1Sternberg (1982a, 1982b), Short (1988), Tannen (1989: 110-19) and Fludernik (1993: 409-14) have pointed out that Direct Speech (DS) can be used to present propositions which cannot possibly be accurate reports, either because, for example, too much time has elapsed for memory of an original to be accurate (this often happens in spoken ‘reports’ of speech, as Tannen points out) or because what is being reported as speech never actually occurred, e.g. in hypotheticals like ‘Get lost’ in ‘I would have said “Get lost” but I was too embarrassed by what he said’ (see Short, Semino and Wynne 2002) or what Fludernik (1993: 11) calls ‘condensed speech acts, in which a brief discourse schematically represents an entire speech event’, a notion which is not unlike, but not identical with, the concept of speech summary which I suggested in Short (1988) and will develop here.
- 2 When reporting is cross-language, even these requirements have to relaxed, of course, although one (...)
2Fludernik, following Sternberg, argues via what is termed the ‘direct discourse fallacy’ that the assumption of faithfulness in discourse report has to be abandoned. I have, with others, already argued against this view in Short, Semino and Wynne (2002). There, we argued (a) that it is only when reporting is involved that issues of faithfulness (which effectively means lexical and grammatical faithfulness),2 and so also the stronger notion of verbatim report (see Clark and Gerrig, 1990), apply and (b) that careful consideration of the context and co-text is needed to be sure that reporting is actually taking place, rather than being merely presentation (as in fictional or hypothetical speech) or representation (for example to bring out a contrasting ideological ‘take’ on the original speech). Hence we suggested that for clarity, and to avoid confusion, we need:
- 3 For discussions of discourse in fiction, I prefer, as indicated in Semino and Short (2004) to use t (...)
(a) to distinguish terminologically among (i) discourse presentation (which refers only to the presenting discourse, the posterior discourse in situations of report and representation), (ii) discourse report (which assumes, for direct discourse presentation, a match between the lexis, deixis and grammar in the anterior and posterior discourses) and (iii) discourse representation (which assumes a mismatch between the lexis, deixis and grammar of the anterior and posterior discourses)3 and
(b) to distinguish systematically among (i) speech, (ii) writing and (iii) thought presentation, only using the term ‘discourse presentation’ and its category-specific equivalents (e.g. ‘free indirect discourse’) when talking very generally or when there is ambiguity or uncertainty as to whether one or another form of discourse is being presented.
3It is important for us to be very clear about exactly what we are talking about if we are to characterize accurately the meanings and effects of the various forms of discourse presentation. Moreover, thought presentation, unlike speech and writing presentation is not the presentation of a form of ostensible inter- personal communication; and a proper understanding of the presentation of communication also needs to take account of the fact that the assumptions we have about (i) speech and (ii) writing, although similar in many ways, can also be different from one another. Indeed, I suspect that it is because of descriptive imprecision from traditional times to the present over speech/ writing that much of the recent confusion concerning the concept of faithfulness has arisen.
- 4 Ikeo (2009) discusses interesting cases of DW in literary reviews, where what is quoted is accurate (...)
4‘Speech’ has always been the default term in discourse presentation, as the oratio directa vs oratio obliqua distinction in Latin rhetoric, in spite of the fact that the recording of spoken language has only been possible over the last hundred years or so, shows. Essentially, most discourse presentation which has been described has, for thousands of years, been found in written (and often fictional) texts. Not surprisingly, then, our canonical assumptions concerning faithfulness actually relate to writing (e.g. written scholarly debates), not speech and it is in writing that the assumptions concerning lexical and grammatical faithfulness to an original in DS presentations are strongest (see Short, Semino and Wynne 2002). As students and scholars, we can be accused of unreasonable manipulation if we misquote from other written texts and on most occasions writers, even tabloid journalists, try not to commit the sin of misquotation (which counts in Gricean terms as a violation of the maxim of quality).4 And as teachers, we punish our students severely if they violate this maxim in the other direction too. Plagiarism, the pretence that the words of others, and the propositions they present, are those of the current writer is as unpardonable a sin as that of misrepresenting what others have said.
5All this suggests that the concept of faithfulness needs to be preserved in the real world, otherwise the attitudes I have just referred to cannot be adequately explained. And our responses to fictional discourse presentation are clearly based on the schemata we have developed from our experience of real- world discourse presentation, in line with Ryan’s (1991) Minimum Distance Principle. Of course the notion of faithfulness in fiction is a chimera, as just about everything is invented by the author. In 1st - and 2nd -person fictional narration there was no actual anterior speech situation for the narrator to ‘report’, even though we pretend to ourselves when reading that there was; and in 3rd -person narration it is arguable that the idea of anterior vs posterior discourse situations does not sensibly apply at all - it usually seems that what is said is being said for the first time ‘in front of our eyes’. Note also that being faithful to an original in real-world direct speech (DS) does not normally involve correspondences in intonation and pronunciation between the anterior and posterior situations, presumably because they are irrelevant in written presentations and would require talented oral mimicry in spoken presentations.
6This also shows the salience of writing presentation in the formation of our discourse presentation schemata. Direct speech in novels, for example, has the lively, dramatic qualities that it has, compared with the less dramatic indirect speech (IS) form, precisely because it is associated schematically with a claim to present accurately the lexis, deixis and grammar of the (putative) original, whereas IS does not.
- 5 This is testable empirically, something which would throw useful light on the ‘faithfulness debate’ (...)
7Speech and writing summaries, like the presentation of hypothetical speech, do not constitute presentational report and so cannot be used as counter examples to the faithfulness account. Moreover, I suspect that speech summary is much more extensive than we have noticed so far5 and that many of the examples of inaccurate DS/DW presentation used to date to argue against the notion of faithfulness in discourse report can be seen to be summary, and so not really counter examples at all.
8In section 3 below I will discuss a series of examples of speech summary, but as a prelude to that discussion, I need to outline briefly, for those who are not familiar with it, my current view of the speech, writing and thought presentation scales. My current position, is slightly different from Semino and Short (2004) and the same as that presented in Short (2007), except that I now think it clearer to use the term ‘Presentation’ (and so the acronym ‘P’) rather than ‘Representation’ (and so the acronym ‘R’) for the various category labels, as this term focuses entirely on the presenting situation and so helps us to avoid the trap of confusing presentation with representation and report. In the past I, like others, have unfortunately run these notions together. For example, in chapter 10 of Short (1996), following on from chapter 10 of Leech and Short (1981), I use the terms ‘speech presentation’ and ‘thought presentation’ for chapter and section headings but then use ‘representation’ (which suggests a change from an anterior situation to the posterior, presenting situation) for the Narrator’s Representation of a Speech Act (NRSA) category and its thought presentation equivalent (NRTA).
9For clarity I will first introduce some speech presentation examples (using, as is traditional, roughly equivalent manipulations of an initial DS string) and associated category labels (including the term ‘presentation’ [P] for the reasons outlined above) and then I will outline the faithfulness scales as I currently see them. Note that our assumptions about the effects of the various presentational categories rest on rearranging a DS string to create the other categories and their resultant effects. This helps to explain why we tend to use such proposition-domain manipulations when introducing discourse presentation to students (I distinguish proposition-domain summary from discourse-domain summary in 3 below).
|
Category
|
Example
|
|
Direct Speech (DS):
|
“Just go – now!” he said grumpily.
|
|
Free Indirect Speech (FIS):
|
She should get out now!
|
|
Indirect Speech (IS):
|
Grumpily he told her to leave.
|
|
Narrator’ s Presentation Speech Act (NPSA):
|
Grumpily he ordered her out.
|
|
Narrator’ s Presentation of Voice (NPV):
|
He spoke grumpily.
|
10These five speech presentation categories, and their equivalents for writing presentation, are each associated canonically with differing sets of proposition- domain faithfulness assumptions, as shown below (where the categories are presented in the reverse order from that above so that I can outline the faithfulness claims in ascending quantitative order from one claim, in NPV, to four claims, in DS):
|
Category
|
Faithfulness claims
|
|
NPV/NPW:
|
Speech/writing took place (1)
|
|
NPSA/NPWA:
|
(1) + speech/writing act specified, optionally with the topic indicated (2)
|
|
IS/IW:
|
(1) + (2) + indication of propositional content (3)
|
|
FIS/FIW:
|
(1) + (2) + (3) (+ 4???)
|
|
DS/DW:
|
(1) + (2) + 3 + words and structure used to express the content (4)
|
11I suggest that the canonical faithfulness assumptions stemming from writing presentation effectively ‘wash over’ straightforwardly onto speech presentation. The faithfulness claims increase, one at a time, as we move down the list, one category at a time, from NPV/NPW to DS/DW, except for FIS/FIW (free-indirect discourse is famously a semantic halfway house between the direct and indirect forms). Indeed, the indeterminacy with respect to faithfulness claim 4 (in novelistic terms, raising the issue of whose words are being used, narrator or character) explains why the free indirect category is perceived by readers and hearers in the way that it is. The NPSA category (and NPWA) is often associated with summary, precisely because the most it can contain is a speech act value plus an indication of the topic of speech. For NPSA/NPWA, unlike DS/DW, FIS/FIW and IS/IW, there is no separate propositional form for the presented string. Discussions of faithfulness in discourse report usually centre on the direct categories, as this is where the largest number of faithfulness claims are made, and rarely consider in any detail the two categories with the least faithfulness claims, precisely because propositional faithfulness is not at issue with these presentational forms. When we move from NPSA/NPWA to NPV/NPW in report, where all we are told is that speech or writing occurred, the faithfulness claim is so weak that the term ‘faithfulness’ no longer relates to the form or content of the reported discourse at all and so the relationship to the fuller proposition-domain forms is non- existent and even the term ‘proposition-domain summary’ is inapplicable.
12As I have said above, speech presentation and writing presentation, which both involve the presentation of ostensible communication, seem to act in rather similar ways, with the canonical assumptions being even stronger for writing presentation than for speech presentation. However, thought presentation is not the presentation of a communication between people but the presentation of someone’s inner world. So, in the set of discourse presentation scales in Figure 1 below, I separate the thought presentation scale from the other two scales, as in Semino and Short (2004). Square brackets are used to separate off elements which are linked to the discourse presentation scales but are not technically part of the scales themselves. [N] = sentences of the narration of states, events and actions; [NPS]/[NPW]/[NPT] (Narrator’s Presentation of Speech/Writing/Thought = reporting clauses and other, non-clausal, reporting signals):
Figure 1. The discourse presentation scales
- 6 This DS category is wider than that traditionally used, and includes what is usually known as Free (...)
- 7 DW includes FDW for reasons parallel to those noted in note 3 above.
- 8 IN = Internal Narration, covers the narrator’s descriptions of internal cognitive states which are (...)
- 9 Whether or not we need a DT/FDT distinction on the thought presentation scale needs careful empiric (...)
|
Speech and writing presentation
|
|
[N]
|
[NPS]
|
NPV
|
NPSA
|
IS
|
FIS
|
DS6
|
|
|
[N]
|
[NPW]
|
NPW
|
NPWA
|
IW
|
FIW
|
DW7
|
|
|
|
|
|
|
|
|
Norm?
|
|
Thought presentation
|
|
[N+IN8]
|
[NPT]
|
NPT
|
NPTA
|
IT
|
FIT
|
DT
|
FDT9
|
|
|
|
← Norm?→
|
|
|
|
13Because thought presentation is not the presentation of ostensible communication, the thought presentation scale is constructed on the basis of a less than perfect analogy with the other scales. Some category effects seem to be roughly equivalent, for example the dramatic effects of DT (and some FIT), compared with IT, seem similar to those on the speech and writing presentation scales. But others are not. I argued in chapter 10 of Leech and Short (2007 [1981]) that the differences in effect between FIS (distance from the speaker, irony etc) and FIT (closeness to the speaker, sympathy etc) are a consequence of the fact that the norm for speech presentation (because it is ostensible communication) is DS whereas the norm for thought presentation must be more indirect; consequently the free indirect category represents a move in a different direction from the norm on the thought presentation scale, compared with those for speech and writing presentation scales. The bigger difference between thought presentation and the other forms of discourse presentation is also seen in the need for an extra category of Internal Narration (IN). And, perhaps most importantly for this paper, it is not at all clear that NPTA has a proposition-domain summarizing effect, as NPSA and NPWA usually do, because summary does not seem sensibly to apply to a form of discourse presentation which does not involve ostensible communication and so an ‘original proposition’ is not available to the presenter to be summarized.
14Typically, when stylisticians discuss the NPSA (and NPWA) category they characterize its effect as summarizing in type, for the reasons I have suggested in 2 above. When we establish the various discourse presentation categories, we typically do so, as I did above, by manipulating a proposition in a DS string into the forms associated with the other presentation categories. In other words, the kind of summary that is involved in NPSA and NPWA is effectively proposition-domain summary. However, there is another form of summary, related to whole discourses or parts of them, which I want to call discourse-domain summary. When I was a grammar school pupil many years ago, I was trained to write summaries of texts, to varying lengths (100 words, 500 words, and so on). This was, in effect, I assume, training for possible administrative roles in later life; when secretaries in institutions take the minutes in meetings, what they create, and then present in their ‘published’ minutes of meetings, are discourse-domain writing presentation summaries of anterior speech, sometimes of individual turns in the meeting and sometimes of sequences of turns, summarized together.
15Once we see that speech and writing can be summarized a proposition at a time or a larger stretch at a time, we can see that there might be ambiguities between whether what is being presented in the NPSA/NPWA form is the summary of a proposition or of some larger piece of discourse. This has led me to realize that I made a mistake in section 10.1.3 of Leech and Short (1981) with two invented examples, [12] and [13], which I used to illustrate what I was then calling the Narrative Report of Speech acts (NRSA; now Narrators’ Presentation of Speech Acts [NPSA]):
- 10 These invented examples were changed to be more accurate in the second edition of Style in Fiction.
He promised to return.
He promised to visit her again.10
- 11 I would like to thank Geoffrey Leech for pointing this out to me.
16In fact, both of these examples are formally IS, as the presented string is clearly a clause (albeit a short, non-finite one).11 I suspect that I was assuming without realizing it that the summary was of more than one proposition in each case, and so mistakenly assumed that, because they were summaries, they were examples of proposition-domain NRSA (now NPSA) summary, as NPSA is the obvious proposition-summary category on the speech presentation scale.
17Below, I discuss a series of examples of discourse-domain speech summary. I do not have a full catalogue of summary examples yet (in particular, I am still looking for writing presentation examples), but essentially I want to suggest as a consequence of the analyses below that:
(a) in addition to the more ‘standard’, one-proposition-at-a-time, presentation (including proposition-domain summary for NPSA and NPWA), speech presentation and writing presentation can also be used to present summaries of longer stretches of speech and writing (discourse-domain summary), including whole discourses/texts;
(b) as all of the categories on the presentation scales appear in principle to be usable for presenting discourse-domain summary, we effectively need two discourse-domain summary presentation scales (speech and writing) in addition to the three proposition- domain discourse presentation scales (speech, writing and thought), as set out in Short and Semino (2004), with the minor modifications I have suggested in section 2 above;
(c) there is no equivalent discourse-domain thought presentation summary scale as the notion of thought summary does not make much sense – summary can only reasonably occur when an original is available to be summarized, something which is arguably impossible even when we present our own thoughts, let alone those of others;
(d) the establishment of the discourse-domain speech and writing summary presentation scales helps us more easily to identify and describe accurately a range of interesting presentation ambiguities.
18As I have suggested that the above invented examples from Leech and Short (1981) can be seen as IS discourse-domain summary, I will begin with a clear textual example of this category. Here, and from now on, where relevant I will bolden the stretch of text I am focusing on:
At other times the daughter, heart-stoppingly voluptuous in her tight Californian pants, would lead me by the hand through the ruined garden, to the last clump of still-rooted myrtles, then crouch, bare-kneed, and pull me down beside her, and demand to know my ideological convictions.
(Laurie Lee 1969 As I Walked Out One Midsummer Morning, 35)
19This example from an autobiography (which is another example of IS involving a non-finite clause) looks more like the presentation of a summary of what was said rather than of a single proposition, mainly because of the clash between the single-proposition structure and abstract lexis on the one hand, and, on the other hand, the schematic assumptions we have for emotionally- charged interactions. It is implausible that the young woman would have uttered just one, rather abstract, single-proposition sentence like ‘What are your ideological convictions?’ or ‘What are your political views?’. Even though the text is not fictional, in this case, because we do not have access to the original very private conversation referred to, a decision as to whether the presentation is of a discourse-domain summary or not can only be based on what is in the presenting text and the relevant schematic assumptions the reader brings to the text. In real life, it is sometimes possible in principle at least, to check a posterior discourse presentation against a recorded original, although in practice that often turns out not to be possible. The same is true of the following DS examples.
20When, in Short (1988) I discussed newspaper headlines like:
UGH! GET RID OF MY SQUINT
(The Sun, 21 June 1984)
You’ve given me a squint, said Maggie
(Daily Express, 21 June 1984)
21I pointed out that it was unlikely that Margaret Thatcher, the UK Prime Minister at the time, ever said what she was presented as saying in these DS headlines, and that, indeed, there was no contextual evidence in the ensuing articles for her having used the words presented. The implausibility here relates mainly to character and role. Mrs Thatcher typically spoke rather formally in public in any case, but when she was Prime Minister she also had a duty to keep her language formal to reflect her position. Effectively, then, as with the Laurie Lee example, our assessments of whether we have proposition-domain presentation or discourse-domain summary presentation will be based on textual clues in the presenting text and schematic assumptions related to situation, speaker role and so on.
22I went on in Short (1988) to consider whether the above examples might be speech summary, but concluded that the DS form weakened that interpretative possibility (even though I noted that IS could sometimes be used to present summaries of stretches discourse longer than one proposition) and came to the conclusion (followed up on, in more detail, in the proposals in Short, Semino and Wynne 2002) that faithfulness constraints varied depending on factors like genre (for example news reports in popular newspapers might be less faithful than those in serious papers) and textual position (for example that headlines might be allowed more faithfulness leeway than the main body of news reports). In other words, although I raised the possibility of speech presentation being used for discourse-domain summarizing purposes I did not really follow the idea through, something which I am beginning to do in this paper. If these examples are indeed DS discourse-domain speech summary, as I now believe, then the standard speech presentation faithfulness considerations do not apply, the only faithfulness constraint being that the wordings, whatever their style, represent a reasonable summary of what was said overall. Of course the standard proposition-domain presentation interpretation is still possible, leading to a possible reading ambiguity. Whether readers respond to the above headlines and equivalents as discourse-domain summaries or the presentation of particular propositions is an empirical issue, of course, which could be tested in future research.
23It is clear in the next, fiction, example that an NPSA discourse-domain summary interpretation of a part of the conversation makes most sense:
. . . one of these questions related to our manner of living, and the place where, because I had heard he had a great plantation in Virginia, and that he had talked of going to live there, and that he had talked of going to live there,
and I told him I did not care to be transported.
(Daniel Defoe 1906, The Fortunes and Misfortunes of the Famous Moll Flanders, 41)
24There is only one clause (no ‘reporting clause + reported clause structure’), the speech act (question) is specified and the presented string indicates that two topics were asked about, suggesting that more than one clause (and maybe even more than one turn) was uttered.
25In the extract below, a group of characters are discussing preparations for an expedition they intend to undertake:
And thus it was agreed. They would depart in the spring, to avoid the malarial menace of the later seasons. Each would require a portable bedstead, an air mattress and a pillow; they would take some Oxley's essence of ginger, some good opium, quinine and powders; a portable inkstand, a match-box and supply of German tinder; umbrellas against the sun and flannel belts to ward off cramps of the stomach during the night.
(Julian Barnes 1989, A History of the World in 10½ Chapters, p. 149)
26Arguably all of this extract is discourse-domain speech summary. The first sentence is NPSA. For a group of people all to agree, there must normally be more than one utterance of agreement, so the NPSA must be the summary of a number of contributions. The NPSA summary introduces a stretch of FIS, which again appears to be discourse-domain summary, this time of an extended stretch of interaction. The first of the two FIS sentences has a plural subject, again suggesting more than one speaker and so more than one conversational turn. This in turn suggests that the subsequent sentence, which is in effect a long list of the items that the group would need to take, is also a summary of an extended interaction among the participants about what they would need, probably with different individuals suggesting different items.
- 12 I am grateful to Chang Shuchen for pointing out this example to me.
Breathless, half-choking, she told the dreadful story.
(Katherine Mansfield, ‘The Garden Party)12
27I have classified this example as NPV, not NPSA, discourse-domain summary because the telling of a story (in this case Laura’s description of her encounter with the family of a working-class man who has just been killed in an accident) is unlikely to involve just one proposition and we cannot know what specific speech acts were used in the telling of the story. It would be even more clearly discourse-domain summary if Mansfield had added the topic of the story (e.g. ‘. . . she told the dreadful story of the dead man’). The above sentence could conceivably be NPSA summary if we assume that all of the sentences uttered were statements. This provides support for the theoretical likelihood that there can be inter-category ambiguities on the discourse-domain speech and writing summary scales as well as on the ‘standard’ discourse presentation scales.
28In Leech and Short (2007 [1981]: 10.1.3) I suggested that ‘Mr D’Arcy came from the pantry, fully swathed and buttoned, and in a repentant voice told them the history of his cold. Everyone gave him advice . . .’ was what I am now referring to as NPSA. A more accurate account, given the above, is that ‘Mr D’Arcy . . . told them the history of his cold’ is ambiguous between NPSA proposition-domain summary presentation and NPV discourse-domain summary presentation, and that ‘Everyone gave him advice’ is NPSA discourse-domain summary presentation.
29Given that I have now provided examples of speech presentation discourse-domain summary using each of the standard speech presentation categories, and that it is likely that examples can be found of each of the categories throughout writing summary too, I would like to propose a discourse-domain speech summary presentation scale and a discourse-domain writing summary presentation scale, to match the standard proposition-domain speech and writing presentation scales (where the subscript ‘s’ below indicates a discourse-domain summary interpretation):
Figure 2. Discourse-domain summary speech and writing presentation scales
|
Discourse-domain summary speech presentation
|
|
[N]
|
[NPS]
|
NPVs
|
NPSAs
|
ISs
|
FISs
|
DSs
|
|
Discourse-domain summary writing presentation
|
|
[N]
|
[NPW]
|
NPWs
|
NPWAs
|
IWs
|
FIWs
|
DWs
|
30The introduction of a set of discourse-domain summary presentation scales would allow us to be more accurate in our interpretative and analytical claims for the stretches of presented discourse under discussion. They would also enable us to describe more exactly the ambiguities and uncertainties that can occur between proposition-domain presentation and discourse-domain summary presentation. The NPSA/NPSAs and NPWA/NPWAs category pairs are quite likely to be ambiguous with one another, as the NPSA and NPWA presentational categories are prototypically associated with summary on both the proposition-domain and the discourse-domain scales.
31We have already seen in 3.4 that it is possible to have ambiguities both between (a) proposition-domain presentation categories which are adjacent on the cline and (b) proposition-domain presentation and discourse-domain summary presentation, at least in cases where the speech is presented relatively minimally. This, in turn, raises the possibility of similar sorts of ambiguities in relation to adjacent categories.
32What I have also come across, however, are some examples of presentations which effectively constitute (i) proposition-domain presentation and (ii) discourse- domain summary presentation at the same time. Consider the example below (I have numbered the sentences for ease of reference), which comes at the beginning of the garden party referred to in the title of the story, with Laura welcoming the guests as they arrive:
- 13 Also supplied to me by Chang Shuchen.
‘Darling Laura, how well you look!’ (1)
‘What a becoming hat, child!’ (2)
‘Laura, you look quite Spanish. (3) I’ve never seen you look so striking.’ (4)
And Laura, glowing, answered softly, ‘Have you had tea? (5) Won’t you have an ice? (6) The passion-fruit ices really are rather special.’ (7)
(Katherine Mansfield, ‘The Garden Party’)13
- 14 The story is a fiction, of course.
33(1), (2) and (3)–(4), because they are each contained within separate sets of inverted commas look like conversational openers produced by different people commenting on Laura’s appearance, They would thus seem to be DS proposition-domain presentations of the individual utterances of three different characters arriving at the party, with no matching individual response turns from Laura being provided. The DS of sentences (5)–(7), on the other hand, as they cohere together pragmatically and are all contained within one set of inverted commas, are apparently14 all excerpted from one of Laura’s responses to one of the visitors. Hence each DS example in (1)-(4), seen on its own, is traditional proposition-domain speech presentation. But they are clearly also representative parts of three separate interactions and, together, the three sentences of Laura’s presented speech also count, by inference, as the enaction the sort of response Laura would have made to all of her guests, including the three who produce sentences (1)-(4). So, overall we have what amounts to a quotative summary, which quotes representative parts of at least three conversational openings and one (part of) a representative reply, with the rest of the discourse omitted. In other words, the DS proposition-domain speech presentation is being used at the same time as (DS) discourse-domain summary presentation.
34This discussion in turn brings me to another example I now realize I did not get quite right in Style in Fiction:
Mr Shepherd hastened to assure him, that Admiral Croft was a very hale, hearty, well- looking man, a little weather-beaten to be sure, but not much; and quite the gentleman in all his notions and behaviour; ― not likely to make the smallest difficulty about terms; ― only wanted a comfortable home, and to get into it as soon as possible; ― knew he must pay for his convenience; ― knew what rent a ready-furnished house of that consequence might fetch; ― should not have been surprised if Sir Walter had asked more; ― had enquired about the manor; ― would be glad of the deputation, certainly, but made no great point of it; ― said he sometimes took out a gun, but never killed; ― quite the gentleman.
(Jane Austen, Persuasion Ch. 3, quoted in Leech and Short 2007 1981): 10.1.4)
- 15 I would like to thank Elena Pallarés-García for pointing out this example and sharing her disserta (...)
35I described this extract, correctly, I think, as FIS. But in the light of the DS example from the Katherine Mansfield sentence above, I think it is more accurately described as FIS quotative summary, as the dashes and elliptical syntax clearly suggest that we are being presented with excerpted snippets of a longer speech (and maybe even of a series of Mr Shepherd’s turns, with the contributions of Sir Walter and others omitted). The introduction of the kind of careful proposition-domain presentation and discourse-domain summary presentation analysis I have been arguing for in this paper thus helps us to characterize better the detailed effects of such examples. A similar Jane Austen example (which is also arguably ambiguous between FIS and FDS) is discussed in Pallarés-García (2008: 63), who refers to it as ‘an interesting mixture of quotation and summary’.15
36These examples are similar to what Fludernik (1993: 411) calls contraction, for which she provides a DS example, derived from Page (1988 [1973]: 32) and indeed my notion of summary shares some similarity with what she calls ‘condensed speech acts, again giving DS examples to illustrate what she means.
37This paper clearly builds on the work of others (e.g. Sternberg, Page, Tannen, Clark and Gerrig, and Fludernik) as well as my own, including some of my own earlier imprecisions. Analytical false steps are, of course, an inevitable, and indeed welcome, consequence of stylisticians’ attempts to be empirical and analytically and interpretatively precise. I would be pleased, of course, for others to help fill in the blanks I have referred to above and correct any mistakes, inaccuracies or gaps. Similarly, I am very interested in hearing from others about different kinds of discourse presentation ambiguities and uncertainties they have discovered. I would also like to suggest that empirical work is conducted on whether or not real readers arrived, while reading, at the kinds of discourse-domain summary interpretations I have suggested.
38Finally, I would suggest that discourse presentation analysts need also to spend some concentrated time on investigating the pragmatic processes involved in inferring whether a presentation is what I have called proposition- domain presentation or discourse-domain summary presentation (or both at the same time). As I have suggested in some of the discussion of individual examples above, the co-text may contain information to suggest that a discourse-domain summary is involved, the presentation itself may have relevant summary-suggestion features and we clearly use schematic knowledge of various kinds to infer that the presentation of what Ryan calls the Textual Actual World is summarized. How we perceive and respond to discourse presentation ambiguities and vaguenesses, both within and across the presentation scales, also merits inferential pragmatic investigation. Indeed, given that, to date, the definition of the discourse presentation categories has been dominated by structural considerations (e.g. syntax, lexis), it is arguable that the elephant in the room in discourse presentation theory and analysis is the relative weighting of formal, contextual and pragmatic factors when deciding upon categorizations, discourse-presentation types (e.g. proposition- domain presentation and discourse-domain summary presentation) and the effects associated with them.
