Previous Page

Return to listing of notes in this Guide | Return to Appraisal Homepage

5. Engagement and Dialogistic Positioning

An outline of Engagement

In this set of notes we are concerned with the diverse range of resources by which speakers/writers adjust and negotiate the arguability of their utterances. Under the Appraisal framework, such resources are grouped under the heading of "Engagement". The category of Engagement includes values which have been analysed in the literature under headings such as attribution, modality, hearsay, concession, polarity, evidentiality, hedges, boosters and metadiscursives.1

As indicated above, these Engagement resources provide the means by which speakers/writers adjust and negotiate the arguability of their propositions and proposals. More particularly, they are the means by which any utterance, whether in single-party discourse (e.g. writing) or multi-party discourse (e.g. conversation), can be construed so as to reveal its inherent dialogistic potential - that is to say, its location and functionality with reference to past, present and future processes of communicative exchange. Thus by the use of these resources, the terms of the arguability of any utterance can be varied by adjusting the dialogistic status of the utterance, by varying the way in which it is positioned to engage with past, present or future communicative exchanges.

Key Engagement resources include meaning which can be grouped together under the following headings,

1. Disclaim: includes Denial and Counter-Expection

2. Proclaim: includes Expectation and Pronouncement

3. Probabilise: includes Evidence, Likelihood and Hearsay

4. Attribute: e.g. The head of Clinton's security division says this will damage trust. As a number of security experts have indicated, this will damage the trust between President and body guard.)2

The notion of "Dialogism"

Now, I have said these are all meanings by which the speaker/writer can adjust the dialogic terms or status of an utterance. What do I mean by `dialogic terms' and in what ways do these resources achieve such a rhetorical outcome?

This notion of `dialogism' is inspired by the now widely influential view of the communicative process as set out in the work of Bakhtin/Voloshinov. (Bakhtin 1981, Voloshinov 1995) The following quotation sums up this perspective.

The actual reality of language-speech is not the abstract system of linguistic forms, not the isolated monologic utterance, and not the psychological act of its implementation, but the social event of verbal interaction implemented in an utterance or utterances.

Thus, verbal interaction is the basic reality of language.

Dialogue, in the narrow sense of the word, is of course only one of the forms - a very important form, to be sure - of verbal interaction. But dialogue can also be understood in a broader sense, meaning not only direct, face-to-face, vocalised verbal communication between persons, but also verbal communication of any type whatsoever. A book, i.e. a verbal performance in print, is also an element of verbal communication. ...[it] inevitably orients itself with respect to previous performances in the same sphere... Thus the printed verbal performance engages, as it were, in ideological colloquy of a large scale: it responds to something, affirms something, anticipates possible responses and objections, seeks support, and so on. (Voloshinov 1995: 139)

My point then, is that the resources included within Engagement are all `dialogistic' in this sense - they are all means by which speakers/writers represent themselves as engaging in a `dialogue' to the extent that they present themselves as taking up, acknowledging, responding to, challenging or rejecting actual or imagined prior utterances from other speakers/writers or as anticipating likely or possible responses from other speakers/writers. Or, to put it in other terms, they are dialogic in that, to different degrees and in different ways, they all acknowledge or invoke representations or points of view which are to some degree different from the representation/point of view currently being advanced by the text. It is with this alternative position, therefore, with which the speaker/writer presents themselves as engaged dialogically.

I will now consider more specifically the terms of dialogistic positioning which associated with the different sub-choices within Engagement.

Disclaim: deny

Under `disclaim' (which includes Deny and Counter-Expect) we are concerned with resources by which some prior utterance or some alternative position is invoked so as to be rejected, replaced or dismissed as irrelevant or some way communicatively inactive. From a dialogistic perspective, we can see Denial (negation) as a resource for introducing the alternative positive position into the dialog, and hence acknowledging it and engaging with it, and then rejecting it. Thus in these interpersonal/dialogistic terms, the negative is not the simple logical opposite of the positive, since the negative carries with it the positive, while the positive does not reciprocally carry the negative.3 This aspect of the negative, though perhaps at odds with common-sense understandings, has been quite widely noted in the literature - see for example, Leech 1983: 1014, Pagano 1994 or Fairclough 1992: 121.) Consider, for example, the following extract from an advertisement placed in magazines by the British Heart Foundation.

We all like something to grab hold of. But sometimes you can have too much of a good thing. And a man whose table diet consists of double cheeseburgers and chips can end up looking like a tub of lard. There's nothing wrong with meat, bread and potatoes. But how about some lean meat, wholemeal bread and jacket potatoes?

Here the denial, `There is nothing wrong with meat, bread and potatoes', is clearly dialogic in the sense that it invokes, and presents itself as responding to, claims/beliefs that `This IS something wrong with meat, bread and potatoes'. A prior and alternative position is thus clearly engaged with dialogistically.

Disclaim: Counter-Expect

Here we are concerned with formulations which represent the current proposition as replacing and supplanting a proposition which would have been expected in its place. Consider, for example,

Surprisingly, McGuinness is especially scathing about `the chattering classes', of which he has long been a member. (Dissent: p.6, Number 4, Summer 2000/2001)

Here, the writer invokes the alternative proposition, that `McGuinness would not be scathing of the chattering classes' but indicates that it does not, after all, apply. The dialogism, therefore is with the alternative position (the unrealised expectation) which is rejected. In situations where a comment adjuncts such as `surprisingly', `amazingly', `bizarrely' are used, the actual or imagined communicative respondent is represented as sharing the unrealised expectation with the writer/speaker. In other instances, however, the expectation is more clearly presented as that of the imagined respondent, and not of the speaker/writer. Consider for example,

They [Kevin and Ian Maxwell, sons of Robert Maxwell] have a lot to prove in the coming years. Now they will not only seek to make their own fortunes but to clear their father's besmirched name. They grew up to see him as the eternal outsider, the man who had fought Establishment prejudice and pettifogging bureaucracy to get where he was. Sure, he broke rules. Yes, he ducked and dived. Admittedly, he was badly behaved. But look at what he had achieved. From nothing, he had become a multinational businessman with an empire stretching across the world, the confidant of statesmen and just as famous himself. (From the Bank of English UKMags corpus)

The extract (from The Times) is concerned with the notorious British businessman, newspaper magnate and former Labour MP, Robert Maxwell (now deceased) and his two sons, Kevin and Ian. In the extract, the writer seeks to explain, even justify, why the two sons might have continued to regard their father favourably, despite the negativity with which Maxwell had come to be viewed generally. (Maxwell had been found after his death to have secretly diverting millions of dollars from two of his companies and from employee pension funds in an effort to keep his business empire solvent.). Here, very obviously, the writer engages with an imagined dialogic partner or voice which is represented as the source of arguments that Maxwell `broke rules', `ducked and dived' and `behaved badly'. The point here, of course, is that, the obvious inferences these arguments give rise to (that Maxwell was a bad man) are represented as not holding, at least for the Maxwell sons. Through a dialogic interaction, certain views are referenced and then rejected. In this, therefore, we see just how dialogic a single-party written text of this type can become through the use of such counter-expectational resources.5 (For a further discussion of the dialogism of concessives, see Hunston 2000: 178-181. For further discussion of the similar functionality of comment adjuncts such as `amazingly' and concessives, see Thompson and Jianglin Zhou 2000)

Counter-Expect includes both the comment adjuncts of the type discussed above (amazingly, surprisingly) and the wide array of formulations for realising what is usually termed `concession'. These formulations typically include some sense of `although', `however' or `but'. A related sense of Counter-Expectation can also be found in many uses of only, just, even, already and still (For a more extended discussion see Martin 1995: 230-234.)

Proclaim: Expect

Under `proclaim' (which includes Expect and Pronounce) we are concerned with formulations which can be interpreted as heading off contradiction or challenge from potential dialogic respondents . They are meanings which increase the interpersonal cost of any such contradiction by adding additional support or motivation for the current proposition/proposal.

Through values of Expect, the speaker/writer represents the current proposition/proposal as uncontentious within the current speech community, as a `given, as being in accord with what is generally known or expected. Consider by way of example the use of `of course' in the following.

When, belatedly, their selectors chose Paul Adams, who would assuredly have won them the second Test in Johannesburg, their attack became `very good' in the opinion of Trevor Bailey, who has seen a few in his time. Bailey, of course, was that rarity, a cricketer who at his best was world-class with both bat and ball. (From the Bank of English OzNews corpus)

Here the writer represents himself/herself as simply agreeing with the reader, as recounting a view (that Bailey was a cricketing rarity) which is already held by the dialogic partner and by people in generally. The location of the current proposition within a dialogistic exchange is thus employed to increase the cost of any subsequent challenging or rejecting of the proposition.

Proclaim: Pronounce

Under `Pronounce' we are concerned with formulations by which speakers/writers interpolate themselves directly into the text as the explicitly responsible source of the utterance. This `pronouncement' may take the form of an explicit interpolation of the speaker into the text (`I'd say this will lead to mistrust.'), an intensifying comment adjunct (`Really, this will lead to mistrust'), stress on the auxiliary (`This did lead to mistrust'), or through structures such as `It's a fact that...'. (See Fuller 1995: Chapter 4 for a discussion of `interpolation'.) Such formulations are dialogistically prospective. The author thereby increases the interpersonal cost of any rejection/doubting of their utterance in future communicative exchanges, rendering such a direct challenge to the author's dialogic position. Of course, through such a strategy, by confronting the possibility of rejection, the author integrates that possibility into the text and thereby acknowledges the dialogistic diversity of meaning making in socially diverse social contexts.

Probabilise (Evidence, Likelihood and Hearsay)

Under Probabilise, I include all resources by which the current proposition/proposal is represented as just one of a range of possible propositions/proposals. It includes,

  • evidential formulations such as it seems, apparently, the evidence suggests
  • forms which represent the proposition/proposal as more or less likely (including modals of probability and related forms such as I think/I suppose, as well as certain `rhetorical' uses of questions),
  • hearsay/quotatives such as I hear and It's said.

Such formulations have often been classified as `hedges' and have often been seen as indicating that that the speaker is uncertain or tentative. Within frameworks inspired by the concerns of formal logic, they are often interpreted by reference to notions of `truth-value' - they are seen as indicating that the writer/speaker declines to commit to the truth of his/her proposition. (See, for example. Lyons 1977: 452) Such interpretations all operate within a framework by which the communicative process is seen as a form of self-expression, a process by which the speaker/writer's primary purpose is to convey their inner thoughts and beliefs to the outer world. Thus, if a speaker frames an utterance with a formulation such as `it seems to me', then this usage is seen as necessarily revealing some aspect of the speaker's current state of mind, some condition of the knowledge or beliefs they are seeking to communicate - presumably the speaker's uncertainty or lack of commitment to truth-value.

From a dialogistic perspective, however, we come to see such resources rather differently. We see their functionality in terms of the dialogistic negotiation which all speakers/writers undertake. By the inclusion of an `it seems', a `probably' or an `I hear', the speaker actively represents the proposal/proposition as contingent, as located in some individual subjectivity, in some individual assessment of likelihood or of the available evidence. The utterance is thus construed as but one of a range of possible utterances, since different contingencies and different individual subjectivities may well result in different assessments of likelihood and the available evidence. Thus, by the use of values such as It seems..., probably..., I hear... to frame a proposition/proposal, the writer/speaker opens up the space for dialogistic alternation, for a potential response which in some way challenges or differs from the current utterance. In a sense, such forms acknowledge that such alternation is expected or at least possible and accordingly provide an interpersonally more favourable context for such alternation. Thus, as Hyland has observed ( 2000: 88) such formulations anticipate the affect that the current utterance is likely to have upon actual or potential interlocutors and, as Myers has observed ( 1989), reveal the writers/readers purposes in negotiating their claims with these interlocutors. By way of brief illustration of these points, consider the following extract from a linguistics text book on language learning/acquisition. Here the writer is arguing a case with respect to Genie (the young woman who, in a celebrated case, had been found to be almost entirely without language.)

What can we say, then, in answer to the question as to whether Genie acquired language, and acquired it normally? The passage from Curtiss quoted above shows that she thinks Genie had acquired language, but hadn't acquired it fully. It seems to me that we can barely allow Genie into the category of those who have acquired language, and certainly we can't allow her into the category of those who have acquired it naturally and fully. (Cattell 2000l: 199)

Here we find probabilising formulations (it seems to me, certainly) which are associated with assertions to which the writer is strongly committed - there is no sense of uncertainty, tentativeness, equivocation or lack of commitment. The formulations here operate with an obvious dialogic functionality - they serve to mark these propositions as points of contention in the current debate over language acquisition and hence as points at which a difference of opinion is expected from anticipated dialogic respondents.

Attribution and extra-vocalisation

Attribution has been considered at some length in the previous set of notes (Stage 4). Here we consider attribution in the context of dialogistic positioning and introduce an additional term, "extra-vocalisation". Within Engagement, we distinguish two broad categories of resources for negotiating dialogistic or inter-subjective positioning - what are termed "intra-vocalisation" and "extra-vocalisation". Under extra-vocalisation, we are concerned generally with what has previously been termed attribution, with resources which involve the inclusion in the text of some explicitly external voice (hence the term extra-vocalisation). This extra-vocalisation contrasts with resources in which the voice involved in the dialogistic positioning is an internal voice, that is to say, the voice of the speaker or author or writer. All the resources considered earlier in this set of notes (for example, those of Proclaim, Disclaim, Probabilise etc) involve "intra-vocalisation" since the voice which proclaims or disclaims or probabilises is the internal voice of the speaker or writer. I will return to this distinction subsequently. For the moment I take up the question of the dialogistic functionality of extra-vocalisation (attribution)

As already indicated, extra-vocalisation involves the quoting or referencing the statements or points of view of external sources. The rhetorical functionality here somewhat complex because it involves both dialogistic positioning and what can be termed `heteroglossic'6 positioning. Attribution is obviously `heteroglossic' in that it introduces an additional voice into the text - a text with attributions will necessarily be multi/diversely-voiced. Our concern here, however, is not so much with the relationship which the writer/speaker enters into with the quoted source (a relationship of heteroglossic positioning which was explored at length in the previous set of notes) but with the way that writers/speakers uses extra-vocalisation to position themselves dialogistically with respect to actual and potential communicative partners. Consider for example, the following,

Christian Jacq, perhaps the world's most prominent Egyptologist, has argued compellingly that when it came to backroom intrigue and regional betrayal, the modern Middle East still has a lot to learn from ancient Egypt.

Here we encounter a rhetorical manoeuvre which has two aspects. The first is heteroglossic - a second voice is introduced into the text and that voice is evaluated as highly authoritative and convincing. The second is dialogistic - the proposition that ancient Egypt was a place of intrigue and betrayal is associated with an individual subjectivity (that of the attributed source) and is thereby construed as contingent and hence arguable in the current dialogistic context. The degree of arguability that the writer allows, however, is rather limited as a consequence of the high expertise associated with the source. Thus the rhetorical effect of such a formulation is somewhat akin to that of the Pronouncements discussed previously.

There are a number of factors which determine the dialogistic positioning which can result from a given extra-vocalisation. These include the degree of authority which is indicated of the source and the degree to which the writer/speaker endorses (or dis-endorses) the attributed material. Thus the following involve different dialogistic positionings.

  • As X, perhaps the world's leading authority on Y, has demonstrated, ... (high authority / authorially endorsed, the writer indicates they share responsibility with the source for the proposition/proposal)
  • X says that... (neutral with respect to endorsement)
  • Some Xs have claimed that...(dis-endorsed, author disavows responsibility for the proposition/proposal)

Dialogistic positioning and terms of arguability

In the above, therefore, I have described the individual sub-systems of Engagement essentially as a list. In order, however, to be able to apply the system usefully to critical text analysis, we need more systematically to consider what consequences for the arguability of a given utterance result from choosing one Engagement option, and hence one mode of dialogistic positioning, over another. In order to do this, I will explore two parameters by which rhetorical effect and terms of arguability can vary across the Engagement options. The first parameter relates to what I term dialogistic expansion/contraction and turns on the degree to which the text acts to engage with the diversity of view point (the heteroglossic diversity) activated by the current utterance - whether the text is opening up or closed down to this diversity. The second parameter relates to what I previously termed extra-vocaliations/intra-vocalisation (or externalisation/ internalisation), a distinction which turns, as already indicated, on whether the voice of the current proposition/proposal is represented as external or internal to the text, whether or not the voice is that of the author/speaker or that of some external source. I will discuss each of these parameters in turn.

Dialogistic expansion and contraction (open/close)

Engagement resources present the speaker/writer as dialogistically engaged. The nature of this engagement can differ according to whether the Engagement value employed presents the speaker as opening up the dialog to more or less divergent positions or as closing it down so as to suppress or at least limit such divergence. Resources grouped together under Disclaim are generally contracting or closing since, while they acknowledge alternative positions within the dialogistic context, they either reject or directly challenge these. In the case of Denial and Counter-Expect, alternative positions are closed down by being directly rejected or by being replaced. Through Expect and Pronounce, the space for dialogistic diversity is contracted by what amounts to a pre-emptive rhetorical action - the writer/speaker is presented as seeking to constrain possible dialogistic divergence by overtly and strongly indicating their personal investment in the current proposition/proposal. Under Disclaim, then, the terms of arguability are adjusted so that any challenge or questioning of the current utterance puts more at stake interpersonally. Any challenge necessitates a direct confrontation with the speaker writer and in the case of Expect, a confrontation with what is represented as `common-knowledge' or `public opinion'.

In contrast, the resources of Prababilise (Evidence, Likelihood and Hearsay) act to expand or open the space for dialogic diversity and difference. By the use of such resources the speaker/writer indicates that the current assertion is but one of a number of possible alternative assertions and simultaneously indicates that these alternatives are, at least to some degree, anticipated and hence dialogistically authorised. Under Probabilise, the terms of arguability are thus adjusted so that any challenge to, or questioning of, the current utterance would puts less at stake interpersonally. Under Probabilise, the proposition/proposal is, in fact, overtly characterised as arguable or contentions and hence challenge, contradiction or alternation are explicitly authorised dialogistically.

Formulations of Attribution will either expand or contract according to a range of variables, including the authoritativeness of the attributed source and the extent of authorial endorsement of the attributed proposition. An endorsement-neutral formulation such as `Some researchers argue...' will tend towards dialogistic expansion, since the proposition/proposal here is not afforded any enhanced argumentative force. It is represented as simply one view among many. In contrast, endorsed formulations (for example, `As X has so compellingly demonstrated) will tend to contract the scope for dialogistic diversity. Through such formulations, the writer not only indicates their personal investment in the current argument, but adds to the argumentative force by representing the current view as one which is not theirs alone but one which is shared with, for example, the wider community or with relevant experts.

We can, then, make a broad distinction, then, between Engagement resources which contract the space for dialogistic diversity and difference (Denial, Counter-Expect, Expect, Pronounce and authorially-endorsed Attribution) and those which expand the space (Evidence, Likelihood, Hearsay and some values of endorsement-neutral Attribution). I represent this distinction below.

Contracting dialogistic diversity

(Disclaim:Denial) McGuinness doesn't have anything positive to say about `the chattering classes', of which he has long been a member.

(Disclaim:Counter-Expect) Surprisingly, McGuinness is scathing about `the chattering classes', of which he has long been a member.

(Proclaim:Expect) McGuinness is, of course, scathing about `the chattering classes'

(Proclaim:Pronounce) You'll have to agree with me that McGuinness is especially scathing about `the chattering classes'

(Extra-vocalise: authorially-endorsed) A number of leading media analysts have compellingly argued that McGuinness is scathing about `the chattering classes'

Expanding dialogistic diversity

(Probabilise:Evidence) It seems that McGuinness is scathing about `the chattering classes'

(Probabilise:Likelihood) It's possible that McGuinness is scathing about `the chattering classes'

(Probabilise: Hearsay) I hear that McGuiness is scathing about `the chattering classes'

(Extra-vocalise: endorsement neutral) Some writers hold that McGuinness is scathing about `the chattering classes'.

In the above, I have represented the `dialogistic expansion' versus `dialogistic contraction' relationship as binary or taxonomic - Engagement formulations are represented as either one or the other. It is possible, however, to see the resources as lying along a cline between most contracting (Disclaim) and most expanding (endorsement-neutral Attribution).7

The rationale for this ordering is as follows.

Disclaim:Deny: Under Denial, a specific dialogistic alternative (the contrary position) is directly rejected. Thus, while the alternative is referenced, it is given minimal dialogistic space - the formulations allows little scope for negotiation of alternative positions.

Disclaim:Counter-Expect: Under Counter-Expect a particular expectation or inference is invoked. This expectation is not directly rejected, as is in the case with Denial, but, rather, is replaced by an alternative. Thus the expectation (in the above examples) that McGuinness would speak relatively favourably of the social grouping of which he is a member is not denied outright. It is simply replaced with the alternative, that he is `always criticising the chattering classes unfairly'. Accordingly, I conclude that with such an instance of Counter-Expect there is, relative to a Denial, more dialogistic scope for the alternative - the alternative is somewhat more arguable.

Proclaim:Expect: Under Proclaim (and all the remaining Engagement options) alternative positions are no longer directed invoked, though they are anticipated. (We move from retrospective to prospective dialogistic positioning). With Proclaim:Expect formulations such as `of course' or `predictably', the speaker/writer indicates a high level of commitment to the proposition and by this, perhaps paradoxically, renders the utterance relative or contingent since it is thereby associated with a given individualised subjectivity (presumably the shared subjectivity of the writer, reader and possibly `people in general'). The apparent paradox here is similar to that observed by Halliday in association with high values of modality (e.g. `he must be corrupt', `he is definitely corrupt') where, by the indication of their conviction, the speaker/writer renders the utterance less absolute or less invariable than the bare assertion (`he is corrupt')8. The relativity or contingency of the utterance is the basis of its dialogism. As subjectively based, it opens up a certain limited space for dialogistic alternation - the possibly of dialogistic diversity is acknowledged. The high degree of the writer/speakers conviction, however, and the fact that it is represented as being expected by `people in general' means that the scope for dialogic alternation is relatively minimal.

Disclaim:Pronounce: Values of Pronounce have a similar `paradoxical' rhetorical functionality to values of Expect. By explicitly indicating their conviction, the writer/speaker renders the proposition relative or variable, but, of course, only minimally so. I would argue that Pronounce is somewhat more expansive dialogistically than Expect since here the argument is represented as based in the single subjectivity of the speaker/writer rather than in the more generalised subjectivity of speaker/writer plus reader plus `everyone'.

Probabilise (Evidence, Likelihood, Hearsay): Formulations such as It seems, I think, perhaps, it's possible, I hear, It's said are unproblematically more expansive dialogistically than values of Proclaim. While Pronouncement formulations such as `we can but conclude', `I contend' `undeniably' actively confront, challenge and hence discourage dialogistic alternatives, the Probabilise values effectively invite them. They characterise the current proposition as `just my opinion with which you may well disagree'. (Complications which arise from the arguability of intensified options such as `it's certain', `definitely' and `I'm sure' will be discussed below.)

Endorsement-neutral or Dis-endorsed Attributions: Attributions such as `X says...' (endorsement neutral) or `X claims...' (dis-endorsed) typically are the most dialogistically expansive of Engagement resources9. Here the speaker/writer explicitly distances themselves from the attributed source, indicating that they share no responsibility for the material being asserted. The asserted material is thus represented as simply the observation or view point of one individual among many. Such utterances are maximally arguable since they may be challenged or questioned without any direct confrontation of the authorial voice.

The `undialogised' utterance - the dialogistic status of bare assertions

It remains to consider the status of the bare assertion (e.g McGuinnes unfairly criticises the chattering classes) from this dialogistic perspective. Following Bakhtin, we observe that all communication operates in a social world dominated by heteroglossia, by a diversity of `voices' and socio-semiotic positions. Hence, when we speak, when we adopt a particular socio-semiotic position, our utterances necessarily interact with, or enter into a `dialog' with, all the various more or less divergent social positions activated by that utterance. Utterances which employ some value of Engagement acknowledge this `dialogic imperative' (Bakhtin 1981: 426) Those which do not, which employ the form of the bare assertion, ignore or deny this dialogic imperative and thereby suppress the basic heteroglossic nature of social reality. Accordingly, from this perspective, we do not see the bare assertion as in some way `neutral', `unmediated' or factual - as in some way being the communicative default. Rather we see them as adopting a particular socio-semiotic position, an `undialogized' (Bakhtin 1981: 427) stance by which the inherent dialogism of the communicative process is denied. Thus we see such `undialogized' language as rhetorically, interpersonally and socially charged, as entering into relationships of tension with whatever related set of alternative or contradictory utterances it brings into play. The degree of that tension will, of course, vary according to the social context. It is a function of the number and the social status of those alternative socio-semiotic realities under which the utterance at issue would be problematised. Consider, for example, the difference between an utterance such as `Australia was terra nullius, an empty land, when the first European settlers arrived' and `In the view of some historians/It seems/I think/It's my contention, Australia was an empty land when the first European settlers arrived.'. Under a commonsensical, truth-functional perspective we might view the difference simply as one between `factuality' and `assessment/opinion'. Under the dialogistic perspective we see the first utterance (the bare assertion) as highly charged inter-subjectively since it denies or suppresses the significant heteroglossic diversity and difference within which it is situated and which it will inevitably activate

Accordingly, we categorise the bare assertion as another option within the system of Engagement, an option by which particular dialogistic terms can be set for the current utterance, though of course in this instance, the dialogistic terms are those of denial or suppression. I therefore distinguish broadly between the monologism of bare assertion (I say they `monologise') and the dialogism of the all formulations I have discussed above (I say they `dialogise').

Dialogistic contraction and expansion - a more delicate level of analysis

At this broad level of analysis, then, we associate different Engagement subsystems (for example, Disclaim versus Proclaim versus Probabilise versus Attribute) with different degrees of dialogistic contraction/expansion. It is possible, however, to see this parameter of variation operating more narrowly and at a more delicate level of analysis, as a cline of variation operating within, rather than between, the Engagement sub-systems. Within Likelihood, for example, it is possible to identify different levels of force or intensity - thus, `This is possibly a bad idea' (low), `This is probably a bad idea' (median) and `This is definitely a bad idea' (high). Clearly these different options within Likelihood vary from more dialogistically expansive (possibly) to more dialogistic contracting (`definitely). This optionality is available with many of the Engagement subsystems.

I note in passing, that the more contractive values of Likelihood (e.g. I'm certain this is a bad idea., This is definitely a bad idea, This must be a bad idea) seem quite close in the their rhetorical functionality to values of Pronounce generally. I would certainly not want to argue that there is any major difference here in terms of dialogistic contraction/expansion. The difference is a relatively subtle one - a meta-discursivity distinction turning on whether the speaker indicates that the speaker expressing an assessment of high likelihood or whether the speaker interpolate themselves explicitly into the text as committed `sayer'.

Extra-vocalisation versus Intra-vocalisation

The second parameter by which terms of arguability can be varied turns on the distinction between what I previously termed extra-vocalisation (externalisation) and and intra-vocalisation (internalisation). Under extra-vocalisation, responsibility for the arguability of the proposition/proposition is assigned to some external voice, typically some attributed source. This contrasts with internalising options (intra-vocalisation) where responsibility for arguability is text internal - it remains with the internal authorial voice. This distinction has been very widely examined in the literature in the context of, for example, considerations of the functionality of reported speech and of intertextuality/heteroglossia. In this regard, Hunston, for example, distinguishes between attribution (externalisation/extra-vocaliation) and averal (internalisation/intra-vocalisation). (Hunston 2000) For the most part this a relatively straightforward distinction between material which the authorial voice offers on its own behalf ( This is probably a bad idea / This seems a bad idea. / I hold this to be an undeniably bad idea etc) and material which is attributed to an outside source (e.g. Larry has indicated that he sees this as a bad idea.)

In such clear-cut cases (for example, Some of the agents see this is a bad idea [externalised] versus I think this is a bad idea [internalised]), the difference in arguability and dialogistic positioning is straightforward. As already indicated, this type of externalisation maximally expands the scope for dialogistic negotiation by introducing an additional, external voice to whom the writer/speaker assigns responsibility for the proposition/proposal.

But as has also been indicated above, the distinction and the resultant rhetorical consequences are not always so clear cut. Thus in instances of authorially-endorsed extra-vocalisation (e.g. Larry has compellingly demonstrated that this is a bad idea.), the utterance can be seen as both extra and intra vocalised - it contains both externalising aspects (the attribution to a quoted source) and internalising aspects (the indication that the authorial voice endorses the proposition/proposal). Such formulations can thus bee seen as dialogistically multiple, so to speak - combining both extra-vocalisation and intra-vocalisation so that responsibility for the arguability of the proposition/proposal is ascribed to both the inner and the outer voice .

Values of Hearsay (e.g. it's said.../ I hear...) share this multiple functionality to some degree. On the face of it, Hearsay formulations introduce an external, heteroglossic voice. And yet the fact that the speaker/writer declines to specify or identify the source of the externalised material indicates that they to some degree endorse the proposition/proposal there conveyed - they share some communicative responsibility for the proposition/proposal. The `I hear' formulation takes this one step further by grammatically internalising the extra-vocalisation. However, while authorially endorsed extra-vocalisations combine voices, Hearsay values tend more towards dialogistic ambiguity. The situation remains unclear as to whether responsibility for the utterance lies with the inner or outer voice.

The relationship of various Engagement options to externalisation/internalisation is represented diagrammatically below.

This extra/intra distinction also operates more narrowly (at a level of greater delicacy) to distinguish between options within subsystems. Consider for example the difference in arguability terms between `it seems to me' and `it seems' - both terms within Probabilise:Evidence. The first (it seems to me) is maximally internalised. The `seeming' upon which the arguability of the utterance relies is unambiguously and solely associated with the authorial subjectivity. This is not so absolutely the case with `it seems'. In such a formulation the `seeming' is less clearly specified, it is more generalised or objectified - the `seeming' is not just a matter of inferences drawn by the speaker/writer but, presumably or possibly, by the reader/listener or by the relevant speech community generally. Thus `it seems' - at least when contrasted with `it seems to me' - invokes at least some sense of `externalisation'. It references a dialogistic position which is to some degree separate and hence external from that of the speaker/writer. The same distinction can be seen to underlie the different rhetorical functionality of such pairs as `I think' and `It's probable'. Under the systemic functional framework, such formulations are seen as `grammatical metaphors' of modality (Halliday 1994: 354-9) with `I think' being characterised as the `subjective' and `It's probable' as the `objective' option. From the perspective I have been exploring here, I reconstrue the distinction as a matter of dialogistic positioning, and more specifically as one of different degrees of externalisation. I illustrate these relationships diagrammatically below.

Summary: The cline of extra versus intra vocalisation

More to Follow: power and solidarity

For the moment that's where this introductory course ends. However, more material will appear hear shortly. The additional material will include a section in this set of notes exploring the application of the Engagement framework set out above to issues in critical text analysis. For example, I'll be demonstrating how Engagement can be applied to exploring the issue of the discursive construction of relations of power and solidarity,

More to Follow: Graduation (Force & Focus)

I will also be adding a new section on what in the Appraisal framework is termed Graduation - the set of resources by which the force or tone or intensity of an utterance may be raised or lowered or by which the we vary to "focus" or preciseness of the semantic categories we employ in our communications.

Reference List

Bakhtin, M. 1981. The Dialogical Imagination, M. Holquist, (ed.), C. Emerson & M. Holquist, (trans.), Austin, University of Texas Press.

Cattell, R. 2000. Children's Language - Concensus and Controversy, London & New York, Cassell.

Chafe, W.L. & J. Nichols, (eds), 1986. Evidentialty: The Linguistic Coding of Epistemology, Norwood, N.J., Ablex.

Crismore, A. 1989. Talking With Readers: Metadiscourse As Rhetorical Act (American University Studies Series XIV : Education, Peter Lang Publishing.

Fairclough, N. 1992. Discourse and Social Change, Cambridge, UK, Polity Press.

Halliday, M.A.K. 1994. An Introduction to Functional Grammar, London, Edward Arnold.

Hunston, S. 2000. 'Evaluation and the Planes of Discourse', in Evaluation in Text: Authorial Stance and the Construction of Discourse, Huston, S. & Thompson, G. (eds), Oxford, Oxford University Press.

Hyland, K. 1996. 'Writing Without Conviction: Hedging in Science Research Articles', Applied Linguistics 17 (4): 433-54.

---. 2000. Disciplinary Discourses - Social Interactions in Academic Writing, Edinburgh Gate, Pearson Education Limited.

Jakobson, Roman. 1957. Shifters, Verbal Categories and the Russian Verb. Russian Language Project. Harvard: Dept of Slavic Languages and Literature.

Körner, H. unpublished. 'Moving Between Two Worlds: The Construction of Inter-Discursivity in Legal Judgements', Workshop: Exploring Ingterpersonal Grammar - Systemic Functinal Workshop on Interpersonal and Ideational Grammar, Katholieke Universiteit Leuven, Nov.

Leech, G. 1983. The Principles of Pragmatics, London & New York, Longman.

Lyons, J. 1977. Semantics, Cambridge, UK., Cambridge University Press.

Markkanen, R. & Schröder, H. (eds) 1997. Approaches to the Analysis of a Pragmatic Phenomenon in Academic Texts, The Hague, Walter De Gruyter & Co.

Meyer, P.G. 1997. 'Hedging Strategies in Written Academic Discourse: Strenghtening the Argument by Weakening the Claim', in Hedging and Discouse - Approaches to the Analysis of a Pragmatic Phenomenon in Academic Texts, Markkannen, R. & Schröder, H. (eds), Berline & New York, Walter de Gruyter.

Myers, G. 1989. 'The Pragmatics of Politeness in Scientific Articles', Applied Linguistics 10: 1-35.

Pagano, A. 1994. 'Negatives in Written Text', in Advances in Written Text Analysis, Coulthard, M. (ed.), London, Routledge.

Palmer, F.R. 1986. Mood and Modality, Cambridge, UK, Cambridge University Press.

Thompson, G. & Jianglin Zhou 2000. 'Evaluation and Organization in Text: The Structuring Role of Evaluative Disjuncts', in Evaluation in Text: Authorial Stance and the Construction of Discourse, Hunston, S. & Thompson, G. ( Oxford, Oxford University Press.

Voloshinov, V.N. 1995. Marixism and the Philosophy of Language, Bakhtinian Thought - an Introductory Reader, S. Dentith, L. Matejka & I.R. Titunik, (trans), London, Routledge.


1 For modality, see for example, Palmer 1986and Lyons 1977), for evidentiality see Chafe and Nichols 1986), for hedging Jakobson 1957, Myers 1989, Markkanen and Schröder 1997 and Meyer 1997), for `boosters' (Hyland 1996) and for `metadiscursivity', see (Crismore 1989).


2 Engagement includes a greater range of resources than is typically included under headings such as modality and evidentiality (Chafe and Nichols 1986) and a narrower range than is typically included under the heading of meta-discourse (Crismore 1989). It is, perhaps, most noteworthy that Engagement includes resources of Denial (negation and Counter-Expectation/Concession) The reasons for this are set out in the following discussion.


3 I have noted one instance where the positive did acknowledge the negative. On a sign at the verge of a wide expanse of neatly mown lawn by a footpath in Toronto, Canada, the following `Please Walk On The Grass.'


4 Leech makes essentially this point when he states, `In fact, the [Co-operative Principle] will predict that negative sentences tend to be used precisely in situations when... [the speaker] wants to deny some proposition which has been put forward or entertained by someone in the context (probably the addressee).


5 There is an interesting additional layer of dialogism here which I haven't really attended to. What we have here, in fact, is the writer representing the position of the Maxwell sons as they represent the position of those who view their father negatively.


6 Once again following terminology from Bakhtin.


7 For the inspiration for this topological approach to dialogistic positioning I am indebted to Henrike Körner. See for example, Körner unpublished


8 As Halliday puts it , `you only say you are certain when you are not' 1994: 89


9 It is, of course, crucial to keep in mind that the functionality of the `X says' formulation (and related formulations) will typically vary significantly from context to context. Thus the writer may have indicated elsewhere in the text that the source X is highly regarded or is highly convincing or has a high status and hence any citation of X will carry with it a sense of authorial endorsement and hence result in a narrowing of the scope of dialogistic negotiability.

Previous PageTop Of Page

Return to listing of notes in this Guide | Return to Appraisal Homepage