Stance and academic promotionalism: a cross-disciplinary comparison in the soft sciences.
Key words: stance; evaluative adjectives; soft sciences; academic promotionalism; academic rhetoric; disciplinary writing
El concepto de 'evaluacion'-o la expresion de la 'voz' del autor-en la prosa academica ha sido tema de investigacion de especial interes. Sin embargo, los estudios sobre evaluacion en las ciencias no experimentales han sido relativamente escasos comparados con los de las ciencias experimentales. Mediante la comparacion de dos disciplinas no experimentales, este articulo explora los perfiles semantico y sintactico de los adjetivos evaluativos en las introducciones de los articulos de investigacion-una seccion en la que los escritores crean un 'hueco' en la investigacion para convencer a los lectores del valor y/o de la utilidad de su investigacion. El articulo examina tambien la funcion de estos adjetivos como recursos retoricos que los escritores emplean para evaluar los procedimientos de la investigacion y dar relevancia a los resultados de su investigacion. Los datos cuantitativos muestran que los adjetivos evaluativos comparten un perfil semantico-sintactico similar en las dos disciplinas. Sin embargo, el analisis retorico revela convenciones especificas para cada disciplina respecto al uso de los adjetivos evaluativos dentro del modelo CARS (Create-a-Research-Space) establecido para escribir introducciones. Dichas diferencias podrian atribuirse a la naturaleza del conocimiento disciplinar, asi como a las convenciones de escritura cientifica de cada sub-comunidad academica.
Palabras clave: evaluacion; adjetivos evaluativos; ciencias no experimentales; retorica academica; comunidades disciplinares; promocion academica.
To the question of how novice researchers should write academic papers, influential stylebooks recommend the canons of clarity, completeness, impartiality, accuracy and objectivity (Barras 1987; Turk and Kirkman 1996). More specifically, style guides addressed to scholars in the hard sciences claim that the main purpose of research report writing is "to explain to others in the field what the objectives, methods and findings of the study were" (Weissberg and Bucker 1990: 1). (1) The contrary is said to lead to speculation, unsupported claims, overgeneralisations and, broadly speaking, a lack of scientific rigour. Along similar lines, style manuals for the soft sciences explain that since academic prose disseminates new knowledge by focusing on entities, events and processes it should display an object-oriented character. In this vein, Wilkinson specifically remarks that academic writing "cannot, therefore, be the vehicle for the writer's personal relation to the objects studied or personal objectives in the research" (1991: 11).
However, a long-standing debate on the objective/subjective nature of research prose has been opened by linguistic research in the past decade. This research strand has investigated several structural, discoursal and functional features of linguistic expressions of writers' subjectivity and stance in published academic papers (Hunston 1993; Hyland 1998; Hunston and Thompson 2000; Conrad and Biber 2000; Biber 2006). Other studies have paid attention to the pragmatic devices and interpersonal mechanisms that scholars use to interact with audiences (Meyer 1997; Varttala 2003; Markannen and Schroder 1997; Hyland 2002). Intercultural rhetoric research has also been prolific in studies of stance in academic writing, ranging from analyses of personal pronouns (Breivega, Dahl and Flottum 2002), metatextual references (Mauranen 1993) and metacomments (Dahl 2004) to examinations of epistemic indicators and modality expressions (Vassileva 2001; Salager-Meyer 2005), among other stance markers. This broad theoretical spectrum may be taken as evidence that, although academic writing obviously takes objective positions when reporting new knowledge, it also contains subjective elements that lay overt the writers' stance.
2. Overview of the study: the lexical marking of stance
Style guides for research writing in the fields of the soft sciences tend to value content over form and emphasise that the main focus of academic prose should be conceptual. Wilkinson, for instance, explicitly recommends the avoidance of evaluative adjectives such as certain, considerable, dramatic, fair, good, great, marked, poor, possible because they are "too weak and too imprecise for use in scientific writing" (1991: 451). Notwithstanding the valuable guidance that stylebooks offer to those in need of being acculturated in the practices of academic writing, grammars of contemporary English have specifically noted that writers' overt positioning in academic prose is linguistically enacted through "the use of an evaluative lexical item, usually an adjective, main verb, or noun" (Biber et al 1999: 968). Relying on cross-register comparison, Biber et al further remark that the most common predicative adjectives (difficult, important, likely, necessary, possible, true, etc.) are used as explicit indicators of evaluation in academic prose, and that "several of the most common attributive adjectives in academic prose [are] also evaluative" (1999: 968-69)--for example, adjectives such as appropriate, good/best, important, practical, and useful.
Lexical markers of stance have been a major focus of attention of scholarly research in the past decade. Among others, Grabe and Kaplan (1997) have compared the use of stance adjectives in professional and popular writing in the hard sciences. Thetela (1997) has examined lexical evaluation in several disciplines of the soft sciences, the type of entity to which evaluation is attached and its ascribed values. Shaw (2003) has carried out an analysis of stance adjectives in introductory sections of articles on economics, Stotesbury (2003) has compared attitudinal lexis in abstracts of humanities, social and natural sciences, and Biber has studied complement clauses "controlled by adjectives" (usually in extraposed constructions) and their functional role in evaluating "the likelihood of information, or the desirability/possibility of some action or event" (2006: 110).
Because studies on stance in the soft sciences have been relatively scarce compared with the numerous investigations carried out on disciplines in the hard sciences, this paper sets out to explore the extent to which writers in the soft sciences--more specifically, those in the disciplines of applied linguistics and information science--use stance adjectives in the introductions of their papers, a section in which writers are expected to highlight the significance of the new findings in order to persuade readers of the value of these findings.
The reason for selecting these two disciplines is grounded in Becher and Trowler's (2001) classification of disciplinary communities in the soft sciences according to the nature of the knowledge they deal with. These authors distinguish between soft-pure disciplines and soft-applied disciplines. Knowledge in the soft-pure disciplines is defined as "reiterative, holistic, organic/river-like, concerned with particulars, qualities, complication; personal, value-laden; dispute over criteria for knowledge verification and obsolescence; lack of consensus over significant questions to address; results in understanding/interpretation" (Becher and Trowler 2001: 36). In the soft-applied disciplines, knowledge is "functional, utilitarian (know-how via soft knowledge), concerned with enhancement of [semi-]professional practice, uses case studies and case law to a large extent, results in protocols/procedures" (2001: 36). Following this distinction, I initially hypothesised that applied linguistics and information science, both of them considered to be soft-applied disciplines because of their experimental nature of knowledge, were going to behave similarly as regards the use of evaluative adjectives despite the recommendation of style guides to adhere to objectivity when reporting on experimental research.
To validate the hypothesis, this paper quantifies the presence of stance adjectives--and other semantic categories of adjectives--in the introductory sections of research articles in the two disciplines. The paper also describes the syntactic associations of stance adjectives across the selected disciplines and the way these associations build up what Hunston and Sinclair define as a "local grammar of evaluation" (2000: 74). In a qualitative, context-sensitive analysis, the paper looks at the functional behaviour of stance adjectives across the rhetorical moves established for writing introductions (i.e. Swales' 1990 Create-a-Research-Space [CARS] model). By this means, it explores possible explanations of the rhetorical functions that stance adjectives perform in the transmission of knowledge in each discipline.
3. Corpus and methodology
To develop the present research a pilot corpus of 20 research article introductions from the above mentioned disciplines was selected. On a broader scale, the selection of this corpus was intended to determine whether these disciplines were suitable for the compilation of the humanities component of SERAC (the Spanish-English Research Article Corpus), a corpus specifically designed for contrastive analyses of stance markers in academic prose. (2) The texts on applied linguistics were taken from the Applied Linguistics Journal (numbered AL/1 to AL/10). The texts on information science were selected from the Journal of the American Society of Information Science and Technology (JASIST/1 to JASIST/10). Both journals operate on a double-blind peer-review process. The corpus consisted of a total of 15,683 words, relatively evenly distributed across the two subcorpora (8,034 words for the AL subcorpus and 7,649 for the JASIST one). Only the introductions of these texts were selected for the analysis of stance adjectives since, as Swales remarks, this section privileges "an environment in which originality (especially in theory) tends to be highly prized, competition tends to be fierce, and academic promotionalism and boosterism are strong" (2004: 226). As such, introductions were assumed to contain a fair number of these stance markers.
Because stance adjectives are highly context-sensitive items, the frequency count and the comparative distribution of semantic groupings of adjectives in the two disciplines were carried out manually and intuitively. Later, following Swales and Burke (2003), stance adjectives were further categorised into two sub-groups: relevance and assessment evaluators. Colleagues and English native scholars supervised the categorisation of adjectives and provided useful help in those cases where semantic categories of adjectives overlapped. Grammars and dictionaries were also consulted to corroborate the decisions. For the purpose of comparison, the frequency counts of the taxonomy of adjectives in the introductions were all normalised on a 1,000-word basis.
Since the occurrence of adjectival evaluators proved to be statistically significant when compared with the other semantic groupings of adjectives, it was deemed necessary to identify recurrent syntactic associations of stance adjectives. The identification of these patterns of evaluation sought to provide a preliminary description of the preferred lexico-grammar of stance in each discipline. Although corpus limitations for substantive comparison were acknowledged, implications regarding how stance adjectives vary syntactically in the two soft sciences disciplines were expected to be found at this stage. Finally, lexico-grammatical patterns containing stance adjectives were analysed in context. More specifically, the distribution of relevance and assessment evaluators across each move of the Create-a-Research-Space model was quantified and possible explanations of the rhetorical intentions behind the use of stance adjectives were tentatively provided.
4. Results and discussion
4.1. Semantic categories of stance adjectives in introductions
As a preliminary research step, a total of 823 adjectives were manually identified in the corpus. Following Biber et al (1999: 512-13), two major groups of adjectival tokens were established: descriptors (437 tokens) and classifiers (386 tokens). The first group comprised size/amount descriptors (small, high, low, etc.), time descriptors (new, old, etc.), and evaluative descriptors (good, important, possible, likely, necessary, etc.). In the second major group both relational classifiers (used to delimit the referent of a noun, e.g. same, general, basic, following, etc.), and topical classifiers (delimiting the domain of a head noun, e.g. local, physical, social, etc.) were included. In the total count a small number of adjectives were found to behave both as a classifier and as a descriptor when looked at within a context of use. This was the case of adjectives such as different, main, specific, or particular, whose contextual meanings conveyed evaluative overtones. These instances were thus counted as evaluative markers. Table 1 below displays the distribution of the semantic domains of adjectives in the two subcorpora. (3) Raw number of tokens, frequency per 1,000 words and total frequency are shown in the table.
Comparatively, the two disciplines showed a relatively similar distribution of descriptors and classifiers. As regards classifiers, relational adjectives (basic, common, general, single, various, etc.) accounted for approximately 30% of the total occurrences in both AL and JASIST, and were used to restrict a noun's referent in relation to other referents. Much lower frequencies were found in the semantic category of topical classifiers (formal, theoretical, thematic, analytical, conceptual, etc.), with a greater use of adjectives delimiting the domain of a noun in JASIST (16.70%) than in AL (9.28%). As for the group of descriptors, both disciplines showed a very low occurrence of time/size descriptors (previous, recent, subsequent, large, etc.), with a frequency of 7.30% in AL and 6.64% in JASIST, and nil counts of colour descriptors used for describing qualities of Things. Interestingly, the sub-category of evaluators was the most frequent semantic type of adjectives in the two subcorpora (51.12% and 42.40% of the total number of adjectives in AL and in JASIST, respectively).
In addition, and consistent with Thetela (1997) and Perez-Llantada and Neumann (2006), a small repertoire of evaluative descriptors (different, important, possible, likely, difficult, better, good, best, clear and necessary) appears to be used by the two groups of scholars. This selection of adjectives accounted for 30% of the total occurrences of evaluators in the corpus, which might indicate that both AL and JASIST writers tend to select this small range of adjectives for assessing entities, facts and processes or when emphasising their value or usefulness.
Following Swales and Burke's (2003) taxonomy of evaluative adjectives in academic writing and speech, two types of evaluators were quantified--those of relevance and those of assessment. Relevance adjectives are used by writers to highlight research outcomes (central, important, main, major, relevant, unique, crucial, essential, fundamental, significant, considerable, critical, etc.). Assessment adjectives help writers assess procedural aspects of the research process (good, better, best, interesting, difficult, bad, useful, problematic, effective, necessary, complex, appropriate, etc.). Table 2 shows the distribution of the two groups of stance adjectives in the AL and JASIST texts.
Comparative totals indicate that relevance adjectives are slightly more frequent than those of assessment in the pilot corpus, while the comparative frequencies of the two categories of evaluators are also relatively similar across the two disciplines. This context-sensitive analysis of the relevance/assessment categorisation echoes Thetela's distinction between "topic-oriented evaluation" (TOE) and "research-oriented evaluation" (ROE) and, within ROE, the distinction between "research-product" and "research-process evaluative items" (1997: 113). While relevance adjectives in the corpus primarily work for evaluating research products (i.e. results, findings, data, etc.), assessment adjectives tend to function as linguistic markers for evaluating research processes, methods and procedures. Although plausible emendations should be acknowledged, the relevance vs. assessment taxonomy proved significant to complement the study of stance adjectives with the analysis of lexico-grammatical patterns of evaluation across the two disciplines (section 4.2) and of the rhetorical intentions of these adjectives across the three moves within introductions (section 4.3).
4.2. Syntactic profile of stance adjectives in introductions
Having identified similar trends as regards the frequency and semantic profile of stance adjectives in the two disciplines, the second stage of the research looked at the syntactic behaviour of these adjectives. Syntactically, the two disciplines showed a clear preference for attributive rather than predicative positions. In quantitative terms, evaluative adjectives were considerably more frequent in attributive (69.21% of the total occurrences, and a frequency of 16.76) than in predicative positions (30.79%, and a frequency of 7.46). A relatively similar distribution of attributive vs. predicative positions (65.93% vs. 34.07% in AL, and 72.22% vs. 27.78% in JASIST) was observed in the two disciplines. These data seem to support Biber et al's (1999: 506) claim that attributes are much more frequent than predicative adjectives in expository written registers such as academic prose. The subsections below detail how stance adjectives --both attributive and predicative--combine with other linguistic elements, building up recurrent patterns of evaluation.
4.2.1. Syntactic behaviour of attributive evaluators
Nominal pre-modification is the most common syntactic role of evaluative attributes in the two disciplines. Pre-head modifiers describe the quality of a head noun or nominal expression, which, in the selected texts, is usually an abstract noun (a good coverage, an important role, an attractive idea, particular interest, etc.). Though the size of the corpus is very limited, it is interesting to note that the AL introductions seem to show a preference for linguistic combinations with the relevance adjective important (an important assumption, an important connection, an important difference, an important reason). In contrast, JASIST introductions make frequent use of different with evaluative overtones (different indicators, different patterns, different conceptualisation, different levels of expertise). Also, the syntactic pattern formed by two attributive evaluators (long and often tedious process; a complex and multifaceted construct; gradual and unruly nature; a further, more systematic investigation) only occurs in AL.
For purposes of cohesion, evaluative attributors usually combine with nouns performing a phoric function in discourse (issue, idea, focus, question, etc.), as in [t]he major focus of this study is to examine, [...] has raised a critical issue of how to [...]). As detailed in section 4.3, these evaluative units perform pragmatically as they allow writers to place important ideas at the end of the clause, thereby giving greater emphasis to them. Other common linguistic items clustering with pre-head modifiers are adverbial intensifiers (a potentially important marker; a completely different problem; very great control, etc.). Writers in the two disciplines tend to use these stance adverbs to enhance the semantic force of the adjective, that is to say, to emphasise the writer's qualification of a Thing when projecting his/her stance at a textual level.
Stance is also conveyed through noun phrases with an abstract noun head which is post-modified by an of-phrase embedding an evaluative pre-modifier. Word combinations such as identification of relevant items, a problem of central interest, questions of paramount interest, etc., may illustrate a common tendency in the two disciplines to rely on evaluative adjectives for the purpose of enhancing the value of the research reported in the paper. For similar communicative intentions, evaluative attributes in the two disciplines accompany a head noun which is also pre-modified by a classifier adjective, the latter being most frequently a topical classifier (vital strategic importance; unacceptable computational complexity; tedious physical details).
The above observations suggest that the two disciplines share a similar lexico-grammar of stance, which is mostly built upon attributive adjectives. Minor discipline-specific preferences such as the choices of evaluative attributors and the use of double attributive evaluation patterns in the AL introductions would need to be corroborated with a larger sample of texts.
4.2.2. Syntactic behaviour of predicative evaluators
Quantitatively, the two disciplines displayed very low frequencies of stance adjectives in predicative positions (only 30.79% of the total number of evaluators). Also, a different repertoire of linguistic items was noticed in the analysis of predicative modifiers in each discipline. While the post-modifying adjectives interesting, critical and crucial appear to be preferred linguistic choices for the JASIST writers, the AL scholars preferred difficult, likely, better, and necessary for noun post-modification. Future research should reveal whether these observations might entail the existence of a discipline-specific repertoire of stance post-modifiers in the rhetorical practices of these two soft sciences.
Two recurrent syntactic patterns convey authorial stance through predicative evaluators. The first pattern is that of 'anticipatory it + copula verb + evaluative adjective + to-inf./that-clause'. This syntactic construction, in which heavier constituents are placed at the end of the clause, allows writers to thematise attitudinal meaning when expressing a value judgement towards the new findings. In both disciplines, impersonal it-patterns combine with relevance and assessment adjectives modified by to-infinitive constructions to express the writer's evaluation of research processes (it is difficult to test; it is interesting to compare; it is thus critical to develop), and with likelihood adjectives followed by that-clauses conveying the writer's evaluation of research outcomes (such as it is probable that; it is likely that; it is certain that).
The second recurring pattern is that formed by 'nominal expression + copula verb/hinge + evaluative adjective'. Subject predicatives appear both in the AL introductions (the role of language play in SLL is facilitative, though not necessary; the treatment of personality has been inconsistent) and in JASIST (research on this issue is critical; [...] that was relevant to the topic). In both disciplines, the predicative adjective is often preceded by adverbs qualifying the writer's value judgement ([...] are particularly interesting to study; these factors are equally important). At other times, the adjective is complemented by a to-infinitive when the writers evaluate research processes (these theoretical models are difficult to quantify; very hard to publish; very difficult to find). It is worth noting that the intensifying adverb very recurs in process-oriented evaluation and that it stands as the most common stance adverbial in the AL introductions.
In sum, the syntactic description of stance adjectives suggests that both attributive and predicative adjectives should not be regarded in isolation but rather as building up a specific lexico-grammar of evaluation in the introductory sections of research articles. Future comparative studies across these soft-applied sciences should confirm whether the linguistic preferences in terms of syntactic positions, recurring lexical choices and collocational units of stance adjectives commented above are significant in a larger corpus.
4.3. Functional variation of stance adjectives in introductions
On both theoretical and pedagogical grounds, Swales (1990, 2004) and Swales and Feak (1994) represent a major contribution to the consolidation of information conventions in research article sections. These authors establish three main rhetorical moves in the so called 'CARS (Create-a-Research-Space) model' established for writing introductions (Swales and Feak 1994: 174-75): Move 1 sets up the research territory, Move 2 establishes a niche or gap in previous research and Move 3 occupies the niche or gap. As such, the rationale of this information organisation convention brings to the fore the need for persuasion for reasons of "academic promotionalism and boosterism" (Swales 2004: 226). Academic writers rely on stance adjectives either for highlighting the new knowledge or for assessing research processes. By this means, they create a research space for the purpose of convincing the expert audience of the significance of the findings reported in the paper. Furthering the semantic taxonomy of evaluative adjectives described in section 4.1, Table 3 shows the occurrence of relevance and assessment adjectives across each CARS move as well as the total frequency per move of each type of adjectives in the corpus and in the two subcorpora.
Total frequencies per move (cf. right-hand column of Table 3) reveal a higher occurrence of relevance adjectives in Moves 1 and 3 but a slightly higher occurrence of assessment adjectives in Move 2. This distribution of adjectives seems to correlate with the rhetorical intentions of each CARS move: claiming centrality of the topic that the research article discusses (Move 1), conveying authorial stance by showing that the author is knowledgeable about the state of the matter (Move 2), and highlighting the new findings and promoting both research claims and the researcher him/herself (Move 3).
As far as the two disciplines are concerned, the AL texts show the highest frequencies of stance adjectives, both relevance and assessment, in Move 2. In the comparison of the two types of adjectives across moves, relevance evaluators are slightly more common in Moves 2 and 3 and score similar to assessment adjectives in Move 1. In the JASIST texts, the highest frequencies of stance adjectives appear in Move 3, followed by Move 1. Across moves, relevance adjectives outnumber assessment qualifiers in Moves 1 and 3, while assessment evaluators score double than relevance adjectives in Move 2. Using rhetorical analysis, the following subsection describes the functional work of stance adjectives across CARS moves in each discipline and gives possible explanations for discipline-specific rhetorical conventions in the use of these adjectives.
4.3.1. Functional role of stance adjectives in the AL texts
In terms of rhetorical organisation, the AL introductions first provide the general theoretical background of the topic, and then present the literature review to finally lead readers to a specific reference to the research presented in the paper. These three sections would correspond, broadly speaking, to the three moves of the CARS framework, but none the less seem to represent less clear-cut moves than those found in the JASIST introductions. The following context-sensitive observations were drawn regarding the functional variability of relevance and assessment adjectives in the AL subcorpus.
In Move 1, the AL writers provide an introductory sentence that serves to establish the 'research territory' (i.e. sharing background knowledge with their readers). Seven out of the ten introductions claim centrality of the topic by means of relevance adjectives, both in attributive and predicative positions:
(1) The Noticing Hypothesis has been a driving force in advancing research on explicit vs. implicit learning (AL/5)
(2) Language play in general has long been recognised as important for the development and learning [...] (AL/8)
In Move 2 writers summarise previous research and anticipate present research by extending a finding. Nine out of the ten studies are a continuation of an existing line of enquiry, which might explain why Move 2 is rich in assessment and relevance adjectives. Through stance adjectives, both functioning as subject predicatives and nominal pre-modifiers, the AL writers assess the value and usefulness of previous research when referring to previous research knowledge on the specific topic of the paper:
(3) Intuitively, this approach to evaluating texts seems to be very satisfactory. (AL/2)
(4) The intrinsic and extrinsic orientations of motivation, as conceptualized in self-determination theory, constitute a useful framework for studying motivation in educational contexts. (AL/4)
(5) Also, it has been documented that private language rehearsal, that is private language play is an important feature of learners' acquisition. (AL/7)
One of the AL writers uses stance adjectives profusely in Move 2. By explicitly acknowledging the difficulty of controlling several research processes, the writer seems to anticipate to readers that what is being reported in the paper intends to cover a research gap (literally, a widely acknowledged problem). In any case, the abundance of evaluative adjectives in this particular extract is more likely to be attributed to the individual style of the writer when projecting his opinion in the text:
(6) It is difficult to get texts [...]. It is difficult to see how the occurrence [...] it is however very difficult to test claims [...]. It is very difficult to set up large groups [...] and it is very difficult to find an example [...]. Again, this is not surprising. A widely acknowledged problem in experimental research is that studies which fail to show significant results are very hard to publish, and for this reason, studies where LFP appears not to be performing successfully are much less likely to be published than studies where it has been used successfully. (AL/2)
In the literature review section of the introductions both relevance and assessment adjectives help the AL writers maintain the argumentative flow. By anticipating for readers the shift from Move 2 to Move 3, evaluative adjectives not only act "to present information through the organization of the text itself (on the autonomous plane)" but also "to engage readers as to how they should understand it (on the interactive plane)" (Hyland 2005: 8). The examples below, for instance, have a recapitulating/summarising function which helps readers make a connection between the problem and the solution that the writer is going to present. At the same time, through the use of adjectives the writers seek to draw their readers towards a similar line of thought:
(7) The intrinsic and extrinsic orientations of motivation [...] constitute a useful framework for studying motivation in educational contexts. [...]. The major focus of this study is to examine the relationship of motivation to actual learning processes. (AL/4)
(8) On a theoretical note, language play can thus be seen as an important element in language learning. There is, however, very little empirical work on [...]. The present study attempts to situate children's spontaneous language play and L2 learning [...] (AL/7)
In Move 3, the AL writers occupy the research niche by giving the purpose of the present research and describing it. Four out of the ten writers explicitly use relevance and assessment adjectives in this move to validate the new knowledge ([t]he major focus of this study is to examine [...], [o]ur starting point is that it is axiomatic that [...]). The remaining six writers rather seem to avoid evaluative remarks and opt for 'voiceless' statements in which stance is kept to a minimum:
(9) Since it is theoretically plausible that effective use of learning strategies may sustain motivation in language learning, this study will explore [...] (AL/4)
(10) The purpose of this paper is to explore [...] and I focus further on one specific aspect of [...] (AL/6)
This authorial positioning is often mitigated by a hedged discourse which acts as a face-saving strategy when claiming centrality of findings. As in the examples below, combinations of impersonal anticipatory-it patterns with hedges such as conditionals, modals and epistemic lexical verbs might suggest that their writing conventions recommend toning down authorial remarks when stating the purpose of the research in the paper in Move 3. The use of tentative language reveals a very different interpersonal positioning to that found in the JASIST writers (see subsection 4.3.2):
(11) In view of the growing importance of this kind of research, it would be advisable to pursue the possible constraints of ID variables on the processing of L1 pragmatic input. As a follow-up study to Takahashi (2001), the current investigation tends to examine [...] (AL/5)
(12) More specifically, in this paper, I hope to show how advice resisting is accomplished in and through talk. (AL/6)
(13) Since this was our first investigation of E&M instruction of vocabulary, it seemed prudent that we should focus on unitary words of apparent high tractability. (AL/10)
In all, both assessment and relevance adjectives, being most common in Move 2 of the AL texts, function to convey a critical stance and foreground the new research outcomes. Rather than making negative comments about previous research, the AL writers use positive adjectives to evaluate these studies. Accordingly, the main rhetorical function of stance adjectives does not seem to be that of creating a research gap but of validating previous research to offer the continuation of a line of enquiry. It is also worth noting that in this particular discipline the expression of stance through relevance adjectives in Move 3 is often mitigated with hedging devices. As stated above, the combination of hedging devices with attitudinal markers could be indicative of a particular writing convention in the discipline of applied linguistics--namely developing a critical voice yet conveying pragmatic politeness towards the expert readership.
4.3.2. Functional role of stance adjectives in the JASIST texts
As opposed to the AL texts, the JASIST introductions prove to be very systematic in terms of rhetorical organisation as all of them stick very clearly to the three main moves of the CARS model. In a closer contextual analysis, the functional distribution of adjectives across these moves displays the following characteristics.
In Move 1, information scientists mostly have recourse to relevance adjectives (central, relevant, complex, critical, integral) in order to establish the research territory. This rhetorical preference helps writers give emphasis to their claims and provide centrality to the theoretical frame of reference. The ten introductions use an evaluative opening statement containing a relevance adjective that foregrounds the significance of the research-related entity being judged by the writer. See, for instance, the following examples:
(14) Identification of relevant items in large information collections has been a problem of central interest in the information systems area. (JASIST/6)
(15) The spread of end-user database applications has raised a critical issue of how to [...] (JASIST/7)
(16) With society's increasing dependence on Internet search engines, it becomes important to understand how users utilize these sites [...] (JASIST/9)
As also happens with the AL subcorpus, Move 2 in the JASIST introductions links what has been researched on (Move 1) to what is going to be researched on (Move 3), and also serves to invite readers to share the author's critical view towards previous research. While the AL writers relied on the literature to offer a continuation of previous research, the JASIST scholars rely on the literature to indicate a research gap. This gap is indicated by attributive adjectives conveying negative assessment and making explicit the inconsistency of previous research findings (serious problems, unacceptable computational complexity, poor fault tolerance, etc.); to a lesser extent, predicative adjectives are also used for similar purposes. Adhering to Thetela's (1997) definition of research-oriented evaluation (ROE), the adjectives italicised below specifically evaluate research-process entities:
(17) However, the treatment of personality has been inconsistent especially within the communication sciences (JASIST/3)
(18) However, such an assumption may not be valid. [...] In fact, over the past few years, there have been many ineffective, dwindling CN services [...] (JASIST/4)
(19) [...] a systematic methodology is still not clear. (JASIST/6)
All the JASIST writers use Move 2 not to extend a previous research finding, as the AL scholars do, but to criticise problems or shortcomings in previous studies. By this means, statements in Move 2 implicitly suggest to readers that there is a need for carrying out improved research processes:
(20) Often, researchers use differing strategies, based on different conceptualizations, which result in nongeneralizable findings (JASIST/3)
(21) This gap in knowledge is increasingly problematic. (JASIST/7)
(22) Nevertheless, it has still been necessary to make such metadata more interoperable and endurable, or more reliable. (JASIST/10)
Following this problem-solution thematic development, Move 3 is mainly used to fill the research gap or niche created in Move 2. The JASIST writers mainly resort to relevance adjectives (indicative, greater, significant, imperative, etc.), at times syntactically acting as predicative evaluators in impersonal constructions and, at other times, followed by to-infinitive clauses specifying still-to-take-place research processes. These adjectives explicitly indicate the usefulness and significance of the new research process undertaken in the study. As shown in the examples below, thematic options containing subject predicative adjectives as "evaluative enhanced themes" (Tucker 1998) allow the writers to place greater emphasis on their expression of stance. In contrast to the AL scholars, notice the use of a non-hedged, categorical language when giving relevance to the new research presented in the paper:
(23) It is thus critical to develop a holistic understanding of the effectiveness of the three data models [...] (JASIST/7)
(24) Much of the existing literature in information seeking suggests that users are largely unsuccessful in their search attempts, with failure rates often approaching 50%. Clearly, this presents a significant dilemma [...]. With such a significant portion of the population searching online for information, it is imperative to understand [...] (JASIST/9)
In sum, the contextual analysis of stance adjectives in the JASIST texts suggests a clear correlation between the presence of evaluative adjectives and the rhetorical intentions of each CARS move. Relevance adjectives are recurrently used in Move 1 to give significance to the research territory. In Move 2, assessment adjectives help create the research niche. As opposed to Move 2 in the AL texts, the JASIST writers abundantly employ negative rather than positive assessment adjectives, since the purpose of their research is not to extend previous findings but to indicate a research gap. In Move 3, relevance adjectives, being much more recurrent than in the other moves, serve to highlight the value or usefulness of the new knowledge, thus echoing the academic boosterism and promotionalism that Swales (2004) notes for research article introductions. The clear-cut correlation between the two categories of stance adjectives and the three CARS moves might be interpreted as belonging to the specific rhetorical conventions and writing practices of this soft science.
Despite the neutrality and facelessness often attributed to academic writing, this small-scale study appears to confirm the importance of evaluative lexis in introductory sections of research articles, a section in which writers are expected to find a research space in order to convince their audiences of the value and usefulness of the new knowledge. The initial hypothesis that applied linguistics and information science, both of them soft-applied disciplines, would behave similarly regarding the use of stance adjectives in introduction sections is only partially confirmed by the findings of this study. Although the size of the corpus cannot facilitate empirical generalisations on stance adjectives, several observations are made below regarding the specific semantic, syntactic and rhetorical profiles of stance adjectives in each discipline.
Semantically, stance adjectives appear to stand as key elements of lexical enquiry in the research writing practices of the two disciplines as they account for almost half of the total adjectival occurrences in the corpus. Consistent with previous studies on stance adjectives (Stotesbury 2003; Swales and Burke 2003; Biber 2006), the recurrence of both relevance and assessment adjectives in the selected introductions may give evidence of the writers' overt evaluation of research outcomes, procedures and methods. This suggests that, even if style guides recommend objectivity in academic prose, both the AL and the JASIST writers tend to lay overt their stance in a very similar way.
The syntactic analysis of stance adjectives across disciplines seems to indicate the existence of a local grammar of evaluation (cf. Hunston and Sinclair 2000) in the introductions of the two disciplines. Also, in agreement with Hunston (1993) and Hyland (2005), the context-sensitive analysis further suggests that this lexico-grammar of evaluation performs a metadiscourse function in discourse. That is, this phraseology of stance not only allows writers to evaluate existing knowledge and claim centrality of new findings (on the autonomous plane of discourse), but also engages readers and convinces them that the author's claims are valid or useful (on the interactive plane of discourse). Although the two disciplines seem to share common syntactic features, this level of analysis also unveils some minor discipline-specific linguistic preferences in terms of recurring repertoires of collocational units accompanying stance adjectives.
Rhetorically, the comparison of the two disciplines reveals different discipline-specific conventions regarding the functional distribution of relevance and assessment adjectives across the three CARS moves of introductions. The AL writers mostly restrict their adjectival choices to items conveying positive evaluation with the purpose of extending previous research in Move 2. Conversely, the JASIST writers make wider use of stance adjectives both to seek centrality of findings (in Moves 1 and 3) and to convey negative assessment--and, by this means, indicate a research gap--in Move 2. A further difference noted in the rhetorical analysis is that while the AL writers mostly express their opinion using a cautious, hedged and tentative style in Move 3, the JASIST writers lay overt their stance when giving significance to the claims of new knowledge. As stated above, although the small-scale corpus is limited for substantive comparison, these findings may also be taken to indicate that this functional variability in the use of stance adjectives should be interpreted as a rhetorical response to the publishers' actual requests. Whereas JASIST information for authors solicits "high quality innovative papers addressing significant research questions" (emphasis added), Applied Linguistics "is more interested in the handling of problems in a principled way by reference to theoretical studies" (emphasis added).
From a broader perspective, and in the light of the above analysis, the rhetorical differences in the use of adjectives across these soft sciences might also be attributed to the specific nature of knowledge, lines of reasoning, enquiry procedures and specific conventions in the transmission of knowledge in each disciplinary grouping. Becher and Trowler claim that "the ways in which academics engage with their subject matter and the narratives they develop about this are important structural factors in the formulation of disciplinary cultures" (2001: 23). As illustrated with the cross-disciplinary comparison of soft scientists' specific choices, functions and rhetorical intentions through the use of stance adjectives, there appears to be a need for redefining the specific writing conventions of each disciplinary grouping according to the nature of knowledge and the ethos of each academic sub-community--hence the linguistic richness of research prose.
Received 1 June 2007
Revised version received 8 January 2008
Barras, Robert 1987: Scientists Must Write. A Guide to Better Writing for Scientists, Engineers and Students. London and New York: Chapman and Hall.
Becher, Tony and Paul R. Trowler 2001: Academic Tribes and Territories. Intellectual Enquiry and the Culture of Disciplines. Buckingham and Philadelphia: The Society for Research into Higher Education and Open UP.
Biber, Douglas 2006: 'Stance in Spoken and Written University Registers'. Journal of English for Academic Purposes 5: 97-116.
Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad and Edward Finegan 1999: Longman Grammar of Spoken and Written English. Essex: Longman.
Breivega, Kjersti, Trine Dalh and Kjersti Flottum 2002: 'Traces of Self and Others in Research Articles: a Comparative Pilot study of English, French and Norwegian Research Articles in Medicine, Economics and Linguistics'. Applied Linguistics, 12.2: 218-39.
Conrad, Susan and Douglas Biber 2000: 'Adverbial Marking of Stance in Speech and Writing'. Susan Hunston and Geoff Thompson, eds. Evaluation in Text: Authorial Stance and the Construction of Discourse. New York: Oxford UP. 56-73.
Dahl, Trine 2004: 'Textual Metadiscourse in Research Articles: A Marker of National Culture or of Academic Discipline? Journal of Pragmatics 36: 1807-25.
Grabe, William and Robert B. Kaplan. 1997: 'On the Writing of Science and the Science of Writing: Hedging in Scientific Text and Elsewhere'. Raija Markkanen and Hartmut Schroder, eds. Hedging and Discourse. Approaches to the Analysis of a Pragmatic Phenomenon in Academic Texts. Berlin and New York: Mouton de Gruyter. 151-67.
Hunston, Susan 1993: 'Evaluation and Ideology in Scientific Writing'. Mohsen Ghadessy, ed. Register Analysis: Theory and Practice. London: Pinter. 5-7.
Hunston, Susan and Geoff Thompson, eds. 2000: Evaluation in Text: Authorial Stance and the Construction of Discourse. New York: Oxford UP.
Hunston, Susan and John Sinclair 2000: 'A Local Grammar of Evaluation'. Susan Hunston and Geoff Thompson, eds. Evaluation in Text. Oxford: Oxford UP. 74-101.
Hyland, Ken 1998: 'Persuasion and Context. The Pragmatics of Academic Metadiscourse'. Journal of Pragmatics 30: 437-55.
--2002: 'Options of Identity in Academic Writing'. ELT Journal 56.4: 351-58.
--2005: Metadiscourse. London and New York: Continuum.
Mauranen, Anna 1993: Cultural Differences in Academic Rhetoric: A Textlinguistic Approach. Frankfurt: Peter Lang.
Markkanen, Raija and Hartmut Schroder, eds. 1997: Hedging and Discourse. Approaches to the Analysis of a Pragmatic Phenomenon in Academic Texts. Berlin and New York: Mouton de Gruyter.
Meyer, Paul G. 1997: 'Hedging Strategies in Written Academic Discourse: Strengthening the Argument by Weakening the Claim'. Raija Markkanen and Hartmut Shcroder, eds. Hedging and Discourse. Approaches to the Analysis of a Pragmatic Phenomenon in Academic Texts. Berlin and New York: Walter de Gruyter. 21-41.
Perez-Llantada, Carmen and Claus-Peter Neumann. 2006: 'A Genre-based Study of Authorial Presence in Research Articles and University Lectures'. Paper presented at the Conference in Honour of J.M. Swales. Ann Arbor, Michigan, June 2006.
Salager-Meyer, Francoise 2005: 'The Expression of Conflicting Knowledge Claims in Academic Discourse: A Study in French/English Contrastive Rhetoric'. Pilar Duran, Guadalupe Aguado and Ana M. Roldan, eds. Reflections on Language Use in the Academic Context. Madrid: DLACT. 29-48.
Shaw, Philip 2003: 'Evaluation and Promotion across Languages'. Journal of English for Academic Purposes 2: 343-57.
Stotesbury, Hilkka 2003: 'Evaluation in Research Article Abstracts in the Narrative and Hard Sciences'. Journal of English for Academic Purposes 2: 327-41.
Swales, John M. 1990: Genre Analysis: English in Academic and Research Settings. Cambridge: Cambridge UP.
--2004: Research Genres: Exploration and Applications. Cambridge: Cambridge UP.
John M. Swales and Amy Burke. 2003. '"It's really fascinating work": Differences in Evaluative Adjectives across Academic Registers'. In Pepi Leistyna and Charles F. Meyer eds. Corpus Analysis. Language Structure and Language Use. Amsterdam, New York: Rodopi. pp.: 1-18.
Swales, John M. and Christine Feak 1994: Academic Writing for Graduate Students. Ann Arbor: the U of Michigan P.
Thetela, Puleng 1997: 'Evaluated Entities and Parameters of Value in Academic Research Articles'. English for Specific Purposes 16.2: 101-18.
Tucker, Gordon H. 1998: The Lexicogrammar of Adjectives. A Systemic Functional Approach to Lexis. London and New York: Cassell.
Turk, Christopher and John Kirkman. 1996 (1982): Effective Writing. Improving Scientific, Technical and Business Communication. London and New York: Spon P.
Varttala, Teppo 2003: 'Hedging in Scientific Research Articles: A Cross-disciplinary Study'. Giuseppina Cortese and Philip Riley, eds. Domain-specific English: Textual Practices across Communities and Classrooms. New York: Peter Lang. 141-74.
Vassileva, Irena 2001: 'Commitment and Detachment in English and Bulgarian Academic Writing'. English for Specific Purposes 20: 83-102.
Weissberg, Robert and Suzanne Buker 1990: Writing up Research. Experimental Research Report Writing for Students of English. New Jersey: Prentice-Hall.
Wilkinson, Antoinette 1991: The Scientist's Handbook for Writing Papers and Dissertations. New Jersey: Prentice Hall.
Ma Carmen Perez-Llantada Auria
University of Zaragoza
(1) Becher and Trowler (2001: 14) distinguish between hard sciences (i.e. those belonging to the scientific, technological and biomedical fields) and soft sciences (i.e. those belonging to the social sciences, education, humanities and arts fields). As these authors contend, this distinction is grounded in the fact that disciplinary communities of scientists use different "methodological approaches, conceptual and theoretical frameworks" depending on the specific nature of the knowledge. Applied linguistics and information science, the two disciplines analysed in this paper, fall into the category of the soft sciences (cf. also UNESCO international nomenclature for the classification of these humanities disciplines as 5701- Applied Linguistics).
(2) This paper is a contribution to projects 'El ingles y la ciencia: la voz del autor en la expresion y difusion del conocimiento cientifico' (Spanish Ministry of Education, HUM2005-03646) and 'Interpersonalidad en el lenguaje academico escrito' (Government of Aragon, 245-100).
(3) It might be thought, at first sight, that the statistical significance of the differences in distribution could be assessed by means of the chi-square test. However, this test is appropriate only when the data points, both between cells in the contingency table and within them, are independent of each other. Although for the data reported here there is independence between cells, the data points represented in each cell are not independent, since many observations are derived from each text, and some texts might well contribute far more instances to a given cell than others, the effect of this being exaggerated by the small sample used in this pilot study. For this reason, the differences in distribution have not been tested statistically. Here, I would like to acknowledge the valuable guidance provided by the Atlantis specialist in the statistical procedures of this paper. I also wish to thank the anonymous reader who gave generous help in improving the precision of expression in the text.
Table 1: Distribution of semantic categories of adjectives in AL and JASIST introductions Descriptors Descriptors Evaluators AL 26 182 3 22.65 0 51.12% JASIST 31 198 4.05 25.89 6.64% 42.40% Classifiers Relational c. Topical c. AL 115 33 14.31 4.11 32.30% 9.28% JASIST 160 78 20.92 10.2 34.26% 16.70% Totals AL 356 44.31 100% JASIST 467 61.05 100% Table 2: Distribution of relevance and assessment stance adjectives across disciplines Relevance adjectives AL JASIST Raw number of tokens 97 112 Frequency per 1,000 words 12 14.64 Total % in subcorpus 1 56.57% Total tokens 209 Frequency per 1,000 words 13.33 Total % 55.00% Assessment adjectives AL JASIST Raw number of tokens 85 86 Frequency per 1,000 words 10.70 11.24 Total % in subcorpus 0.47 43.43% Total tokens 171 Frequency per 1,000 words 10.9 Total % 45.00% Table 3: Relevance (R) and assessment (A) adjectives across CARS moves AL R A Move 1 25 25 3.11 3.11 50.00% 50.00% Move 2 40 36 4.98 4.48 52.63% 47.37% Move 3 32 24 3.98 2.99 57.14% 42.86% JASIST R A Move 1 39 21 5.1 2.75 65.00% 35.00% Move 2 10 22 1.31 2.88 31.25% 68.75% Move 3 63 43 8.24 5.62 59.43% 40.57% Total frequencies R A Move 1 64 46 4.08 2.93 58.18% 41.82% Move 2 50 58 3.19 3.69 46.30% 53.70% Move 3 95 67 6.06 4.27 58.64% 41.36%
|Printer friendly Cite/link Email Feedback|
|Author:||Perez-Llantada Auria, Ma Carmen|
|Publication:||Atlantis, revista de la Asociacion Espanola de Estudios Anglo-Norteamericanos|
|Date:||Jun 1, 2008|
|Previous Article:||'Making something out of nothing': lesbianism as liberating fantasy in the children's hour.|
|Next Article:||Ann Thompson and Neil Taylor, eds. 2006: Hamlet. The Arden Shakespeare.|