3. INFORMATION STRUCTURING: FROM THE FIRST STUDIES
3.3. After the Prague school: sentence, text, discourse
3.3.2. Spoken and written language
Oral and written language have had very different statuses regarding their use and usability in studies that deal with Information Structuring. The first studies that have been made reference to in this thesis deal with written language, but as oral language becomes more salient as an object of investigation by itself, the attention turns to oral language and to the possibilities it offers to linguistic investigation. When it became clear that the traditional morphosyntactic approaches are not always convenient to study the phenomena of orality, then the question arises of opposing these two codes by assigning to them different investigation methods: should these two domains be kept separetely because of their admittedly different objects?
Among others, French syntacticians C. Blanche-Benveniste & C. Jeanjean (1987) have deplored the fact that these two variants of the language have been placed in opposition as two completely different codes, with the idea that one of them is the ideal form and the other one is necessarily incorrect. P. Linell could also be mentioned, having elaborated quite an extensive critique called The written language bias in linguistics, in which he proposes alternative views on all the problems that come from the formalistic, objectivising approach to language (Linell 2005). Ayres-Bennett & Carruthers (2001: 4) observe that both codes contain a large variety of subtypes and registers which do not necessarily correspond to the general idea of oral language being more informal and written code being more formal (there can be very formal uses of oral language and,
similarly, completely informal uses of written language), so that the general opposition between the two is not justified.
In the present study the problem of opposing or separating oral language from written language is not a central one: the main analysis is carried out on examples of oral language, which I consider as primary when investigating real language use, and secondly also takes as a point of departure the assumption that detached constructions are above all a construction typical to oral speech. I agree rather with J.-M. Adam (2004: 38), who considers these two objects (oral and written language) as two aspects of the same language and where the huge differences come, on the one hand from the difference between the genres, and on the other hand from the differences between the conditions of interaction in these two cases.
When spoken and written languages have been compared, there are several specificities that have been pointed out: the constructions used in spoken and in written language are in some measure different. However, it is not correct to argue that spoken language uses systematically more simple constructions, or that it is a sort of simplified written language due to the fact that speakers want always to make only minimal effort. The main linguistic characteristics of spoken language can be summed up as follows (Miller & Fernandez-Vest 2006: 13):
- information is carefully staged, a small quantity of information being assigned to each phrase and clause;
- spontaneous spoken language typically has far less grammatical subordination than written language and much more coordination or simple parataxis;
- the syntax of spontaneous spoken language is, in general, fragmented and unintegrated; phrases are less complex than phrases of written language; the clausal constructions are less complex;
- the range of vocabulary in spontaneous language is not as large as in written language;
- a number of constructions occur in spontaneous spoken language but not in written language, and vice versa.
These properties are in a large proportion made possible by the special conditions in which spoken language occurs and by its intrinsic properties:
availability of context and extra-linguistic information in the situation, the use of gestures and other non-verbal means and the use of pitch, amplitude, rhythm etc., which help to disambiguate and guide the information flow in the right direction.
There are also approaches that underline some other aspects typical to spon-taneous oral language, for example Chafe 1982 distinguishes between contex-tualized speech with strong interpersonal involvement (spontaneous speech) and decontextualized speech (more formal uses) which shows the opposite features.
The contextualized language is characterized by the following properties:
- concreteness and imageability (use of details)
42 - use of 1st- and 2nd-person pronouns - people and their relations highlighted
- actions and agents more emphasized than states and objects - feelings and thoughts (evaluation) reported
- hedged signals used
- feedback signals and repairs used when needed
Information Structuring as an inherent property of any sensed production can consequently be analysed equally in written language as well as in oral productions by exploring to some extent different markers, for example referring to the prosody in oral language.
Different approaches can be found in this domain: there are studies which rely mostly on the evidence of written language, by using terms and definitions that have been worked out on the basis of written language data and by using traditional morphosyntactic categories (these are, for the most part of the studies, dealing with the sentence level), but seem not to exclude the study of e oral language, by just adding some typical features to it. An example of such an investigation is the recent collection The expression of information structure (Krifka & Musan 2012) which unfortunately does not go further in this direction: the examples are in most cases created by the authors and there are not many new ideas in the light of recent analysis of the spoken language.
The question arises as to whether it is possible to use the terms and relevant phenomena which have been described, first in relation to written language, for example as concerns frames etc. and to try to apply them to oral language bearing in mind that these two language uses have certain divergent features.
Many influential studies still deal mainly with written language or self-created examples or just very short oral sentences, because it becomes very complicated to take into consideration full oral paragraphs, which means that there are necessarily more unclear and questionable sequences. However it was con-sidered here that although the textual/discursive approach has in this context some weaknesses, it should be attempted to find a way to first get to the data of the oral language only and to consider what it is possible to find out by working on this type of corpus. Of course, as the linguistic material (oral or written) shares a certain amount of common features and is undoubtedly organized following universal underlying principles, it is not necessary to oppose them on all levels, but to try in some way to push into the background the evidences that we already have concerning the written language.
M. M. J. Fernandez-Vest (1994: 144–158, 2009: 194–198) has shown in which manner the features of oral language are modified when an oral com-munication (given based on written notes, though) is to be ‘translated’ into a normative written text. As for the frequency of detached constructions, the comparison given in Fernandez-Vest 2009: 196 as an example of the occurrence of detached constructions in interviews and their written and published transcriptions speaks for itself: in the written version, no final detachments remained and only 5% of Themes are detached.
It is generally admitted that the choices concerning transcription and editing principles depend on the objectives of the research: the transcription should be on the one hand as informative as possible and at the same time usable by the researcher in his or her particular field of study (the principles that have been followed in the present research are delineated in sections 1.2 and 1.3).
Sometimes, the features of oral language are to some extent also preserved in transcriptions used in more normative registers of language; in these cases one can question the particular effects this creates.
Such an interrogation could be raised about the next example, which shows that the principle of economy is not always followed in transcriptions of oral language; the excerpt comes from a partial transcription of a video broadcast on the public TV channel during the main news programme. The person interviewed is Estonian cross-country skier Andrus Veerpalu after his discharge at the Court of Arbitration for Sport. It should be noted that this affair was highly controversial and generated passionate debates over two years in Estonian society. The interview itself is somewhat longer, the interviewer asks about ten questions, but for illustration a short excerpt has been selected.
Q : Kas olete kahe aasta jooksul endale selgeks teinud, kust kohast [tuli] siis see tõdemus, et Andrus Veerpalu on dopingupatune?
V : No... seda on nagu raske öelda, kust see... See on piiripealne, kui see test ei ole veel oma õigeid piire kätte saanud, siis sportlased ongi suht sellised...
teistsugused ja kui neid asju pole nii täpselt uuritud, ega sealt võibki mingi viga sisse tulla.
Q : Andrus Veerpalusse mitteuskujad on kogu aeg öelnud, et vaidlustatakse ju testi, seda, kas test on pädev, mitte seda, kas on tarvitatud dopingut või mitte.
Kuidas nendele vastate?
V : No... eks igaüks sai nagu... ise mõelda ja olla, mis ta tahtis, ega mina saanud kellelegi mõtteid pähe panna.
Q: ‘Have you been able to understand after the last two years, where the conclusion that Andrus Veerpalu has used drugs came from?’
V: ‘Well, it is, like, difficult to say, from where it… It is a borderline, when this test has not yet reached its limits, sportsmen are indeed, like, … different and when these things have not been investigated so thoroughly, then some mistakes can indeed be made.’
Q: ‘Those who didn’t believe in Andrus Veerpalu have said all the time that it is the test that’s being contested, in order to find out whether the test is efficient, not the fact of whether drugs have been used or not. How would you respond to those?’
V: ‘Well, then everyone could, like, ... think and believe, whatever they wanted, I could not put ideas into anyone’s heads.’
What is the role of this type of short written summary? Should it give an overview to those who do not (cannot) watch the video? Should it generate the reader’s interest towards the online video?
When we compare the summary with the original interview, we find naturally that many features typical to oral language are edited (word order, repetitions, repairs etc), which is predictable, but the question is to what purpose are the numerous incoherencies proper to spontaneous language use preserved?
This sportsman is known as someone who is not at ease with public appearances, so it could be another illustration of this weakness, corresponding also to the expectations of the public or intentionally presented as such. The original interview shows that the answer to the second question is more developed after this hesitant take-off, however the editor chose not to transcribe the whole response, so that it seems here that the oral language devices used in sensitive matters can act as rather discrediting means.
In her analysis about the initial detachments in French, B. Barnes (1985: 114) put forward the hypothesis, based on the distinction of foreground and background in speech and on the suggestions of Lambrecht (1984), that detachments are a feature of foregrounded parts of the discourse – the back-grounded parts display the characteristics of spontaneous or unplanned use of language less. The definitions of ‘foregrounded’ and ‘backgrounded’ portions come from Hopper and Thompson’s (1980: 280) analysis of narratives (written language): foregrounded parts represent the main storyline, while backgrounded portions assist, comment or amplify the foregrounded parts.8
The detachment constructions are almost constantly characterized as primarily oral devices and this is presented in different works as a cross-linguistic universal, so that as a rule this assumption seems not to be questioned by researchers working on different languages. However, McLaughlin (2011) proposes a more subtle approach to detachment constructions in French and tries to find out whether the general idea, according to which the detachment constructions in written texts are used primarily in order to give an effect of orality, can be supported by the evidence found in the corpus. She explores the occurrences of detachments in three types of language use: spoken, journalistic and literary (fiction novels) and comes to the conclusion that detached constructions are used in written texts to create different effects, not only to represent orality: on the one hand, they certainly can be used in order to imitate orality (with other ‘mimetic features’ present such as discourse particles etc.), but on the other hand, in many cases the surrounding features typical to written
8 Lambrecht makes use of them in order to distinguish between the uses of lexical subject NPs and detached NPs.
code without any characteristics of oral speech suggest that the detached element also has other functions, related to the inner structuring of the text (stylistic effects), simple establishment of the topic and others (McLaughlin 2011: 225–226). However, the author does not explore more specifically the other possible functions of detachments in her article.
In this chapter an insight was proposed into the complex problematics of relations between oral and written language uses in connexion with detached constructions and we saw that the claim about the orality of these constructions in all uses should not always be taken as an immutable principle, although a consensus seems to be established about their primarily oral nature.
3.3.3. Mental processes linked to the treatment of the referents