ANALYSING STANCE IN AMERICAN AND SPANISH BUSINESS MANAGEMENT RAS: THE CASE OF SENTENCE-INITIAL ‘RETROSPECTIVE LABELS’

When writing research articles (RAs), scholars can use certain lexico-grammatical traits that enable them to encode their attitudes, judgments and opinions, thus functioning as markers of stance. It is believed that sentenceinitial nouns preceded by a deictic –‘retrospective labels’ in Francis’s terminology (1986, 1994)– can be considered one of those traits. The aim here is to explore whether there are any language-driven differences in the use of ‘retrospective labels’ as markers of stance within a particular disciplinary discourse, namely, Business Management. ‘Retrospective labels’ were analysed in a corpus of 12 RAs on the above-mentioned discipline, 6 in American English and 6 in Spanish. The focus is placed on the contrastive analysis of the frequency of use of these ‘retrospective labels’, the type of head nouns and modifiers which most frequently form part of them and the extent to which these ‘retrospective labels’ convey attitudinal meaning. As a general implication, it is believed that the differences drawn from analyses of this type should be borne in mind by Spanish Business Management scholars when writing their RAs in English. 137 Journal of English Studies, vol. 4 (2003-2004), 137-154 1. A preliminary version of this piece of research was presented at the Conference “Discourse, Ideology and Ethics in Specialized Communication” held at the Università degli Studi di Milano (Italy) from 11th to 13th November 2004. What I present here is the result of a thorough revision of the paper presented there. This study was carried out within the framework of the project entitled “El inglés y el conocimiento científico: análisis pragmático-cognitivo de la metáfora gramatical y su utilización en la creación, expresión y divulgación del conocimiento científico” (Universidad de Zaragoza V.I. Apoyo B 2004 245-100). I am indebted to the Ministerio de Educación, Cultura y Deporte as this research was undertaken while holding a scholarship from that Ministry.

1.A preliminary version of this piece of research was presented at the Conference "Discourse, Ideology and Ethics in Specialized Communication" held at the Università degli Studi di Milano (Italy) from 11th to 13th November 2004.What I present here is the result of a thorough revision of the paper presented there.This study was carried out within the framework of the project entitled "El inglés y el conocimiento científico: análisis pragmático-cognitivo de la metáfora gramatical y su utilización en la creación, expresión y divulgación del conocimiento científico" (Universidad de Zaragoza V.I.Apoyo B 2004 245-100).
I am indebted to the Ministerio de Educación, Cultura y Deporte as this research was undertaken while holding a scholarship from that Ministry.

INTRODUCTION
It is generally assumed that there are a number of lexico-grammatical traits in research articles (RAs) that reveal the presence of a writer in the text and through which writers establish a particular interactive relationship with their readership, conveying their attitudes, judgments and opinions.Those traits that can be considered markers of 'writer's stance' (Biber et al. 1998a;Hyland 1999) or 'evaluation' (Hunston 1993(Hunston , 1994(Hunston , 2000) ) 2 perform, then, a significant interpersonal function, signalling the writer's attitude and contributing to the construction of a convincing argument.It is believed that the use made of markers of stance or evaluation can be dependent both on the discipline of the RA and on the language in which it is written.
It is believed that those nouns which encapsulate a preceding argument and are accompanied by a deictic -'retrospective labels' (Francis 1994)  3 -might act as significant stance markers.'Retrospective labels' perform an outstanding textual, cohesive function.Through these labels writers re-take previously stated information and present it as given in thematic position, which enables them to build on it and, therefore, to continue developing their argument.Besides this discoursal function, 'retrospective labels' might also perform an interpersonal function, (i) indicating the readers how to interpret the preceding piece of information, and/or (ii) encoding the writer's attitude towards what has been previously stated.Through 'retrospective labels' authors might try to ensure that their texts are understood and interpreted as intended.In the following examples the 'retrospective label' clearly signals the writer's interpretation of the previous stretch of discourse and indicates the reader how it is to be understood.

Example 1
We also restricted our sample to late movers that were clearly able to make an entry into the U.S. market.As suggested by Robinson et al. (1994), we distinguished firms that were actually able to make an entry into the market from those who attempted but failed in their attempt to enter.This definition of an entrant […].(SMJ2) 4 .

Example 2
Makadock (2001) notes that picking and deploying resources should not be viewed as alternative, but rather as complementary perspectives for achieving desired strategic outcomes.How do these two approaches interact in the context of acquisitions?(JM1).

. (AD1).
This interpersonal potential of 'retrospective labels' cannot be disregarded if the reader wants to fully ascertain the writer's argumentation.
As stated above, it is here believed that not only the discipline but also the language in which RAs are written and, more broadly, the culture in which each set of articles is produced and distributed, might condition the use made of stance markers, including 'retrospective labels', taking an 'intercultural rhetoric' approach (Connor 2004a(Connor , 2004b)).In this sense, a contrastive analysis of these linguistic units in RAs from a single discipline will be here proposed.Sentenceinitial 'this/these/such+Noun' constructions will be analysed in a corpus of Business Management RAs originally written in English and in Spanish.The aim here is to adopt a contrastive point of view and analyse: (1) the frequency of use of these 'retrospective labels' in both languages, (2) the types of head nouns more frequently used in each language, (3) the kind of modifiers more frequently included as part of the 'retrospective labels' in both sets of articles, and (4) the extent to which these 'retrospective labels' convey an attitudinal meaning.

CORPUS AND METHODOLOGY
The analysis of 'retrospective labels' will be carried out in a comparable corpus, which is in fact formed by two sub-corpora: (1) a corpus composed of 6 RAs on Business Management written in American English and (2) a corpus composed of 6 RAs on Business Management written in Spanish. 5The number of words and sentences of each sub-corpus is presented in the The same criteria have been followed for the compilation of both sub-corpora so that they lend themselves to a contrastive analysis.Accordingly, both the English and the Spanish sub-corpora are composed of RAs written by native speakers, published during the years 2003 and 2004 and covering a similar domain or field, namely, applied studies to Business Management.As Corpas Pastor (2001: 158) states, a comparable corpus is "aquel corpus que, en relación a otro u otros corpus de lenguas distintas, incluyen tipos similares de textos originales.Al haber sido compilados de acuerdo con los mismos criterios de diseño, dichos corpus permiten la equiparación lingüística de sus elementos integrantes".
In the English sub-corpus all 'this/these/such + Noun' constructions in thematic position and in the Spanish sub-corpus all the instances of sentence-initial 'este/esta/ estos/estas/tal/tales/dichos/dichas + Noun' constructions were searched for.Their context was then analysed to decide whether they could be considered tokens of 'retrospective labels' or not.Throughout this selection process the two criteria stated by Francis (1986: 3-4) in order for a noun to be considered an "A-noun" were very much taken into account: First, it must be functioning as a pro-form and as such be an anaphorically cohesive device, referring metadiscursively to a stretch of discourse preceding it in terms of how the writer chooses to label or interpret the latter for the purposes of his/her argument.Second, it must also face forwards: it must be presented as the given information in terms of which the new propositional content of the clause or sentence in which it occurs is formulated.
In this way, some examples such as 'this paper', 'this study', 'this research', 'este trabajo', 'este estudio' or 'esta investigación' that refer to the authors' current research and not to a previous stretch of discourse were dismissed, as well as those instances that did not project the argument onwards.Francis (1986: 31) further states that A-nouns are new lexical items.Therefore examples such as the following ones in which the head noun of the 'retrospective label' is a verbatim repetition of a previous lexical item were disregarded: Una de las principales razones por las que las empresas temen externalizar alguna de las funciones de su sistema de información es la incertidumbre que esta decisión conlleva.Esta incertidumbre [...].(AD2) Those instances of 'retrospective labels' that are preceded by an Adjunct (e.g.'However, these findings' (SMJ1), 'En concreto esta falta de significatividad' (DyO1)) were counted as tokens since they still link a previous stretch of discourse with a subsequent one and, together with the Adjunct, they can be analysed as 'multiple Themes' (Halliday 1994: 52-54, Thompson 1996: 133-138).Those 'retrospective labels' that form part of the Theme but which do not constitute its head were also counted as tokens (e.g.'The development of such a contingency perspective' (SMJ2), 'One interpretation of this finding' (JM1) 'Los orígenes de esta idea' (DyO2)).Finally, those labels that had the same referent were only counted once.

TOKENS OF 'RETROSPECTIVE LABELS'
The total number of 'retrospective labels' found in both sub-corpora is presented in Table 2.In order for the counts to be comparable, they have been normalised 6 per 100 sentences in 6. "'Normalization' is a way to adjust raw frequency counts from texts from different lengths so that they can be compared accurately" (Biber et al. 1998b: 263).A normalisation per 100 sentences seems to be the most suitable procedure to ensure an accurate comparison because, first, the linguistic phenomenon analysed is sentence-initial, therefore there cannot be more than one instance per sentence and, second, the number of sentences that compose the English sub-corpus is much greater than the number of sentences in the Spanish sub-corpus.
The total counts, as well as most partial counts, show that the use of 'retrospective labels' is slightly greater in the English sub-corpus than in the Spanish one.American scholars pack a preceding piece of discourse and present it as the Theme of the following sentence more frequently than Spanish scholars.In using more 'retrospective labels' American scholars can be seen to proceed more linearly, contributing to guiding their readers in their decoding process, following a 'writeroriented style' (Kaplan 1966(Kaplan , 1987(Kaplan , 1988)).Spanish scholars, however, make less use of 'retrospective labels', favouring a more 'reader-responsible style', that is, contributing less prominently to facilitating the reading process.
As Francis (1986: 36) points out "by encapsulating previous discourse in this way, A-nouns function interactively as organisational signals, providing the reader with signposts to guide him/her through the discourse".Spanish scholars, then, can be seen to provide fewer 'organisational signals', which might result in a greater decoding effort on the part of readers who have to deduce the relationship between different parts of the discourse.It can be inferred from this that when using a 'retrospective label' writers implicitly state their views and interpretations on something previously stated and, thus, they establish a particular relationship with their readers.
The difference in frequency of use of 'retrospective labels' could be partly explained in terms of different rhetorical preferences in the two languages and cultures to which the two sets of RAs belong.In her comprehensive research of premise-conclusion retrospective labels in economics and business RAs, which included the analysis of fuzzy (i.e.This), implicit (i.e.conjuncts) and explicit (i.e.deictic + noun in all positions, not just sentence-initial) retrospective labels, Moreno (2004) also found that Spanish scholars tend to make a scarcer use of explicit retrospective labels, which are the ones that concern us here.

TYPES OF HEAD NOUNS IN 'RETROSPECTIVE LABELS'
As Francis (1994: 88) states "any noun can be the head noun of a label if it is unspecific and requires lexical realization in its immediate context […]".That is, any noun can potentially become the head of a 'retrospective label', whether it is realised as such or not depends on its textual context.All types of head nouns in 'retrospective labels' appearing in the corpus have been listed and arranged according to their frequency in Appendix 2. Particular attention will be paid here to the distinction between metalinguistic and non-metalinguistic nouns 7 acting as heads in 'retrospective labels'.
In line with Charles (2003) findings, most head nouns, in both sub-corpora, appeared just once in the corpus.In her disciplinary contrastive research on the use of 'retrospective labels' in theses, Charles (2003) found that metalinguistic head nouns appeared twice as much in Politics as in Materials theses, that is, they were more frequently used in the discourse of a social science discipline than in the discourse of a pure science discipline.In the corpus analysed here, however, it has been found that metalanguage head nouns are very scarce, even though the RAs that compose the corpus also belong to a social science.Only two examples of metalanguage head nouns (This definition and These claims) have been found in the English sub-corpus and only one example has been found in the Spanish sub-corpus (Esta afirmación).Charles (2003) offers two reasons for the greater use of metalanguage nouns as heads of 'retrospective labels' in the discipline of Politics: its text-based nature and the fact that knowledge is reiterative in this discipline, which enourages scholars to take up a position in relation to others' works.In the RAs analysed, however, knowledge seems to be constructed cumulatively, which is more common of natural sciences as Charles (2003) claims.Although Business Management scholars revise the previous literature on the discipline and take up positions in relation to it, they contribute to the discipline by developing and testing new hypotheses and devising enlightening case studies.It can be said that whereas Politics can be considered to fall under the pure social sciences disciplinary sub-grouping put forward by Becher and Trowler (2001: 36), Business Management can be considered to fall under the applied social science sub-grouping.This shows the importance of carrying out discipline-specific analysis, since linguistic and rhetorical differences can also be found even when comparing closely linked disciplines.
The similar outcome in both the English and the Spanish sub-corpora in relation to the use of metalinguistic nouns as heads of 'retrospective labels' can lead us to conclude that the type of 'retrospective label' used might be dependent on the discipline being analysed, whereas the frequency of their use might be more dependent on the language and cultural context in which the research is reported.
It has been observed that in both corpora the head noun of 'retrospective labels' is, very frequently, a nominalisation.

English sub-corpus
Spanish sub-corpus Nominalised head nouns 23 (46.93%) 10 (45.45%)Out of the 49 different head nouns in the English sub-corpus, 23 types (46.93%) are nominalisations.In the case of the Spanish sub-corpus, 10 out of a total of 22 are also nominalisations, amounting to 45.45%.These nominalisations can be considered instances of what Halliday (1994) calls 'grammatical metaphor', i.e. incongruent realisations of qualities and processes as entities.These grammatical metaphors facilitate the accruement of lexical information in the 'retrospective label', "allowing a lot of information to be presented as a single given package" (Francis 1994: 96).The following examples illustrate the lexically dense stretches of language grammatical metaphors as heads of 'retrospecive labels' give way to: The accruement of lexical items in 'retrospective labels' will be further explored in the next section.

MODIFIERS IN 'RETROSPECTIVE LABELS'
As Francis (1994: 95-96) states, the modifiers that accompany 'retrospective labels' can perform an ideational function: "they add to the meaning of the head noun by classifying it or defining it, making its participant role more explicit", a textual function: "textual modifiers contribute directly to the organizational role of labels" or, most commonly, an interpersonal function: "they evaluate the propositions they encapsulate".Only modifying adjectival and nominal groups were looked into.In the English sub-corpus, out of a total of 66, 18 'retrospective labels' contain a modifying adjective or noun, whereas in the Spanish sub-corpus 7 'retrospective labels' out of a total of 25 have been found to contain a modifier of this type.Here are some examples of such modified 'retrospective labels':

Example 12
Makadock (2001) notes that picking and deploying resources should not be viewed as alternative, but rather as complementary perspectives for achieving desired strategic outcomes.How do these two approaches interact in the context of acquisitions?(JM1)

Example 13
If employees adopt a highly generous pattern of favor exchange, they may accrue social status, but their productivity may also decline.Conversely, if employees obtain reciprocation for favors they are owed, it may improve their productivity, but decrease their social status.This apparent paradox [.The percentage of 'retrospective labels' containing a modifier is very similar in both sub-corpora (27.27% in the case of the English sub-corpus and 28% in the case of the Spanish sub-corpus).Ideational modifiers are the most common type in both sub-corpora.Although Moreno (2004) looked into different types of modifiers, not only adjectival and nominal ones, she also found ideational modifiers to outnumber interpersonal and textual ones both in her English and in her Spanish corpora.
Against Francis (1994: 96) statement that modifiers in 'retrospective labels' most frequently perform an evaluative function, Spanish Business Management scholars do not tend to use modifiers to evaluate previous stretches of discourse or at least not to the same extent as American scholars.Moreno (2004) found no examples at all of textual or interpersonal modifiers in her Spanish corpus.In any case the scarce (or no) inclusion of interpersonal modifiers in 'retrospective labels' can be interpreted as a further signal that Spanish scholars prefer their readers to infer the writer's stance rather than making it explicit to them.

EPISTEMIC AND ATTITUDINAL 'RETROSPECTIVE LABELS'
It is believed that choosing a 'retrospective label' to package a preceding stretch of discourse could in itself be interpreted as an interpersonal action: "given that all labels are constructions of the writer, it follows that their discourse-organising function will reflect the purposes and views of the writer and hence indicate the writer's stance" (Charles 2003: 317).'Retrospective labels' as markers of stance can convey epistemic or attitudinal meanings (Biber et al. 1998a).The attitudinal or evaluative meaning of the 'retrospective label' can be conveyed either by the head noun itself (example 16) and/or by the modifiers that accompany it (examples 17 and 18).Epistemic 'retrospective labels', on the other hand, are considered to be those which "comment on the status of information in a proposition" (Biber et al. 1998a: 972) but which do not add any evaluative connotations (example 19 and examples 1-4 above).In example 19, for example, the choice of the word claims signals the status the writer attachs to the previous information; in this sense, the author's assessment of the truth value of the previous statement would have been different, had the author chosen another head noun such as, for instance, assertion or argument.This particular 'retrospective label' acts, then, as a marker of the writer's epistemic stance.

Example 17
Most of the research to date has also presumed that the success of the earlier entrant would make it difficult for subsequent entrants to make any significant inroads into the market.More recently, this presumed ability of a pioneering firm to pose hurdles for subsequent entrants is

Example 19
Some researchers argue that an acquisition may offer the opportunity to get rid of managerial deadwood (Walsh & Ellwood, 1991).While these claims may be true, the findings of this study suggest that the successful appropriation of knowledge implied by a good acquisition may require maintaining those that know the target operations best -its managers.(JM1) Biber et al. (1998a: 974)   Again the results show that Spanish scholars are less inclined to present their viewpoints and attitudes straightforward.In that sense it could be argued that American scholars establish a closer or at least more direct interpersonal relationships with their readers.The results again coincide with those of Moreno's (2004) wider analysis.She found (i) that non-evaluative labels were more frequent than evaluative ones in both languages, and (ii) that evaluative labels were more common in the English corpus than in the Spanish corpus, as reported here.

CONCLUDING REMARKS
American scholars in Business Management have been seen to make a slightly greater use of 'retrospective labels' in their RAs than Spanish scholars.The inclusion of these labels enables writers to proceed linearly in their arguments and to guide readers in their interpretation process -as well as to encode their attitudes, judgements and opinions.The greater use made of 'retrospective labels' by American scholars can be considered to be consistent with the general belief that the English language tends to favour a 'writer-responsible style', whereas Spanish is believed to favour a more 'reader-responsible style', that is, Spanish readers are more frequently left alone in their interpretation process.
It has been shown that metalinguistic nouns as heads of 'retrospective labels' are very scarce in both sub-corpora.Nominalisations account for about fifty per cent of all head nouns in both sub-corpora.These similarities in the type of head nouns being used in the English and Spanish sub-corpora can lead us to conclude that whereas the frequency of use of 'retrospective labels' might be dependent on the language/culture in which the RA is produced, the type of 'retrospective labels' used might be more dependent on the disciplinary discourse.
As far as modifiers are concerned, it has been seen that, even though Francis (1994: 96) states that "[t]he most common modifiers in labels […] are those which encode interpersonal meaning unequivocally", her statement is valid only for the English sub-corpus and not for the Spanish one.In the latter more 'ideational' than 'interpersonal' modifiers have been found.In the same line, fewer attitudinal or evaluative 'retrospective labels' have been spotted in the Spanish sub-corpus.Both findings might be seen to hint at the Spanish scholars' preference to allow their readers to infer the author's viewpoints and attitudes for themselves, which probably leads to a more distant relationship with their readers.
The contrastive analysis presented here shows that there are certain rhetorical differences between English and Spanish RAs at least in this specific discipline, which Spanish scholars need to take into account if they wish to produce successful pieces of writing when drafting their RAs in English.Moreover, the fact that differences have been found between the rhetorical practices of scholars belonging to the same disciplinary community but to two different cultural contexts means that intercultural rhetorical analyses (Connor 2004a(Connor , 2004b) ) are necessary and desirable.

Table 1 .
Description of the corpus.

Table 2 .
Tokens of 'retrospective labels' in sentence-initial position in both sub-corpora.

Table 3 .
Tokens of 'retrospective labels' in sentence-initial position normalised per 100 sentences.
point out that epistemic markers of stance outnumber attitudinal ones in the four registers they investigate, namely, conversation, fiction, news and academic prose.The results obtained in the present research are in line with their findings:

Table 6 .
The use of 'retrospective labels' as markers of attitudinal or epistemic stance.