Evaluative language in medical discourse: a contrastive study between English and Spanish university lectures

Academic spoken discourse has been a dominant issue for discourse studies researchers for the last 25 years or so. Different spoken academic genres have been analysed (Swales, 1990, 2004; Berkenkotter and Huckin, 1995; Bhatia, 2001, 2002; Mauranen, 2001; Juzwik, 2004; Crawford-Camiciottoli, 2004, 2007; among others) thanks to the compilation and the easy access to electronic spoken corpora , 1994) in higher education. Here, I analyse the use of evaluative language in medical discourse lectures. A contrastive study between Spanish and English medical lectures is carried out. To my knowledge, little attention has been paid to the analysis of evaluative language in medical discourse. The present study employs a quantitative and a qualitative approach to analyse four Spanish and English medical discourse lectures with an average of 35,000 words. The English lectures have been taken from the Michigan Corpus of Academic and Spoken English (MICASE) and the Spanish lectures have been recorded and transcribed in the Degree in Medicine course at a Spanish university for the purpose of this study. Corpus analysis tools have been used to analyse attitudinal language expressing explicit evaluation. The findings show similarities and also differences in the use of evaluative markers in academic medical discourse.


Background
Evaluation is one of the most outstanding features of interpersonal discourse, , 2004: 13). Intercultural and interpersonal discourse has been crucial to discourse analysis, since it joins language use to its social, cultural and educational contexts. Most of the research carried out up to now has centred on the use of evaluative language in academic writing in different academic texts such as research articles (Hunston, 1994;Hyland, 1998Hyland, , 1999Hyland, , 2000Hyland, , 2001Hyland, , 2004Hyland, , 2010Oakey, 2005;Okamura, 2005), essays (Barton, 1993), book reviews (Shaw, 2004;Römer, 2005Römer, , 2008, abstracts (Stotesbury, 2003;Belles-Fortuño and Querol-Julián, 2010) or text books (Poppi, 2004;Freddi, 2005), among others. To my knowledge, little research has dealt with the use of evaluative language in spoken academic discourse and even less with spoken medical lecture discourse from a cross-cultural perspective. Among studies of spoken academic corpora, and more specifically the MICASE corpus, we find Mauranen's (2001Mauranen's ( , 2002Mauranen's ( , 2003, although these studies deal with the use of evaluative language from a broad perspective, taking into account forms of implicit evaluation. In this study, I depart from Thompson and Hunston's definition of towards, viewpoint on, or feelings about entities or propositions that he or she is texts it is multi-, 2000: 177), the current study focuses on those audience-oriented attitudinal forms of explicit evaluation without neglecting previous disagreements and the lack of consensus among researchers on which lexical items are evaluative (Thetela, 1997: 102).
Recent research on evaluation in discourse discusses the complexity of evaluative phenomena, Alba-Juez and Thompson (2016: 6) argue on the through a variety of linguistic resources. They insist on the discursive process of evaluation as an important element for negotiating meaning between addresser and addressee and which . 13). Following Thompson and Alba-Juez research and expanding on the above presented definition of evaluative elements, these are seen in the current study as of the discursive context that affect the final evaluative meaning both at the : xi) Any language is related to specific cultural and institutional contexts. When analysing academic speech, we are entering a didactic environment where instruction is an essential point of departure followed by observation and feedback, which involves socialisation into the academic community (Mauranen, 2002). In this study, evaluation is seen as a powerful rhetorical tool for this academic community to convey interaction within more pedagogical genres such as the lecture in a specific discipline: Health Sciences. Some studies have approached the use of interactive features in academic genres such as the medical conference monologue (Webber, 2005); yet such genres seem to involve a far more expert-to-expert type of communication and are basically audience-oriented with little spontaneous conversation. The academic lecture, in contrast, usually involves expert-to-novice communication and questions may arise at any time during the lecture.
Non-native speakers and listeners can find it harder to process lectures in real time since they may not have the same control over the language and rhetorical tools as natives have. Contrastive linguistic studies such as the one carried out here represent a potential for pedagogical application for L2 students (English or Spanish) to cope with evaluative language patterns as well as providing insights for the field of EAP teaching and learning.

1.2
Choice of features to be analysed If we aim to explain the interaction between speaker and hearer and how evaluation is used in the communicative event (in our case the lecture), we can distinguish three different relational categories among discourse elements that can easily be detected, namely: i) relation part of discourse-part of discourse, ii) relation speaker-hearer or vice versa, and iii) relation speaker-speech. These three relations can be conveyed in many different ways: kinesics, visuals or, the most common, the use of linguistic units by means of evaluative items. The main goal of these relations is to express meanings throughout the discourse utterances. Halliday (e.g. 1994) distinguishes three functional components of meaning ideational, interpersonal and textual (or discoursal in this particular case).
If we look closely at the relational categories among the discourse elements explained above, we can see that they can somehow correspond to mentioned find corresponding relational discourse categories that can be conveyed throughout the discourse utterances. In Figure 1 we can see how the three functional meanings are distributed according to the relational categories.  Hyland (2004) distinguishes between two ways of managing interaction: Stance and Engagement. Stance intrude into the discourse to stamp their personal authority onto the arguments or step back from their , 2004: 15). On the other hand, Engagement (Hyland, 2004: 18); it is the way writers pull readers along with their argument or try to call their attention by guiding them, and this type of interaction with the audience is essential for the analysis of persuasion. This distinction is, however, not always crystal clear and there may be some overlaps.
This study focuses on those rhetorical tools that convey evaluation through signalling to the audience a salient point of the lecture, those evaluative ttitudinal interpersonal dimension and that have to do with the discourse relations speaker-hearer and/or speakerspeech, according to Figure 1 above. These are the expressions that, in written discourse, Hyland (2004) would include within the interactional concept of Engagement explained above this time applied to lecture discourse as the way speakers bring listeners into discourse. The study aims to detect and analyse usually metadiscursive phrases containing discourse deictics, that is, what Crawford-Camiciottoli (2004) has called audience-oriented relevance markers interpret as important (Crawford-Camiciottoli, 2004: 82). These expressions include lexico-grammatical patterns similar to the ones suggested by Hunston and Sinclair (2000). a) Discourse deictic + is + adjective of relevance Ex:. This is important b) Discourse deictic + is + adjective of relevance + metalinguistic noun Ex:. That is the main point Additionally, other forms of audience-oriented relevance markers that can refer not only to anaphoric but also to cataphoric elements have been taken into account for the analysis. Therefore markers such as or The point is es were also considered.
Conscious knowledge of different syntactic elements has proven to potentially aid the lecture comprehension process. Discourse markers (hereafter DMs) are part of these linguistic features that enhance lecture comprehension. Authors such as Schiffrin (1987) examined DMs in the spoken discourse of ordinary conversation and identified DMs of different nature that constructed coherence through relations between adjacent units in discourse. She observed different levels of DMs interaction: a) The semantic reality, b) the logical (epistemic) level, and c) the speech act (pragmatic) level (1987: 24). Later Fraser (1999) also attributes DMs three properties: syntactic, semantic and pragmatic. Both Shiffrin (1987) and Fraser (1999Fraser ( , 2004 agree that DMs have a core meaning and are pragmatically ambiguous. In the present study coherence is understood as an asset for the construction of discourse relations (see Figure 1) which are usually expressed by linguistic units whose meaning is strongly context dependent. According to Fraser (1999) the discourse coherence approach to the study of markers goes far beyond the analysis of linguistic entities and takes into consideration the effect on the interpretation of discourse in specific contexts.
The study takes into consideration relevance markers defined as lexicogrammatical devices that overtly mark the relative importance or relevance of points in a lecture (adapted from Deroey and Taverniers, 2012: 222). They are interactive features of lecture discourse and marker distinctions between more and less important discourse. and Taverniers, 2012: 222). According to Deroey and Taverniers (2011) relevance markers act as discourse organizers emphasising important lecture points, whichever the name they have been given in literature: evaluators (DeCarrico andNattinger, 1988), emphasizers (Siepmann, 2005), attitudinal stance bundles (Biber, 2006) or audience-oriented markers (Crawford-Camiciottoli, 2004, 2007, to name a few. The latter is the name adopted for the current study. As regards the analysis of audience-oriented relevance markers in the Spanish corpus, these can fit into what has been defined by Spanish researchers (Llorente, 1996;Portolés, 1998) as operadores discursivos ( discursive operators ), that is, markers which link communicative acts taken by discourse (1996: 14) words discursive operator is its capacity to help in the realisation of pragmatic and dis development of discourse, relate discursive acts and, in sum, are aimed at facilitating the processing of information. [my translation] 1 In previous studies on the analysis of markers in Spanish lecture corpora (Bellés-Fortuño, 2004, 2016a, I realised that trying to find Spanish marker counterparts starting out from English markers is not always the most scientific and successful procedure, but some other studies did follow this process (Fraser and Malamud-Makowski, 1996). Similar forms of markers may be found translated from English into Spanish, although these forms might not be the most common in the Spanish language and so other forms of markers can be of greater interest according to their frequency of use. Despite departing from expressions with lexico-grammatical patterns similar to the English ones, as suggested by Hunston and Sinclair (2000) and described above, a closer look at the Spanish lecture corpus would certainly give us original Spanish language patterns of audience-oriented relevance markers which may not correspond to translations from English.
The objective of this study is to analyse evaluative language features, such as audience-oriented relevance markers, from a cross-cultural perspective within a contrastive linguistics framework by comparing and contrasting medical discourse lectures from a small corpus of lectures. The aim is to detect and identify similarities and/or differences in the use of explicit evaluative forms of attitudinal language in Spanish and English academic university lectures.

2.
The study The study departs from the following research questions: a) Do explicit evaluative forms of attitudinal language differ in type, form and frequency between Spanish and English lectures in the field of Health Sciences?, b) Is language more evaluative in any of the corpora analysed? If so, how does this affect the relations between speaker-hearer at an attitudinal interpersonal level?

Methodology
Corpus linguistics techniques and discourse analysis methodologies were used for the analysis of both lecture corpora. The whole corpus (consisting of English and Spanish lectures) was searched electronically using concordancing software (AntConc 3.4.3; Anthony, 2014) to find relevant patterns of frequency of audience-oriented relevance markers starting out from the English medicine lectures and using corpus query language. The initial data were then exposed to a hand-editing process to eliminate fake expressions that did not fall into the category of audience-oriented relevance markers and to try to find other relevance audience-oriented markers contextualised in the original transcripts that the electronic search might have skipped. The analysis is thus a combination of both empirical and hermeneutic approaches.

2.2
The data The corpus under study consists of four lecture transcripts: two of them are in English and belong to the English Corpus (EC), and the other two Spanish lectures make up the Spanish Corpus (SC). The English lecture transcripts have been taken from the MICASE (Michigan Corpus of Academic Spoken English) (Simpson et al. 2002), which is available online thanks to the English Language Institute at the University of Michigan (United States) 2 . The Spanish corpus (SC) consists of two lectures recorded within the medicine degree course at a Spanish university and later transcribed for the purpose of this study. During the audiorecording of the Spanish lectures, I was present to observe the actual lecture event and take notes on contextual data.
The lecture attributes have been taken from the MICASE to make both corpora homogeneous. Therefore, the primary discourse mode, which refers to the predominant type of discourse, corresponds to the monologic type, from highly monologic in the case of the EC and monologic in the SC. As to the number of students, MICASE distinguishes between small lectures (LES), a class of 40 or fewer students, and large lectures (LEL), a lecture class with more than 40 students. In an attempt to homogenise the corpus, the lectures under study are mostly LEL, except for the Spanish LE4, which is an LES. The four lectures share the same disciplinary field, i.e. that of Health Sciences. The four lectures in the corpus were delivered to undergraduate students. The total corpus, including the EC and the SC, has a total number of 34,281 words, with an average of 8,570 words per lecture, thus satisfying the characteristics of a small corpus study (see Table 1 for further details). The whole corpus, although not extremely large in number of lectures has proven to be a valid sample for the analysis of audience-oriented relevance markers and some evaluative tokens.

2.3
Context and procedure The process of isolating audience-oriented relevance markers from the corpus has not been easy. Starting with the English lectures and the results obtained from the concordancer, the lexico-grammatical patterns were detected. These results were later hand-edited to remove instances and patterns that did not correspond to the previously established lexico-grammatical patterns and in order to avoid any confusing evaluative item indicating the lecture of a specific object, phenomenon or entity instead of being instances of evaluation signalling salient points of the lecture to the audience. This distinction, although quite easy to identify in writing, is not always unambiguous in spoken discourse. Consequently, I had to narrow the search in the corpus to the restricted lexico-grammatical patterns previously established.
Following earlier studies on the MICASE and taking into account the importance of the word thing in this corpus (Swales, 2001(Swales, , 2004 as the commonest noun, all instances of the word thing in combination with evaluative adjectives were carefully observed. It has been argued that the word thing is downtown home version of point , 2004: 35). Therefore, in the EC, instances like the important thing is, the point here is, etc., where thing or point are metalinguistic nouns, were carefully observed. In this way, evaluative adjectives as linguistic items within the audience-oriented relevance markers patterns were also studied.
Regarding the Spanish lectures, a preliminary search with the concordancer revealed some recurrent and relevant lexico-grammatical patterns similar to the English ones that allowed similar audience-oriented relevance markers to be identified and isolated. However, as this is a contrastive linguistic study, syntactic and grammatical differences between the two languages resulted in diversity as to the number of marker patterns identified and their variety. Grammatical aspects such as gender differences and number (singular or plural) in nouns and adjectives were considered as the analysis was being performed.

Results
This section presents the findings according to sequence of analysis, the relevance marker patterns are studied and compared in both corpora (EC and SC), followed by the study of evaluative adjectives.

Relevance markers in the EC
The most outstanding marker patterns found in the EC, including the number of instances and total counts per million words, can be seen in Table 2 below. The results from the analysis and observation of the relevance discourse markers in the English lectures have shown different varieties of syntactic forms that go far beyond the initial guiding patterns originally presented. In total, seven main syntactic patterns have been identified. The two outstanding marker patterns have 13 and 11 instances respectively. The most frequent relevance marker pattern corresponds to the use of relative clauses introduced by the pronoun what. A closer look at these frequently used markers revealed some collocational patterns with the particles and or so. This presents the idea of an organizational structure of spoken discourse at the attitudinal and interpersonal level in order to facilitate comprehension by the audience. The collocational patterns of discourse markers have already been identified in previous studies (Swales and Malczewski, 2001;Bellés-Fortuño, 2008) as repetitive cooccurrences that tend to form clusters with more identifiable markers. In the following examples taken from the EC, there is a slight discourse break or pause before the particle followed by the initiation of a discourse unit introduced by an evaluative audience-oriented marker.
(1) so what you have here, are pike per second (LE2/EC) (2) so what you can conclude is

(3) And so what you record is
(4) And what you get here is The audience-oriented relevance marker with the pattern: deictic+verb to be (present or past form)+adverb+a meta noun also has a significant number of occurrences (11 instances). This seems to be a preferred form with the use of a modifying adverb followed by a meta-noun, although very closely followed by the same syntactic structure with the use of adjectives plus meta-nouns, which are also relevant evaluative markers in the EC. Some examples of these recurrent uses of marker patterns are shown below: (5) .the only case where ly a problem . with most cancer drugs (LE1/EC) (7) .and this is always the assumption It is worth pointing out here that not only adverbs and adjectives are relevant audience-oriented markers. Observations from the corpus revealed that verb clauses are also important, giving evidence of other patterns which, although not falling into any of the previously identified lexico-grammatical patterns, are also used as evaluative token attention and guiding them through the lecture discourse. Some of these expressions are introduced as discourse units, usually by particles such as so, then or now.
(8) then this just outlines the synthesis (LE2/EC) (9) this just shows the same kind of data (LE2/EC) (10) This now brings us to the third step (LE1/EC) (11) So this clearly shows that the plasma (LE1/EC) Previous research on relevance markers also discussed about the predominance of verb patterns as relevance markers in lectures especially imperative patterns containing mental verbs such as remember attributing this aspect to a and Taverniers, 2012: 229).

Relevance markers in the SC
The most outstanding marker patterns found in the SC, together with the number of instances and total counts per million words can be seen in Table 3. This being a contrastive study between two different languages within the same spoken genre and discipline, syntactic and lexico-grammatical Spanish rules had to be taken into consideration. Thus, in the case of relative clauses, the Spanish expression lo que is a combination of the neutral article lo followed by the relative pronoun que. The neutral article lo in Spanish does not have any identical counterpart in isolation in English, although the combination lo que closely corresponds to the uses of the English translating counterpart what in The neutral article lo in Spanish can, in other cases, change the grammatical category of the adjacent adjective to a noun, a grammatical process called sustantivización in Spanish, that is, nominalisation . Observe the examples below that fit into the marker pattern Det+adj+ is 3 : The important thing (LE3/SC) (13) Lo más importante aquí es que los teji The most important thing (LE3/SC) In examples 12 and 13 above, the adjective important used in its zero and superlative degrees undergoes a nominalisation process due to the fact that it is preceded by the neutral article lo. In the English lectures this marker pattern is also frequent with the variation that an abstract noun seems to be necessary and would appear in the expression. Such nouns include: thing, point, issue, etc., as already discussed in the previous section (note the underlined thing in the translation of the previous examples).
Although fewer audience-oriented relevance markers that could fit into fixed lexico-grammatical patterns were found in the SC, it is worth pointing out All Spanish examples extracted from the SC have been provided with an English translation for better understanding. that some other expressions and utterances were found with an evaluative function that explicitly indicate to the audience what they should consider relevant. Most of these expressions tend to be rhetorical questions, these being rather more recurrent in the Spanish lectures compared to the English lectures. A total of 73 instances of the 2 nd person plural of the present tense of the verb entender appear in the SC in the form of a rhetorical question, usually preceded by a short pause and followed by another question or rephraser such as es decir, relevance markers in the SC are presented below: These forms of rhetorical questions in the Spanish medicine lectures are not aimed at eliciting an answer illocutionary force to produce an effect in the audience and highlight that the information they are giving is relevant.

Evaluative adjectives in the EC
Adjectives are an essential part of the audience-oriented relevance markers studied here. The most frequently used evaluative adjectives within the marker patterns analysed in the EC can be seen in Table 4 below. Earlier research on evaluative language in the MICASE corpus identified a series of relevant adjectives that proved to be significant (Swales and Burke 2003), namely: important, central, main, major, relevant, crucial, essential, fundamental and key. This previous information has been of great help in order to facilitate the identification of some audience-oriented relevance markers in the English medicine lectures. Six highly frequent evaluative adjectives have been identified as part of the marker patterns. Important has the highest number of instances followed by main, interesting and key; other adjectives in the corpus taking part of the lexico-grammatical patterns are critical or significant, with isolated instances. Some evaluative adjectives coincide with the ones established by Swales and Burke (2003), while others are exclusive to the English lectures under study. All these relevant adjectives as part of the marker patterns become evaluative and they clearly cause audience engagement to the discourse by triggering a reaction. Some examples from the corpus can be read below. Evaluative adjectives in the SC The most frequently used evaluative adjectives as part of the audience-oriented relevant marker patterns analysed in the SC can be seen in Table 5 below. In the case of relevant adjectives in the SC, the search had to include number and gender differences, and thus masculine and feminine forms of adjectives were analysed in their singular and plural forms. At first glance, it was observed that the medicine lectures in Spanish presented a smaller variety of adjectives than the English lectures. Some of the most recurrent adjectives coincided in both corpora, such as importante interesante where importante the top of the list in both corpora. A closer look at the SC revealed that, in spite of a smaller number of types, relevant adjectives were more frequently used in the Spanish lectures as part of audienceoriented marker patterns. Another significant characteristic of evaluative adjectives in the SC is the use of their superlative form. The English medicine lectures rarely used pure superlative forms when expressing evaluation, whereas, in contrast, the Spanish lectures tend to use the superlative form of adjectives in their two grammatical structures, the ending -ísimo/a/s (the adj+-est, in English) and the form más+adj (the most + adj, in English The most important (LE3/SC) (28) Esto es curioso pero sale repetidamente sale repetidamente que las mujeres tenéis peor calidad d This is curious but it has been shown over and over again (LE3/SC) In general, Spanish medicine lectures seem to prefer the use of adjectives to express explicit evaluation in their attributive or predicative forms, whereas more complex syntactic patterns were found in the EC in the forms used to express audience-oriented relevance markers. There is no exact match in the use of relevant adjectives between the SC and the EC, which supports the idea that identical translating correspondences between two language systems do not always occur and should not be taken as the basis for foreign language teaching.

Conclusion
This study departed from the following research questions: a) Do explicit evaluative forms of attitudinal language differ in type, form and frequency between Spanish and English lectures in the field of Health Sciences?, b) Is language more evaluative in any of the corpora analysed? Regarding the use of audience-oriented relevance markers in academic lectures in the field of Health Sciences in English and Spanish, the results showed that although some similarities were found in the identification of marker patterns and the use of relevant adjectives, there were differences in terms of the variety of those patterns and the use of such adjectives. These differences, resulting mainly from the different linguistic systems, indicate that equivalent translation of counterparts from one language to another is not always an effective and reliable pedagogical asset for foreign language teaching. Moreover, contextual analysis revealed the use of some collocational patterns in both corpora that could be included in language teaching syllabi and taught to language students. The creation of valid closed classifications of lexico-grammatical and syntactic patterns counterparts between two languages is by no means a trivial task. Some previous studies on evaluation have focused on disciplinary variation, stating that in the hard sciences explicit evaluation is less common as a consequence of a more impersonal style (Hyland, 2004). Generalisations of this sort cannot be drawn from the study presented here, since all the lectures corresponded to one single discipline: Health Sciences. The results found in the study show evidence that the use of evaluation in the form of audience-oriented relevance markers is crucial for the functional relationships between speakerhearer and vice-versa throughout the discourse elements. Adjectives are carefully selected and slightly preferred by the SC to express evaluation, the Bellés-Fortuño, 2016b: 71), that is, the ability to create interpersonal relations with the listener through expository clarity.
The variety of audience-oriented relevance patterns found in both corpora and the use of other forms of evaluative expressions, such as rhetorical questions or other pragmatic devices (pauses, changes in the pitch of the voice, etc.), reaffirm the spontaneity of academic speech, which is closer to conversation than to academic prose on some occasions (Swales, 2004). This variety may also be subjected to other factors such as idiolectal variations of the speaker (lecturers in this case) or their own individual lecturing styles. The contextualisation of marker patterns has shown features of evaluation in the form of clusters or collocations in medical discourse, evaluative units have been identified thanks to some collocational evaluative marker patterns with conjunctions such as and -fillers such as well such as I mean The English and Spanish medicine lectures used here have shown characteristics of conversational discourse where the audience-oriented relevance markers analysed correspond to those metadiscourse elements of spoken interaction between speaker and listener.
The findings of this study, while conditioned by the limited data analysed, introduce evidence of the use of evaluation expressed by audience-oriented relevance markers in the academic spoken discourse of Health Sciences. Small corpus studies such as this one allow the combination of a quantitative and an easier qualitative interpretation of data for a deeper analysis of the texts in this genre, rather than other larger corpus studies, where findings are sometimes based entirely on quantitative data.
The results obtained contribute to called contrastive discourse analysis (Taboada, Doval and González, 2013: 2). Lectures, as the most extended spoken genre among Higher Education institutions, have a strong pedagogical objective. When lecturing, lecturers try to build a convincing argument and in doing so they use, among other discursive features, audienceoriented relevance markers such as the ones presented and analysed in this study in order to express their attitudes and opinions towards what they are presenting. By comparing audience-oriented relevance markers in two language systems (English and Spanish) we obtain both similarities and differences. The similarities can help in the comprehension of the two languages, that is, they are , 2009: 1). The differences between lecture discourses in both languages, on the other hand, can help to predict and prevent potential learning difficulties, while at the same time being useful to develop pedagogical teaching material and techniques for foreign language teaching. Attitudinal and emotional linguistic expressions cannot be ignored in the language learning process (Alba-Juez, 2015). Contrastive studies paying attention to relations speaker-hearer and their context can greatly help foreign language teaching and learning processes. This study can therefore benefit undergraduate and postgraduate students, as well as novice lecturers and/or researchers, without forgetting medicine professionals who want to learn about the conventions of medical discourse. Likewise, large Higher Education discourse communities can benefit from this contrastive study, which explores the improvement and success of academic communication English and Spanish.