Exploring lexical bundles in recent published papers in the field of applied linguistics

4.1.1. Proportion of lexical bundles in the ALC

A total of 1,045 different lexical bundles, regardless of their frequency, were identified in the ALC. The most frequent 3-, 4-, 5-, and 6-word lexical bundles were as well as, on the other hand, at the end of the, and It is important to note that respectively. Some of these bundles are also identified as frequent bundles in the literature (See Hyland 2008). The 1,045 recurrent expressions amount to 1,652,275 words (words in bundles with different length multiplied by their frequency) which make up around 10% of the 16 million words in the ALC, highlighting the point that lexical bundles with different length contribute considerably to discourse of applied linguistics academic writing. Figure 1 illustrates this point.

Figure 1.

Percentage of all lexical bundles in the 16-million-word corpus.

4.1.2. Structural analysis of ALC lexical bundles

Structural analysis of lexical bundles generated from the ALC is discussed according to the structural classification proposed by Biber, Conrad, and Cortes (2004). Table 2 illustrates lexical bundles that incorporate verb phrases.

As shown in Table 2, there are many 3-word lexical bundles which do not expand to larger ones in the ALC. For example, the bundle were asked to has occurred 1,399 times in ALC and is not expanded to any larger lexical bundle because further manual analysis of the concordance lines including this lexical sequence showed that it has not combined with any other words to make a frequent 4-word, 5-word, or 6-word lexical bundle. This bundle which incorporates a passive verb mostly pointed to situations where participants of a study needed to perform an action. The following is an example of this lexical bundle:

(1) Once again, teachers were asked to make recommendations about fictional students based on their academic achievement as revealed by the grades on their record cards. Teachers were then invited to reflect on the reasons behind their decisions.

The bundle needs to be (see Table 2) is another example which has occurred 839 times (more than fifty times per million words) and, as the approach to identify lexical bundles in the methods section suggests, is not expanded to any larger lexical bundle in the papers published in the field of applied linguistics. This lexical bundle is always followed by a past participle in the ALC which makes a passive voice in texts. It should also be noted that although in the study by Hyland (2008) this bundle was marked as “a locative or logical relation” (p. 11), the bundle needs to be generally refers to an urgent situation in the ALC.

(2) However, perhaps the larger issue that needs to be addressed at this juncture are the differences between beliefs and ideologies.

Some bundles; however, may seem controversial. As shown in Table 2, recurrent expressions found to be, was found to be, is based on the, is based on the fact that, can be explained by, and be explained by the fact that may seem to be only three different recurrent expressions combined with some words to emerge as larger bundles and; thus, should not be regarded as six different lexical bundles. However, this is not the case. As explained in the method section, all larger lexical bundles were removed from the ALC before the investigation of smaller ones. This means that a 4-word lexical bundle like can be explained by has been frequent enough to meet the frequency cut off point (forty times per million words) and range cut off point (occurring at least once in ten different parts of the corpus) after removing all of the instances of its 6-word be explained by the fact that counterpart. The following are examples of concordance lines of these kinds of bundles:

(3) Gestures have been found to be used for an array of purposes, most of which have been hitherto unreported in the second language studies literature, though some are well-documented in research involving native speakers of English.

(4) The reliability index was found to be acceptable (34 participants, 51 items, Cronbach’s alpha =.79).

(5) As discussed in Section 1.2 above, register research is based on the claim that all linguistic variation is functionally associated with the situational context.

(6) This is based on the fact that grammar learning is a complex process and relies on both memorization of individual items and processing of the relationships among them.

As it is apparent from the first two extracts above, the smaller bundles are not preceded or followed by the same words as the larger bundles do. These results indicate that 3-word lexical bundles reasonably contribute to the discourse of applied linguistics academic writing. As a result, their omission, as it is common in some previous studies (see Cortes 2004; Hyland 2008, for instance), in the belief that they usually expand into 4- or 5-word lexical bundles, may adversely affect the results.

The next structural subcategory proposed by Biber, Conrad, and Cortes (2004) refers to lexical bundles which consist of dependent clauses. Examples of the ALC bundles which fit into this subcategory are presented in Table 3.

Regardless of their length, these bundles were structurally in line with those previously found by Biber, Conrad, and Cortes (2004).

Most frequent lexical bundles which fit into the third structural subcategory proposed by Biber, Conrad, and Cortes (2004), bundles with noun phrases and prepositional phrases, are given in Tables 4 and 5. The reason for dividing this subcategory into two tables in this study is that prepositional phrases could act as adjectives or adverbs in sentences in the ALC; therefore, it was necessary to present them in a new table in order to provide further explanation about the subcategory under discussion.

Table 4 also illustrates bundles with varying lengths and the only one which may seem to be expanded to another lexical bundle is English as a foreign language. This bundle seems to be combined into learning English as a foreign language. The following extracts show that English as a foreign language is mostly used to modify a following noun in the ALC; however, the bundle learning English as a foreign language does not act as an adjective phrase. Accordingly, smaller-length lexical bundles may seem to simply expand into larger ones; however, their structural and functional characteristics may change as they expand, as seen in the results of this study. This point makes the presence of bundles smaller or larger than 4-word ones significant to ensure more comprehensive results.

(7) The effects of focused and unfocused written corrective feedback in an English as a foreign language context.

(8) This study investigated children’s motivation for learning English as a foreign language (EFL) and intrinsic motivation for learning in general.

Table 5 presents lexical bundles with prepositional phrases. As stated above, they can act as adjectives or adverbs in sentences. Lexical bundles which function as adjectives are presented before those which act as adverbs.

The following are examples of bundles which consist of prepositional phrases functioning as adjectives in sentences:

(9) After piloting the test with 30 students in another class from the same department, 14 items were eliminated in cases where more than one answer could be correct or if the item was too easy.

(10) Although the learners in our study were highly familiar with each other and used to working together in class, some pairs and groups were more collaborative and produced more LREs than others, therefore creating more L2 learning opportunities.

As shown in extract (9), the bundle from the same modifies the noun class and acts as an adjective. The recurrent expression in our study functions as an adjective by giving more information about its preceding noun learners in extract (10).

Some lexical bundles in Table 5 function as adverbs in the ALC:

(11) […] findings were also supported in the sense that verb movement and finiteness are not entirely related, as has been demonstrated in L1 acquisition.

The lexical bundle in the sense that in extract (11) functions as an adverb, in that it expresses the relationship between object of a proposition and the verb phrase were supported to answer the question of how? Or under what conditions? The findings in that study were supported.

Although a considerable number of lexical bundles found in the ALC fit into the structural categorization proposed by Biber, Conrad, and Cortes (2004), some lexical bundles found in the applied linguistics academic writing expressed new structural characteristics. The explanation regarding the new lexical bundles found in the ALC is as follows.

There are some lexical bundles which do not include a pure verb phrase, dependent clause, noun phrase, or prepositional phrase but functioned as adjective phrases in the ALC. These bundles are presented in Table 6. It should also be noted that the bundles has been shown to be and can be seen in the seem to fit under the subcategory of lexical bundles with verb phrases, but further concordance line analysis showed that these two bundles demonstrate a specific structural identity, acting as adjective phrases, they can be classified under the structural category of bundles as a new subcategory which act as adjective phrases and are called, structurally, adjectival phrases as well.

To illustrate the identity of the bundles presented in Table 6, an example of these bundles in the ALC is presented:

(12) In addition to a short questionnaire aimed at eliciting biographical information, the instruments used in the present study were (a) difficulty judgement questionnaires for all participants and (b) tests of explicit and implicit L2 knowledge for the learners.

As it is obvious in extracts (12) the lexical bundle used in the present study modifies instruments. This means that this bundle has modified its preceding noun, acting as adjectives. Adjectival phrases as lexical bundles found in the ALC are presented in Table 6.

There were also bundles which acted as adverbs in text and are called, structurally, adverbials: so that the, depending on the. The following is an extract from the ALC to discuss this kind of bundles in more detail:

(13) Circumlocutions are classified depending on the features selected for description. Thus, we distinguish among function, superordinate, description of physical characteristics of object, and locations

The lexical bundle depending on the in extract (13) modifies its preceding verb classified. This shows two points: first, lexical bundles smaller that 4-word ones can perform a specific action in text. Second, lexical bundles can act as adverbs in texts as well.

Some lexical bundles in the ALC incorporate dependent clauses which are not previously found in the literature. They can act as adjective clauses or adverb clauses in texts. Table 7 presents examples of this kind of bundles. It should be noted that some of the bundles presented in this table may also concord with the primary structural tables, but as they act as adjectives or adverbs specifically, they are particularly presented in this table. For a more detailed discussion of them extract (14) is generated from the ALC:

(14) They would like their teachers to provide some more specific follow-up questions, so that they can realize what their errors and mistakes are after clarification request feedback.

As extract (14) indicates, the bundle so that they can modifies the verb provide, in that the action of “providing” is done for a specific reason and the bundle so that they can answers the question of “why?” by expressing the relationship between its preceding and following propositions. In addition, these results show that lexical bundles smaller or larger than 4-word ones can contribute reasonably to the discourse of academic writing in the field of applied linguistics.

The next newly identified type of lexical bundles in the ALC incorporated a specific structural pattern. As shown in Table 8 the specific structure is Be + Adjective/Adverb + to. The Corpus Query Language (CQL) function of Sketch Engine was used to identify the frequency of this specific structural pattern ([lemma = “[a-z]*” & tag = “VB.*”] ([lemma = “[a-z]*” & tag = “JJ.*”]|[lemma = “[a-z]*” & tag = “RB.*”]) [word = “to.*”]). It was found that this structure occurred 20,289 times in the ALC (see Figure 2). And this number was reasonable enough to take this structure as frequent. Extracts (15) and (16) are provided to give examples and discussions for this type of bundles:

Figure 2.

Frequency of the structure be + adjective/adverb + to in the ALC generated by sketch engine.

(15) That is, for a learner to understand a word’s meaning in listening, s/he must be able to recall (rather than recognize) that meaning once the form is recognized.

(16) Slightly better control in grammar, vocabulary use, and fluency by the former group may be due to the fact that students in the F2F class had an opportunity to freely and easily clarify grammar facts in each class whenever they wanted to.

These extracts show that the lexical bundle structure be + adjective/adverb + to can be used to express the ability (see extract 15), or the possibility of a specific reason for a result in a study. The extracts (15), and (16) further show that these 4-word lexical bundles found in the ALC are not expansions of 3-word lexical bundles since the lexical bundle be able to is not followed by the word use to argue for that kind of expansion (be able to use is a 4-word lexical bundle in Table 8).

ALC also included lexical bundles which began with a conjunction. These bundles are presented in Table 9.

ALC also included lexical bundles which incorporate infinitive phrases. Table 10 presents most frequent examples of these bundles in the ALC. Extracts below are given to show how these bundles are used in sentences:

(17) Rather than providing a description of the various components that contribute to the subjects’ motivation and identifying the correlation between motivation and learning outcomes as in social-psychological or cognitive approaches to L2 motivation, this study aims to address the issue of motivation through the following question:

(18) Perhaps thinking in the L1 or resorting to the L1 may be “the only way a leaner can ever begin to communicate in a second language” […]

The lexical bundle to address the in extract (17) is used to signal the purpose of that study. In extract (18), the bundle to communicate in a second language is used to show the result of an assumption. In just two examples, two different functions of this kind of bundles (lexical bundles which incorporate infinitive phrases) were shown: to signal the purpose of studies, and to show the result of an assumption.

Up to now, the structural classification of lexical bundles in the ALC has been presented. The next section deals with the functional classification of the lexical bundles in the ALC.

4.1.3. Functional Analysis of ALC lexical bundles

The classification used for the functional analysis of the ALC lexical bundles was the one developed by Hyland (2008). It should be noted that, both Biber, Conrad, and Cortes (2004) and Hyland (2008) have developed a functional categorization for lexical bundles. The one in Biber, Conrad, and Cortes (2004) covers the discourse in classroom teaching and textbooks. On the other hand, Hyland (2008) has investigated research articles, doctoral dissertations, and master’s theses for lexical bundles. In this study, the classification proposed in the latter study is taken for consideration because of the similarity in the target discourse of that study and the present study. It was found that overall, a considerable number of bundles used in published papers in the field of applied linguistics perform functions similar to those performed by lexical bundles previously found in the literature. In addition to the functional sub-categories proposed by Hyland (2008), some other sub-categories of lexical bundles had to be added to define some new bundles found in the ALC.

Most frequent examples of the bundles found in ALC which share the similar functional identity to those identified by Hyland (2008), which are called research-oriented lexical bundles, are presented in Table 11. It should be noted that, in line with the findings of Hyland (2008), noun phrase + of structures found in ALC were mostly associated with research-oriented function.

The next sub-category of the functional classification proposed by Hyland (2008) is called text-oriented lexical bundles which are “concerned with the organization of the text and its meaning as a message or argument” (p. 13). Most frequent examples of the similar bundles found in the ALC are presented in Table 12. The pattern which was mostly associated with this function in the ALC was prepositional phrase which is in line with the findings of Hyland (2008).

The final functional sub-category of lexical bundles is called participant oriented lexical bundles (Hyland 2008) which are “focused on the writer or reader of the text” (p.14). Most frequent examples of the similar bundles found in the ALC are presented in Table 13. It should be noted that, the pattern which was mostly associated with this function was anticipatory it pattern and this was, also, in line with the findings of Hyland (2008).

Overall, although the functional classification proposed by Hyland (2008), efficiently defined a large number of lexical bundles in ALC, some bundles found in the applied linguistics academic writing expressed rather new functional identities. They are presented in the following.

Some bundles and functions newly found in the ALC tend to refer to two things in the texts: Doubling Lexical Bundles. Most frequent examples of these bundles are given in Table 14.

The following examples are given to discuss the doubling lexical bundles in more detail.

(19) We observed that L2 proficiency interrelates with L1 and L2 influence in different ways depending on the specific mechanism used, so that some mechanisms which are frequent at 4th grade decrease their presence, basically borrowing, and some other tend to increase, basically L2-based mechanisms as linguistic proficiency, awareness and cognitive development increase.

(20) Some researchers […] further argue that culturally responsive pedagogy designed to address these imbalances is relegated to the margins in favor of standardized teaching and learning, but the theoretical literacy grounding provided for educators of language-minority students tends to be founded on frameworks that centralize the norms and values of the White, native-English-speaking population […]

In extract (19) the bundle L1 and L2 is mainly related to the field of applied linguistics and has a twofold identity: first language and second language.

Another type of bundles with a new function found in the ALC referred to a point of divergence in textual context. Contrasting Lexical Bundles are presented in Table 15. The following example is given to show how contrasting lexical bundles are used in the ALC.

(21) In his study of English language and global hip-hop cultures, Pennycook attempts to elucidate the ways in which global languages and cultures offer alternative identities and forms of expression, while at the same time being reshaped to meet local needs, and then being sent back out again with new forms and meanings in a circular or flowing process.

In extract (21), the recurrent expression while at the same time is used to refer to a point of divergence, in that although different languages and cultures can provide different identities, they themselves are not immune to changes.

Another type of lexical bundles found in ALC which expressed a specific function referred to examples in texts of the register under study. The most frequent Exemplifiers in ALC are presented in Table 16.

To discuss how exemplifiers are used the following example is given:

(22) It refers to a teacher’s ability to predict language learner difficulties such as the negative transfer of rules and patterns from the L1 or speech sounds that exist in the target language but not in the L1 system.

In extract (22), the lexical bundle such as the is used to point to instances of learner difficulties.

Some bundles in the ALC were used mostly for research questions. Most frequent Questioning Lexical Bundles found in ALC are presented in Table 17. Examples below are given to show questioning lexical bundles in the text:

(23) This study was designed to address the following questions: 1. To what extent do L2 knowledge skills and L2 processing skills predict communicative success in L2 speaking, in a mixed-proficiency group of adult L2 learners?

(24) Is there a relationship between explicit and implicit attitudes towards NESTs and non-NESTs?

As it is apparent in extracts (23), and (24) these recurrent expressions are mostly used to point to or initiate research questions. Also, this function seems not to be identified in the literature before.

The last new type of lexical bundles which performed a specific function in the ALC referred to cause of something. Most frequent examples of Cause Initiator Lexical Bundles found in the ALC are presented in Table 18. Following examples are given to show cause initiator lexical bundles in the text:

(25) This could very well be due to the nature of the task itself, which focuses on exchanging cultural jokes in a task-oriented environment rather than a social environment.

(26) Those who said they like English very much were more likely to prefer NS accents and NS teachers after the project. This may be due to the fact that their preference toward English originated from their interests in Anglo-American cultures: these English lovers were also more likely to agree to the statement, ‘to learn English well, one must know a lot about American or British cultures’ (r = 0.341 p = 0.012).

The lexical bundle be due to the in extract (25) is used to refer to nature of a specific task in that study as the result of something. The recurrent expression due to the fact that in extract (26) is also used in that study to refer to the cause of why some people were more likely to prefer Native Speaker accents and Native Speaker teachers after a specific project. Also, this function of lexical bundles has not been identified in previous theoretical classifications.