal lexical bundle. Moreover, the study suggests that different discourses are associated with different sets of lexical bundles with different frequency and diversity, due to the different communicative functions they follow. The paper ends with implications for future EAP research, materials development, and pedagogy.

Keywords: Lexical bundles; Research articles; Structural analysis; Chemistry research articles

1. Introduction

Research shows that language is formulaic in nature (e.g., Wray, 2002). Besides, it is claimed that awareness of formulas, i.e. recurrent multi-word combinations, facilitates language learning and leads to successful language production (Conklin & Schmitt, 2012). Research also shows that phrases are learned as unanalyzed wholes or chunks rather than individual words and that learning relies heavily on these expressions in the early stages of language acquisition (e.g., Ellis, 2002; Wray, 2002; Staples, Egbert, Biber & McClair, 2013). Some even equate frequent use of appropriate formulaic sequences or lexical bundles with language development and their absence with “lack of mastery of a novice writer in a specific disciplinary community” (Li & Schmitt, 2009, p. 86) (e.g., Ellis, 1996; Ellis & Simpson-Vlach, 2008). As “recurrent expressions, regardless of their idiomaticity, and regardless of their structural status” (Biber, Johansson, Leech, Conrad & Finegan, 1999, p. 990) that “show a statistical tendency to co-occur” (Biber & Conrad, 1999, p. 183), lexical bundles have received extensive attention in research studies (e.g., Biber et al., 1999; Hyland, 2008a, b; Zare & Naseri, 2020). Research shows that “a distinct set of lexical bundles, associated with [its] typical communicative purposes” is peculiar to a certain genre or register (e.g., Biber & Barbieri, 2007, p. 265). More importantly, bundles are reported to be associated with disciplines (e.g., Cortes, 2004; Hyland, 2008a). Hence, discipline-specific studies of lexical bundles are required. Although the literature abounds with studies of the use of lexical bundles in the discourse of different disciplines (e.g., Cortes, 2013; Coxhead & Byrd, 2010; Hyland, 2008a, b), to the best of our knowledge, no prior study has investigated the structural features of lexical bundles in chemistry research articles. To this end, the present study aimed to investigate the frequency and diversity of the structural patterns of lexical bundles in chemistry research articles (RAs).

2. Review of the Related Literature

Lexical bundles are defined as “words which follow each other more frequently than expected by chance, helping to shape text meanings and contributing to our sense of distinctiveness in a register” (Hyland, 2008a, p. 5). According to Biber et al. (1999), lexical bundles are sequences of three or more words that frequently recur in a genre or register and are usually identified with the use of a computer program. Typically, bundles are not idiomatic in meaning (Biber, Conrad & Cortes, 2003), and do not lend themselves to complete syntactic units (Biber & Conrad, 1999; Biber et al., 1999; Stubbs & Barth, 2003), although they correlate with certain structural categories that are used for classifying them.

Generally, there are three parameters that are taken into consideration when identifying lexical bundles: length, frequency, and dispersion. Length is the number of words each lexical bundle consists of. Three to six-word lexical bundles are usually investigated in studies. Four-word bundles are the most common types of bundles because they are more frequent and varied than other bundles (Hyland, 2012). Frequency is the number of times a sequence needs to occur to be considered as a lexical bundle. Different research studies set different frequencies from 10 up to 40 times per million words in a corpus. Dispersion is the number of times a bundle needs to occur in texts. Different studies set different dispersions, ranging from three to five. Dispersion is computed to make sure that the bundle is typical of the entire corpus (Hyland, 2012).

Studies of lexical bundles have typically explored functional and structural properties of bundles. In terms of function, Hyland (2008a, b) classified lexical bundles into research-oriented, text-oriented, and participant-oriented. Research-oriented bundles are lexical bundles that deal with real-world activities, including location, procedure, quantification, description, and topic bundles. Text-oriented bundles are used to express the organization of the text, including transition, resultative, structuring, and framing signal bundles. Participant-oriented bundles are lexical bundles that turn to the reader or writer, including stance and engagement bundles. Several studies have investigated the functional features of lexical bundles which, due to limit in word count, fall out of the scope of this paper (e.g., Esfandiari & Moein, 2016; Hyland, 2008a, b; Jalali & Moini, 2018; Zare & Naseri, 2020).

In terms of structure, investigating the formal features of lexical bundles is a frequent theme in prior research (e.g., Allan, 2016; Biber et al., 1999; Biber, Conrad & Cortes, 2004; Cortes, 2013; Hong, 2019; Hyland, 2008a, b; Jablonkai, 2010; Jalali & Moini, 2018; Jalali, Moini & Alaee Arani, 2015; Pan et al., 2016; Pérez-Llantada, 2014; Qin, 2014; Rahimi Azad & Modarres Khiabani, 2018; Ruan, 2016; Zare & Naseri, 2020). Biber et al. (1999) made a distinction between phrasal (e.g., in the case of) and clausal bundles (e.g., I don't know what). Phrasal bundles may incorporate noun phrases (e.g., the nature of the) and prepositional phrases (e.g., in the context of). Clausal bundles, on the other hand, may incorporate a simple verb phrase (e.g., have a look at) or a main clause (e.g., I don't know how). Cortes (2013) divided four-word and longer lexical bundles in research article introductions from different disciplines into four main groups: (1) lexical bundles that consist of noun phrase or prepositional phrase fragments (e.g., in the present study); (2) lexical bundles that consist of verb phrase fragments (e.g., little is known about); (3) lexical bundles that include dependent clause fragments (e.g., little is known about); and (4) lexical bundles that incorporate both noun and verb phrases (e.g., the objective of this study was to evaluate). In another study, Biber et al. (2004) compared lexical bundles in university classroom teaching and text books with bundles in conversation and academic prose and divides them into Type 1, Type 2, and Type 3. Type 1 lexical bundles include verb phrase fragments (e.g., it’s going to be); Type 2 bundles incorporate dependent clause fragments besides verb phrase fragments (e.g., I want you to); and Type 3 bundles consist of noun phrases (e.g., the end of the), and prepositional phrases (e.g., of the things that). In general, Type 1 and Type 2 bundles are clausal, whereas Type 3 bundles consist of phrasal components. While lexical bundles in conversations incorporate verb phrases and clause fragments, bundles in academic prose incorporate mostly noun phrase and prepositional phrase fragments. The lexical bundles identified in textbooks consist mostly of noun phrase and prepositional phrases. Bundles in university classroom teaching, however, reflect the characteristics of both written and spoken modes of language use, as they consist mostly of both clausal and phrasal fragments. In another study, Hyland (2008b) examined variation of lexical bundles in form, function, and structure in a corpus of 3.5 million words comprising the three genres of doctoral dissertations, master’s theses, and research articles in four disciplines, i.e. electrical engineering, microbiology, business, and applied linguistics. Hyland found most of the bundles in the corpus to be parts of noun phrases or prepositional phrases and to end with prepositions, articles, and complimentizers. Structural analysis of the bundles showed that “several of these structures reflect the cautious limitations of academic discourse, typically through post-nominal modification, agent-evacuated passives and anticipatory-it patterns” (2008b, p. 48). In another study, Pan et al. (2016) investigated the use of lexical bundles by L1-English versus L2-English academic professionals in Telecommunications research papers and found that the lexical bundles that L2 writers use are mostly bundles with verbs and clause fragments (especially passive verb structures). On the other hand, the lexical bundles that L1 writers use mostly consist of noun phrases and prepositional phrases.

Generally, research on lexical bundles can be categorized into three groups: lexical bundles across proficiency levels, lexical bundles across disciplines, and lexical bundles across genres. Regarding lexical bundles across proficiency level, research shows that the use of lexical bundles varies across different proficiency levels. For example, Staples et al. (2013) investigated the use of lexical bundles in the written responses of learners with different proficiency levels and found that lower-level English learners use more lexical bundles. In other words, learners tend to use fewer bundles, as they gain proficiency in English. In another study, Ädel and Erman (2012) compared the use of lexical bundles in advanced writing by L1-Swedish English learners and native English speakers of English and found that, in terms of diversity, native speakers of English use a larger number of lexical bundles than non-native speakers of English in their writings. In another study, Qin (2014) investigated how advanced non-native English graduate students of applied linguistics at different levels of study use five-unit target lexical bundles in their academic papers and reported that “noun phrase with other post-modifier fragments” are used more frequently by writers at higher levels of study. Allan (2016) examined the three- and four-word lexical bundles found in graded readers, and investigated to what extent these bundles are affected by simplified language. He found that at B1 level, most lexical bundles are verb phrases than noun phrases. At B2 level and in FIC, however, noun phrase lexical bundles are predominant. He concluded that simplifying texts may influence the structural composition of four-word lexical bundles. That is, a higher level of simplification leads to more verb phrase lexical bundles in texts which is a feature of spoken language (Biber et al., 2004). Yet, in another study, Pérez-Llantada (2014) investigated the use of lexical bundles in expert academic writing for L1 English, L2 English, and L1 Spanish learners. Structurally, most lexical bundles were a combination of two structural units, where the last word of one unit is the beginning of the second structure (e.g., a function of the, the rest of the). Additionally, irrespective of the language variable, she found that all the lexical bundles follow the norms of academic written register. That is, the majority of the bundles comprised phrases, rather than clauses.

In terms of discipline, research shows that the use of lexical bundles varies across different disciplines. In other words, the use of lexical bundles is discipline specific (e.g., Hong & Hua, 2018; Hyland, 2008a). Investigating a corpus of 3.5 million words of doctoral dissertations, master’s theses, and research articles from the four disciplines of electrical engineering, microbiology, business, and applied linguistics, Hyland (2008a) found that “writers in different fields draw on different resources to develop their arguments, establish their credibility and persuade their readers, with less than half of the top 50 bundles in each list occurring in any other list” (2008a, p. 20). In two other studies, Cortes (2002, 2004) compared research articles in soft and hard fields and found that research articles in hard fields feature much more lexical bundles than research articles in soft fields and with different structural and functional features.

In terms of genre, research shows that the use of lexical bundles varies across different genres. For example, Biber et al. (1999) found that the number of bundles in classroom instruction is twice as much as and four times bigger than the number of bundles in textbooks and academic prose, respectively. In another study comparing three different genres, i.e. theses, dissertations, research articles, Hyland (2008a) found that the bundles used in theses and dissertations are more phrasal, whereas the bundles used in research articles are more clausal.

The plethora of studies on lexical bundles altogether suggests that investigating lexical bundles is a very important area of inquiry. Yet, due to the inconsistencies found in these studies, reaching a comprehensive image of how lexical bundles are appropriately used requires more elaborate studies. To this end, this paper sought to explore the frequency and diversity of the use of lexical bundles in chemistry research articles from a structural perspective.

3. Methodology

3.1. Corpus

The study was based on chemistry research articles corpus (CRAC), developed specifically for this study with more than four million words. CRAC consists of published articles from Elsevier’s online platform ‘ScienceDirect’. The papers included in CRAC all come from Elsevier’s well-known journals. The papers are equally distributed across the four main subject areas of chemistry, i.e. analytical chemistry, organic chemistry, inorganic chemistry, and physical/theoretical chemistry. To develop the corpus, we selected 10 well-known journals from each subject area randomly, first. However, because of access issues, only eight journals in relation to analytical chemistry were accessible and were thus selected for inclusion in CRAC. Next, eight volumes from each periodical were picked out, except for analytical chemistry journals, for which 10 volumes from each journal were gathered. Consequently, 320 volumes, published from 2003 to 2009 were selected. Finally, the papers, published in one issue from each volume, were compiled in the corpus. Altogether, the number of research papers, compiled in CRAC, reached 1,185. The number of authors was not an issue in compiling CRAC. Hence, papers with a single contributor or many authors were all gathered in CRAC. Additionally, intercultural rhetoric was not considered as an issue in selecting the papers, given the fact that we assumed the members of a specific discipline or discourse community follow their group conventions. Moreover, the English language proficiency of the authors and the status of English as a first, second, or foreign language for them were not considered as well, as we assumed the ability to publish in English well-known journal common among all the contributors.

3.2. Analytical Procedure

A corpus-driven discourse analytic approach, compliant with Hyland’s (2008b) approach, was followed in the present study. The study was corpus-driven, due to the fact that identification of the bundles was not based on an established list of lexical bundles from prior research. Instead, bundles were identified, using corpus linguistic tools. On the other hand, the discourse analytic nature of the study is rooted in the structural analysis of lexical bundles in their specific contexts. Hence, a corpus-driven discourse analytic approach was followed to extract lexical bundles from CRAC and investigate their structural features. WordSmith (Scott, 2015) was used to generate lexical bundles and concordance lines from the corpus. Three criteria were considered for identifying bundles: length, frequency, and dispersion. Because four-word lexical bundles “offer a wider variety of structures and functions to analyze” (Hyland, 2012, p. 151), the length of lexical bundles was set at four. In previous studies, frequency ranges from 10 to 40 times per million words (e.g., Biber et al., 1999; Biber, et al., 2004; Cortes, 2004; Hyland, 2008a, b). Here, a minimum frequency of 20 times per million words was set. In previous studies, dispersion or range has been set from three to five (e.g., (e.g., Biber et al.,1999; Cortes, 2013). Here, occurrence in at least five different texts was set as the range cut-off. Hence, using WordSmith, we generated a list of four-word lexical bundles that occurred at least 20 times per million words in at least five different papers. Next, in a discourse analytic approach, we used WordSmith Concord to code the structural features of the generated lexical bundles, following Biber et al.’s (1999) taxonomy. According to Biber et al., a distinction is made between clausal and phrasal bundles. Phrasal bundles fall under one of the structural groups of ‘noun phrase + of’, ‘other noun phrases’, ‘prepositional phrase + of’, ‘other prepositional phrases’. Clausal bundles, on the other hand, have the structures of ‘passive + prepositional phrase fragment’, ‘anticipatory it + verb\adjective’, ‘be + noun\adjectival phrase’, ‘adverbial clause’, ‘that clause fragment’, ‘pronoun/noun phrase + be’, or ‘others’. The distribution of each pattern was then computed and compared with others. To ensure precision in coding the structural patterns of lexical bundles, we coded them independently. A Cohen’s kappa of 0.92 was computed for inter-coder reliability. In cases where disagreements ensued in coding the bundles, a third coder’s opinion was sought. Moreover, a third coder was invited to code a random selection of 30% of the bundles. Here, a Cohen’s kappa of 0.90 was reached in terms of inter-coder reliability.

4. Results

A total of 223 four-word lexical bundles, totaling 37,756 individual cases, occurring at least 20 times per million words in at least five different articles were found in the four-million-word CRAC. More than 28 bundles occurred over 240 times per million words in CRAC, which is much higher than the set minimum frequency in this article. In addition, more than 86% of the bundles appeared in more than 50 different RAs. Among them, in the presence of, in the case of, as a function of, on the other hand, as shown in figure, the reaction mixture was, are shown in figure, is shown in figure, on the basis of, was found to be, with respect to the, in the range of, as well as the, to a solution of, in the absence of, the formation of the, the presence of the, to the formation of, at room temperature for, and a function of the were the top 20 lexical bundles, occurring at least more than 280 times in the entire corpus. Their frequency from the first to the 20^th most frequent lexical bundle ranged from 1315 to 283. These bundles appeared in at least 150 different RAs. Table 1 shows the results of structural analysis of lexical bundles in CRAC.

Table 1: Structural Analysis of Lexical Bundles in CRAC

As Table 1 shows, more than 55% of all the bundles in the entire corpus were phrasal rather than clausal, i.e. noun phrases and prepositional phrases. The bundles consisting of ‘prepositional phrases + of’ were found to be used more frequently than other bundles in CRAC. Prepositional phrase bundles, in general, accounted for more than one-third of all the bundles in the corpus. The second position is occupied by other phrasal bundles, i.e. ‘noun phrase + of’ and ‘other noun phrases’, which accounted for 22% of all the bundles. It can also be seen from Table 1 that ‘noun phrase + of’ and ‘prepositional phrase + of’ were almost three times more common than ‘other noun phrases’ and ‘other prepositional phrases’.

Clausal bundles constituted 35.21% of all the bundles in the corpus. As Table 1 shows, in terms of both frequency and diversity, bundles with ‘passive + prepositional phrase’ structure were the predominant structural category among all clausal lexical bundles, i.e. ‘adverbial clause fragment’, ‘anticipatory it + verb/adjective phrase’, ‘that-clause fragment’, ‘copula be + noun/adjective phrase’, and ‘pronoun/noun phrase + be’. Bundles with ‘passive + prepositional phrase’ structure constituted almost 60% of all the clausal bundles. All the structural groups of bundles are presented in more details below.

4.1. ‘Prepositional Phrase + of’ Lexical Bundles

Table 2 presents the frequent lexical bundles in chemistry RAs with ‘prepositional phrase + of’ structure.

Table 2: ‘Prepositional Phrase + of’ Lexical Bundles in CRAC

Altogether, as Table 2 shows, there were 29 lexical bundles of this types in CRAC which accounted for about 21.24% of bundles in the entire corpus. These bundles took the first place in the corpus with an overall frequency count of 8021, and accounted for more than 38% of all the phrasal bundles. Some of the bundles in this group were extremely frequent in the corpus and were the most frequent bundles in the whole corpus. For example, in the presence of, in the case of, and as a function of were the first three most frequent bundles in CRAC. Most of the bundles in this structural group were extended to 5-word bundles. For example, in the presence of was usually part of a larger prepositional phrase such as in the presence of the and in the presence of a.

4.2. ‘Other Prepositional Phrase’ Lexical Bundles

Table 3 presents the frequency of ‘other prepositional phrase’ bundles in chemistry RAs.

Table 3: ‘Other Prepositional Phrase’ Lexical Bundles in CRAC

Overall, as Table 3 shows, there were 26 lexical bundles of this type in the corpus. They comprised 11.52% of all the bundles in CRAC. The variety and overall frequency of these bundles were less than the previous bundles, i.e. ‘prepositional phrase + of’. Some of the bundles in this group were very frequent in the corpus. These include on the other hand, with respect to the, and at room temperature for. For example, on the other hand with an overall frequency of 720 times took the fourth position among the first 10 most frequent bundles. Altogether, bundles with prepositional phrases accounted for more than 34% all the bundles in the corpus.

4.3. ‘Noun Phrase + of’ Lexical Bundles

Among phrasal bundles, the second most frequent category was characterized by the structure ‘noun phrase + of’. Table 4 presents the frequency and diversity of different lexical bundles with ‘noun phrase + of’ structure.

Table 4: ‘Noun Phrase + of’ Lexical Bundles in CRAC

Bundles with ‘noun phrase + of’ structure constituted 19.18% of all the bundles in the corpus. In terms of variety, they outnumbered all the other phrasal bundles. As Table 4 shows, there were 52 different types of these bundles in the corpus which is much higher than the 32 different bundles with ‘prepositional phrase + of’ structure. Yet, the overall frequency of these bundles was less than ‘prepositional phrase + of’ bundles. Some of the most frequent bundles of this group were the formation of the, the presence of the, a function of the, and the effect of the. All the bundles of this type were preceded by articles the and a and followed by the, of, and a.

4.4. ‘Other Noun Phrase’ Lexical Bundles

Table 5 presents the frequency and diversity of ‘other noun phrase’ lexical bundles in CRAC.

Table 5: ‘Other Noun Phrase’ Lexical Bundles in CRAC

As Table 5 shows, compared to other phrasal bundles, i.e. ‘noun phrase + of’, ‘prepositional phrase + of’, and ‘other prepositional phrases’, this group of bundles were the least varied and frequently used ones in the corpus. There were 8 types of lexical bundles with this structure in CRAC. With a total frequency count of 1042, these bundles accounted for only 2.60% of all the bundles in the whole corpus.

4.5. ‘Passive + Prepositional Phrase Fragments’ Lexical Bundles

Table 6 shows the frequency and diversity of bundles with ‘passive + prepositional phrase fragments’ structure in CRAC.

Table 6: ‘Passive + Prepositional Phrase Fragments’ Lexical Bundles in CRAC

As Table 6 shows, totally, there were 52 different bundles with the structure ‘passive verb followed by a prepositional phrase fragment’ in the corpus. These clausal bundles accounted for 20.86% of all the bundles in CRAC. After ‘prepositional phrase + of’ bundles, they were the most frequent lexical bundles in CRAC. These bundles were much more frequent than other clausal bundles, i.e. ‘adverbial clause fragment’, ‘anticipatory it + verb/adjective phrase’, ‘that-clause fragment’, ‘copula Be + noun/adjective phrase’, ‘pronoun/noun phrase + Be’. They were more than 10 times as common as bundles with ‘that-clause fragment’ structure. Some of the bundles in this category, such as are shown in figure, is shown in figure, and was found to be were very frequent and appeared in many different texts in the corpus, occurring 128, 125, and 92 times per million words, respectively, and in 327, 315, and 214 different texts, respectively.

4.6. ‘Anticipatory it + Verb\Adjective’ Lexical Bundles

Table 7 shows the frequency and diversity of ‘anticipatory it’ bundles in CRAC.

Table 7: ‘Anticipatory it’ Lexical Bundles in CRAC

As Table 7 shows, there were 11 different ‘anticipatory it’ bundles in the corpus. They formed 4.35% of all the bundles. After ‘passive verb + prepositional phrase’, this group of bundles were the most used clausal bundles, both in terms of frequency and variety. Some of the most frequent bundles of this group were it can be seen, it was found that, and it should be noted. Most of these bundles were extended into 5- and 6-word bundles. For example, it can be seen and it should be noted are part of the larger bundles it can be seen that and it should be noted that, respectively.

4.7. ‘That-Clause Fragment’ Lexical Bundles

Table 8 shows the frequency and diversity of ‘that-clause fragment’ bundles in CRAC.

Table 8: ‘That-Clause Fragment’ Lexical Bundles in CRAC

As Table 8 shows, there were only six bundles of this type in the whole corpus. They constituted less than 2% of all the bundles in CRAC. The most frequent bundle in this group was can be seen that, with a total frequency of 178 in the whole corpus. In most cases, this bundle was part of a larger bundle can be seen that the.

4.8. ‘Be + Noun\Adjectival Phrase’ Lexical Bundles

Table 9 shows the frequency and diversity of ‘Be + noun\adjectival phrase’ lexical bundles in CRAC.

Table 9: ‘Be + Noun\Adjectival Phrase’ Lexical Bundles in CRAC

As Table 9 shows, there were only eight different types of bundles with ‘Be + noun\adjectival phrase’ structure in the corpus. Overall, they occurred 1191 times and accounted for 3.15% of all the bundles in CRAC. As Table 9 shows, is due to the was the most commonly used bundle of this type, occurring 222 times in the whole corpus.

4.9. ‘Adverbial Clause Fragment’ Lexical Bundles

Table 10 shows the frequency and diversity of lexical bundles with adverbial clause fragments in the corpus.

Table 10: ‘Adverbial Clause Fragment’ Lexical Bundles in CRAC

As Table 10 shows, there were only four different bundles with adverbial clause fragments in the corpus. Yet, in terms of occurrence, they were much more frequent (1080 times) than the six bundles with ‘that-clause’ structure (744 times). Although these bundles constituted only 2.86% of all the bundles, some of them were extremely frequent in the whole corpus. As Table 10 shows, the most frequent bundle in this group, i.e. as shown in figure, was the fourth most frequent bundle in the entire corpus. It occurred about 646 times in CRAC and in 321 different texts.

4.10. ‘Pronoun/Noun Phrase + Be’ Lexical Bundles

Table 11 shows the frequency and diversity of lexical bundles with ‘pronoun/noun phrase + Be’ structure in the corpus.

Table 11: ‘Pronoun/Noun Phrase + Be’ Lexical Bundles in CRAC

As Table 11 shows, there were only three groups of lexical bundles with ‘pronoun/noun phrase + Be’ structure. These bundles had the least variety among all the structural groups and accounted for 2.26% of all the bundles in the corpus. The reaction mixture was is the most frequent bundle in this group with 572 occurrences in the whole corpus.

4.11. ‘Other’ Lexical Bundles

Table 12 shows the frequency and diversity of lexical bundles in CRAC that could not be grouped into phrasal or clausal categories. These bundles are referred to as ‘others’ in Biber et al.’s (1999) taxonomy.

Table 12: ‘Others’ Lexical Bundles in CRAC

As Table 12 shows, there were 24 lexical bundles in the corpus that did not fell under phrasal or clausal bundles. Totally, they accounted for 9.58% of all the bundles in the corpus. Some of these bundles were very frequent in chemistry RAs. For example, as well as the and the other hand the appeared 344 and 253 times in the whole corpus, respectively.

6. Discussion and Conclusion

The present corpus-driven study explored the structural features of lexical bundles, used in chemistry research articles. Altogether, the results revealed 223 frequent lexical bundles in the four-million-word corpus of chemistry RAs, totaling 37,756 tokens. More than 55% of the bundles were phrasal. On the other hand, 35% of the bundles were clausal. This is in keeping with the results of Pan et al. (2016), Qin (2014), Cortes (2013), and Zare and Naseri (2020). As Biber et al. (2004) note, the existence of a large number of verb phrases is a feature of spoken language, rather than written texts. Moreover, “careful integration of information in academic prose requires the use of noun phrases and prepositional phrases, which leads to a shift from clausal style to phrasal style in academic prose” (Pan et al., 2016, p. 65). This finding mirrors the results of previous studies that pointed to the dominance of phrasal lexical bundles in academic texts (e.g., Biber & Conrad, 1999; Biber et al., 1999, 2004; Esfandiari & Barbary, 2017; Pan et al., 2016; Qin, 2014).

Among the bundles, in the presence of, in the case of, and as a function of were the first three most frequent formulaic sequences in chemistry RAs. All these bundles took the pattern of ‘prepositional phrases + of’ which was used more frequently than other structural categories in CRAC, accounting for more than 38% of all the phrasal bundles. This is in keeping with the results of Zare and Naseri (2020) and Hyland (2008a, b). Bundles with the structure ‘passive verb followed by a prepositional phrase fragment’ were the second most frequent formulaic sequences in the entire corpus, constituting more than 20% of all the bundles in CRAC. This mirrors the results of Biber et al. (1999). In line with Biber et al., the results of the present study show that verb phrases mostly comprise passive constructions, followed by prepositional phrases. This is opposite to what Zare and Naseri (2020) found. The corpus Zare and Naseri used mainly constituted articles from soft disciplines, i.e. linguistics and applied linguistics, whereas the articles in CRAC were compiled from hard disciplines, i.e. chemistry. Such bundles are mostly used to refer to graphical or tabular data in the articles of hard disciplines (Hyland, 2008a). The third and second most frequent category in CRAC and among phrasal bundles, respectively, was characterized by bundles with the structure ‘noun phrase + of’. These bundles which constituted more than 19% of all the bundles in the corpus outnumbered all the other phrasal bundles in variety. These results mirror the findings of Biber et al. (1999), Biber (2010), Esfandiari and Barbary (2017), Jalali and Zarei (2016), Ädel and Erman (2012), and Zare and Naseri (2020). Ädel and Erman found most of four-word lexical bundles in academic writing to incorporate noun or prepositional phrases. As Biber (2010) notes, “70% of the common bundles in academic prose consist of a noun phrase with an embedded prepositional phrase fragment (e.g., the nature of the) or a sequence that bridges across two prepositional phrases (e.g., as a result of)” (p. 172). Moreover, the bundle on the other hand was also among the first 10 most frequent bundles in the whole corpus. This is in keeping with the findings of Hyland (2008a) who found this bundle as the most frequent formulaic sequence in his corpus of electrical engineering, microbiology, business studies, and applied linguistics written texts. The predominance of noun and prepositional phrases in this study is in line with Qin’s (2014) observation that “noun phrases with post-modifier fragments, including prepositional phrases or past participle phrases, are less likely to appear in non-native graduate writers’ writing than in expert writers’ academic discourse” (p. 225). Qin relates this to “the inherently complex structural forms of these bundles, which require writers to pack their message or information in the most economical manner, an important feature of academic writing” (Biber et al., 1999, as cited in Qin, 2014, p. 225).

Among clausal bundles, those with the structure ‘passive + prepositional phrase’ were the predominant structural category in terms of both frequency and diversity. Some bundles with adverbial clause fragments such as as shown in figure, though accounting for less than 3% of all the bundles, were among the most frequent bundles in the whole corpus. Additionally, lexical bundles with ‘that clause fragments’ were the least frequently used formulaic sequences in chemistry RAs.

In terms of diversity, ‘noun phrase + of’ and ‘passive + prepositional phrase fragment’ bundles were the most varied and bundles with ‘pronoun/noun phrase + be’ structure were the least varied formulaic sequences in the corpus. What is important to note is that in terms of frequency, lexical bundles with prepositional phrases outnumbered bundles with noun phrases. However, in terms of diversity, lexical bundles with noun phrases outnumbered bundles with prepositional phrases. This may be taken to indicate that frequency and diversity correlate with the type of phrasal lexical bundle used. Altogether, the results of this research, along with other studies, suggest that different discourses are associated with different sets of lexical bundles with different frequency and diversity, due to the different communicative functions they follow (Tseng, 2018).

Writing a well-developed research paper in English is a very important, yet demanding task. On the other hand, awareness of the recurrent phrases, i.e. lexical bundles, used in research papers, is of great help to the writers of this discourse. Yet, developing a corpus-driven list of lexical bundles is not an easy task. Hence, the findings of this corpus-driven analysis, though by no means conclusive, are useful on many levels. First, the list of generated lexical bundles can be used as basis for comparative research on lexical bundles in other genres. Second, the findings can be used in EAP materials development and pedagogy in chemistry. Therefore, EAP material developers may use the generated list of bundles in their materials for chemistry students and EAP instructors may focus their chemistry students’ attention on these bundles when teaching how to write an effective research paper in English. Practice with the frequent lexical bundles and their contextualized examples raises the chemistry students’ awareness of the kind of language they need to develop in order to be able to publish the results of their research studies.

The findings of this study need to be treated with some caution, due to the following limitations and delimitations. First and foremost, the corpus on which we based our analysis was limited to journals, published by Elsevier only. Second, identifying lexical bundles in this study was only based on their length, frequency, and dispersion. Calculating MI score was not possible. Computing the MI score helps us understand if the words that occur together in a phrase occur more often than expected by chance (Ellis, Simpson-Vlach, & Maynard, 2008, p. 380). Third, due to the subjective nature of coding, different measures need to be taken in order to ensure precision. Fourth, other aspects such as intercultural rhetoric, the number of authors per article, their English language proficiency level, and the status of English as their first, second, or foreign language were not considered in the paper. Hence, future studies need to base their analyses on more comprehensive corpora, compiled from articles, published by well-known publishers, compute the MI score, take different measures to increase the objectivity of coding, and consider aspects of the contributors of papers such as intercultural rhetoric, the number of authors per article, their English language proficiency level, and the status of English as their first, second, or foreign language. Further research may also investigate the diversity of lexical bundles across the sub-fields of chemistry and other disciplines.

References

Ädel, A., & Erman, B. (2012). Recurrent word combinations in academic writing by native and non-native speakers of English: A lexical bundles approach. English for Specific Purposes, 31, 81-92.

Allan, R. (2016). Lexical bundles in graded readers: To what extent does language restriction affect lexical patterning? System, 59, 61-72.

Biber, D. (2010). Corpus-based and corpus-driven analyses of language variation and use. In B. Heine & H. Narrog (eds.), The Oxford handbook of linguistic analysis (pp. 159-191). Oxford: Oxford University Press.

Biber, D., & Barbieri, F. (2007). Lexical bundles in university spoken and written registers. English for Specific Purposes, 26, 263-286.

Biber, D., & Conrad, S. (1999). Lexical bundles in conversation and academic prose. In H. Hasselard & S. Oksefjell (eds.), Out of corpora: Studies in honour of Stig Johansson (pp. 181-189). Amsterdam: Rodopi.

Biber, D., Conrad, S., & Cortes, V. (2003). Lexical bundles in speech and writing: An initial taxonomy. In A. Wilson, P. Rayson, & T. McEnery (eds.), Corpus linguistics by the lune: A festschrift for Geoffrey Leech (pp. 71-92). Frankfurt, Germany: Peter Lang.

Biber, D., Conrad, S., & Cortes, V. (2004). If you look at ...: Lexical bundles in university teaching and textbooks. Applied Linguistics, 25, 371-405.

Biber, D., Johansson, S., Leech, G., Conrad, S., & Finegan, E. (1999). Longman grammar of spoken and written English. London: Longman.

Conklin, K., & Schmitt, N. (2012). The processing of formulaic language. Annual Review of Applied Linguistics, 32, 45-61.

Cortes, V. (2002). Lexical bundles in academic writing in history and biology (Doctoral dissertation). Flagstaff, Arizona: Northern Arizona University.

Cortes, V. (2004). Lexical bundles in published and student disciplinary writing: Examples from history and biology. English for Specific Purposes, 23, 397-423.

Cortes, V. (2013). The purpose of this study is to: connecting lexical bundles to moves in research article introductions. Journal of English for Academic Purposes, 12, 33-43.

Ellis, N. C. (1996). Sequencing in SLA: Phonological memory, chunking, and points of order. Studies in second language acquisition, 18, 91-126.

Ellis, N. C. (2002). Frequency effects in language processing. Studies in Second Language Acquisition, 24, 143-148.

Ellis, N. C., & Simpson-Vlach, R. (2008). Formulaic language in native and second language speakers: psycholinguistics, corpus linguistics, and TESOL. TESOL Quarterly, 42, 375-396.

Ellis, N. C., Simpson-Vlach, R., & Maynard, C. (2008). Formulaic language in native and second-language speakers: Psycholinguistics, corpus linguistics, and TESOL. TESOL Quarterly, 42(3), 375–396.

Esfandiari, R., & Barbary, F. (2017). A contrastive corpus-driven study of lexical bundles between English writers and Persian writers in psychology research articles. Journal of English for Academic Purposes, 29, 21-42.

Esfandiari, R., & Moein, G. (2016). A corpus-driven investigation into lexical bundles across research articles in food science and technology. English Language Teaching, 3(1), 1-30.

Hong, J. (2019). Structural use of lexical bundles in the rhetorical moves of L1 and L2 academic writing. English Teaching, 74(3), 29-54.

Hong, A. L., & Hua, T.  K. (2018). Specificity in English for academic purposes (EAP): A corpus analysis of lexical bundles in academic writing. 3L: The Southeast Asian Journal of English Language Studies, 24(2), 82-94.

Hyland, K. (2008a). As can be seen: Lexical bundles and disciplinary variation. English for Specific Purposes, 27, 4-21.

Hyland, K. (2008b). Academic clusters: Text patterning in published and postgraduate writing. International Journal of Applied Linguistics, 18, 41-62.

Hyland, K. (2012). Bundles in academic discourse. Annual Review of Applied Linguistics, 32, 150-169.

Jablonkai, R. (2010). English in the context of European integration: A corpus-driven analysis of lexical bundles in English EU documents. English for Specific Purposes, 29, 253-267.

Jalali, Z., & Moini, M. R. (2018). Structure of lexical Bundles in introduction section of medical research articles. Procedia Social and Behavioral Sciences, 98, 719-726.

Jalali, Z., Moini, M. R., & Alaee Arani, M. (2015). Structural and functional analysis of lexical bundles in medical research articles: A corpus-based study. International Journal of Information Science and Management, 13(1), 51-69.

Jalali, H. & Zarei, G. (2016). Academic writing revisited: a phraseological analysis of applied linguistics high-stake genres from the perspective of lexical bundles. The Journal of Teaching Language Skills (JTLS), 7(4), 87-114. 

Li, J., & Schmitt, N. (2009). The acquisition of lexical phrases in academic writing: a longitudinal case study. Journal of Second Language Writing, 18, 85-102.

Pan, F., Reppen, A., & Biber, D. (2016). Comparing patterns of L1 versus L2 English academic professionals: Lexical bundles in Telecommunications research journals. Journal of English for Academic Purposes, 21, 60-71.

Pérez-Llantada, C. (2014). Formulaic language in L1 and L2 expert academic writing: Convergent and divergent usage. Journal of English for Academic Purposes, 14, 84-94.

Qin, J. (2014). Use of formulaic bundles by non-native English graduate writers and published authors in applied linguistics. System, 42, 220-231.

Rahimi Azad, H., & Modarres Khiabani, S. (2018). Lexical bundles in English abstracts of research articles written by Iranian scholars: Examples from humanities. Iranian Journal of Applied Language Studies, 10(2), 149-174.

Ruan, Z. (2016). Lexical bundles in Chinese undergraduate academic writing at an English medium university. RELC Journal, 48(3), 327-340.

Scott, M. (2015). WordSmith tools version 6. Lexical Analysis Software.

Staples, S., Egbert, J., Biber, D., & McClair, A. (2013). Formulaic sequences and EAP writing development: Lexical bundles in the TOEFL iBT writing section. Journal of English for Academic Purposes, 12, 214-225.

Stubbs, M., & Barth, I. (2003). Using recurrent phrases as text-type discriminators: A quantitative method and some findings. Functions of Language, 10, 65-108.

Tseng, M. (2018). Creating a theoretical framework: On the move structure of theoretical framework sections in research articles related to language and linguistics. Journal of English for Academic Purposes, 33, 82-99.

Wray, A. (2002). Formulaic language and the lexicon. Cambridge: Cambridge University Press.

Zare, J., & Naseri, Z. S. (2020). Lexical bundles in English review articles. Iranian Journal of English for Academic Purposes, 9(1), 41-56.

[1]Assistant Professor of Applied Linguistics (Corresponding Author), javadzare@gmail.com; Kosar University of Bojnord, Bojnord, Iran.

[2] M.A. in TEFL, leila_valipoor_g@yahoo.com; University of Kashan, Kashan, Iran.