1。1。2 Lexical bundles

Due to the advancements in computers and their use in the analysis of language corpora, there has been a strong shift in the study of formulaic expressions in the last two decades。 In the past, the study of formulaic language was performed intuitively, with researchers making up lists of fixed expressions that they perceived as occurring frequently in the language。 For example, Pawley and Syder (1983), in their seminal article on the importance of formulaic language to non-native speakers of English, explained that “A few minutes’ reflection produced the following sample of clauses that are familiar to the writers as habitually spoken sequences in Australian and New Zealand English。” The explanation of the authors’ methodology to identify habitual expressions was followed by a long list of short and long expressions such as Need any help? Would you like some more? You shouldn’t have said that, and you’ve hurt his feelings, which these authors perceived as frequent formulaic expressions in that geographical register。 Lately, studies have favored more empirically-based research methods to identify recurrent expressions in particular registers。 Advancements in corpus-based analysis make it possible to empirically identify fixed expressions that recur frequently in the language。 These longer expressions, as well as two-word collocations, have often been researched within two traditions: the frequency-based tradition, which emphasizes the frequency of their co-occurrence, and the phraseological tradition, which focuses both on the grade of fixedness that these word combinations hold and on different ways to classify these expressions。 Within the frequency-based tradition, some studies have reviewed the literature on formulaic expressions and checked their frequency in a corpus。 Other studies have used a strict frequency-driven approach to identify this type of expressions。 Using this frequency-driven methodology, Biber et al。 (1999) coined the term lexical bundles to label “the most frequently occurring sequences of words” in a language or register。 Lexical bundles are identified using special software on a large corpus of language。 These expressions are groups of three or more words that frequently recur in a language or in a particular register and meet arbitrarily-established cut-off points for frequency and range。 For example, Biber et al。 (1999) proposed a cut-off point of ten times per million words fora four-word expression to be considered a bundle。 In addition, those expressions had to appear in five or more texts in the corpus under analysis, to avoid users’ idiosyncrasies。 Succeeding studies have been more conservative and established higher cut-off points to ensure that these expressions are really frequent in a particular register with frequency cut-off points set at 20 or 40 occurrences per million words。 As previously introduced, frequent lexical bundles in academic writing are expressions such as a result of, in the case of, and on the other hand, among many others。 Lexical bundles have been studied in a wide variety of registers such as everyday conversation; research articles; university textbooks and lectures, doctoral dissertations and Master’s theses, and English EU documents。 These studies tried to discover tendencies in the use of this type of formulaic expressions in different types of texts and some of these investigations introduced different taxonomies to classify these expressions structurally and functionally。

1。1。3 Lexical bundles and moves as building blocks

The identification of lexical items and lexical and grammatical patterns used to signal the onset of rhetorical moves has always been a topic of interest for researchers studying discourse organization。 Swales (1981) already identified some lexical items as key markers for moves。 For example, he stated that when asserting centrality, a function often expressed in the first move of introductions, authors use expressions that reflect that the issues to be discussed in those articles raise questions of interest using words that denote interest or importance。 A wide variety of expressions have often been intuitively identified as frequently initiating or defining a move。 That is the case of expressions such as it is the purpose of this paper, or the aim of this investigation, which have been found to introduce the arrival of Move 3 (introducing the present work)。 Most of these expressions resulted from the analysis of a reduced number of texts or have been perceived as frequently used by authors to convey that communicative function。 Identifying lexical bundles in the rhetorical moves of RA introductions would empirically demonstrate which expressions are frequently used to initiate rhetorical moves in this section of RAs。 There is another important concept that brings lexical bundles and rhetorical moves together。 Each of these linguistic features has been considered building blocks to be used in the construction of discourse。 Lexical bundles have been defined as recurrent expressions that can be retrieved from our memory to be used as “text building blocks” (。 Hylandalso maintains that bundles have been increasingly seen as “important building blocks of coherent discourse and characteristic features of language use in particular settings”。 In a similar way, rhetorical moves have also been considered building blocks。 Biber, Connor, and Upton explain that move types can be seen as the “main building blocks” of a genre。 Dudley-Evans considered rhetorical moves such an inherent part of a genre that they could be used to teach novice writers how to produce successful texts in that particular genre。