2.3 Textual Sentence Stem and Evaluation

2.3.1 The Multifunctionality of TSSs

Functionally，the term‘textual’owes much to Halliday’s（1994）functional classification of language into the ideational，the interpersonal and the textual.The ideational component refers to resources for construing experience，the interpersonal component relates to humans’attitude and social relations，and the textual component involves the internal organization and communicative nature of a text.This division has important implications for the sociological understanding of human language，but it tends to hide the fact that most textual devices are at the same time interpersonal and evaluative，and vice versa.It is believed by many that there is no clearcut boundary between what is textual and what is evaluative.Thompson&Hunston（2000：6），for example，identify three functions that evaluative language performs.These functions are：

（1）to express the speaker’s or writer’s opinion，and in doing so to reflect the value system of that person and their community；

（2）to construct and maintain relations between the speaker or writer and hearer or reader；

（3）to organize the discourse.

More crucially，as is pointed out by Thompson&Hunston，these three functions are not exclusive.That is，“a single instance of evaluation may well perform two or three of the functions simultaneously”（ibid：6）.

This multifunctional view is echoed by Hyland（e.g.1996c：437）who notes that it is impossible to relate a particular form exclusively to a specific function.Therefore he adopts a fuzzy set model whose essential features are that：

［t］hey allow the gradual transition to membership so cases are not restricted to binary end-points but can range between them，with some examples denoting a greater degree of membership of one category than another.At the core，an expression will most closely approximate to the meaning of that category while examples at the periphery will exhibit less precise meaning.（Hyland 1996c：438）

For example，the phrases I argue that and my purpose is are seen by Hyland as textual metadiscoursal devices under the label of‘frame markers’（1998a：229）.However，both I and argue are highly evaluative in nature and are by themselves typical hedging devices according to Hyland’s observations.He（1996c）therefore suggests that such indeterminacy might reflect Zadeh’s（1972）observation that most classes in the real world are fuzzy.

Despite the fact that it is virtually impossible to provide a single，unequivocal pragmatic interpretation for a particular device，it is（for practical purposes）necessary for this study to identify those multiword expressions whose primary or predominant role is that of expressing the writer’s opinions and that of maintaining writer-reader relationships.However，it must be remembered that many so-called textual devices play a crucial role for expressing interpersonal meanings.It can even be said（though boldly）that many textual devices are at the same time interpersonal.Therefore there is no doubt that it will be equally appropriate if we turn our attention to the organizational aspect of TSSs.

There has been a wide spectrum of studies that set out to investigate how human beings express personal opinions and attitudes towards an entity，a proposition or other people.This phenomenon has been examined under various rubrics.Thompson and Hunston call it evaluation which refers to“the broad cover term for the expression of the speaker’s or writer’s attitude or stance towards，viewpoint on，or feelings about the entities or propositions that he or she is talking about”（2000：5）.Biber et al.use stance to indicate“personal feelings，attitudes，value judgments，or assessments”（1999：966）.Lyons（1977）develops modality to indicate the way to express permission，obligation，volition，and the speaker’s degree of confidence in the truth of the proposition.Chafe（1986）employs evidentiality which includes the reliability of the speaker’s knowledge，the source of the knowledge，the mode of knowing，and the matching of knowledge against the verbal resources that are employed by speakers or against discourse expectations.Martin（2000）develops appraisal theory which is believed to be“the only systematic，detailed and elaborate frameworks of evaluative language”（Bednarek 2006：32）.These studies contribute vitally to the understanding of evaluation.One limitation of these studies，however，is that they commonly treat individual words and set phrases as evaluative devices，and we are still ignorant about how these devices realize attitudinal functions in particular lexical combinations.

In the present book，the adjective‘evaluative’is used in much the same way as‘interpersonal’，‘interactive’and‘attitudinal’.More often than not，we prefer‘evaluative’because of its syntactic and morphological flexibility.For example，‘evaluation’can be seen as a label of functional category，‘evaluator’refers to the linguistic device performing evaluative functions，and‘evaluate’can be conveniently used to refer to the act of expressing personal opinions and attitudes.

Evaluative language has also been investigated under the rubrics of‘hedging’and‘metadiscourse’initially put forward by Lakoff（1972）and Crismore（1989）respectively.The concept of hedging has been most adequately developed since the mid-1990s by Hyland（1994，1996a，1996b，1996c，1998b），who incorporates this concept into a broader framework of metadiscourse in accordance with Crismore but in the more specific context of English for academic purposes（e.g.1998a，1998c，2004，2005b；Hyland&Tse 2004）.We leave these two notions to the next section where a literature review will be conducted with respect to the research on evaluation in academic writing.