MASARYK UNIVERSITY
FACULTY OF EDUCATION
DEPARTMENT OF ENGLISH LANGUAGE AND LITERATURE
COMMENT CLAUSES AND THEIR ROLE IN SPOKEN ENGLISH
Diploma Thesis
Brno 2007
Supervised by: Written by:
PhDr. Renata Povolná, Ph.D. Mojmír Muzikant
DECLARATION
I hereby declare that I have worked on this diploma thesis on my own and that I have used only the sources listed in the references. I also give consent to deposit this thesis in the library at Masaryk University or the Information system of the Faculty of Education and to be made available for study purposes.
________________________
Mojmír Muzikant
ACKNOWLEDGMENTS
I would like to thank my supervisor, PhDr. Renata Povolná, Ph.D., who has helped me with the present thesis, especially for her proposals and valuable advice which I appreciated a lot.
Brno, 20th April 2007
Symbols used in the examples analysed
A, B, C, D speaker identity
a, b, c, d non-surreptitious speaker
*yes* simultaneous talk
(laughs) contextual comment
((yes)) incomprehensible words
yes . yes brief pause
yes – yes unit pause
Table of contents
1 INTRODUCTION... 6
2 PREVIOUS RESEARCH... 7
2.1 PUBLICATIONS ABOUT SPOKEN INTERACTION...7
2.2 GRAMMAR BOOKS...11
2.3 ARTICLES ON CCs...15
3 SPOKEN LANGUAGE... 25
3.1 WRITTEN vs. SPOKEN LANGUAGE...25
3.2 DIFFERENCES BETWEEN SPOKEN AND WRITTEN LANGUAGE...26
3.3 FUNCTIONS OF SPOKEN AND WRITTEN LANGUAGE...28
3.4 FACE-TO-FACE AND TELEPHONE CONVERSATION...28
4 MATERIAL UNDER INVESTIGATION... 30
5 CLASSIFICATION OF COMMENT CLAUSES... 32
6 CONVERSATION ANALYSIS... 47
6.1 FACE-TO-FACE CONVERSATIONS...47
6.2 TELEPHONE CONVERSATIONS...49
7 COMPARISON OF THE TEXTS ANALYSED... 51
8 CONCLUSIONS... 55
9 SUMMARY... 57
10 REFERENCES... 59
11 APPENDIX... 62
1 Introduction
The language is one of the means of communication among people. It helps them understand each other, share their knowledge, express their opinions and attitudes. What is more, it enables them to communicate things and phenomena they would never be capable of without a language.
I have chosen to examine in detail the language of face-to-face and telephone conversation because it is a variety of English that is very close to me. The fact is that the written language represents only a fraction of total language use. Geaney (1996: 26) states that 99 per cent of English use is speech. The paradox is that there was only a little research of everyday speech. Lexicographers still have a tendency to consider the occurrence of a word in print a chief or sole criterion for its inclusion in the dictionary.
Grammarians rarely venture beyond the safe confines of the sentence, a unit that is of doubtful value in the description of casual speech (Svartvik 1980: 167). The fact is that people use spoken language every day – in all kinds of conversations, discussions, quarrels, etc.
For these reasons I have decided to contribute to the study of spoken language. To be more specific, I have chosen as the main topic of my work one of the phenomena of the spoken language – comment clauses. Comment clauses are typical only in the spoken language, so we can barely find them in the written language. According to Concise Oxford Dictionary of Linguistics, comment clauses are ‘used to cover clauses seen as expressing a comment, which is in meaning, at least, parenthetical. Clauses like I think in Nothing, I think, happened or as you know in As you know, nothing happened.’ Another definition is offered by Oxford Dictionary of English Grammar, which says that ‘comment clause is a parenthetical clause, only loosely connected with the rest of the sentence. Comment clauses may be finite or non-finite, and include many clichés and conversation fillers, e.g. you know, generally speaking, you see, as I said or to be frank’.
The topic of my thesis is: Comment Clauses and Their Role in Spoken English. As stated above, the reasons are following: even though there is quite a lot of information about comment clauses in literature – the information is not systematic and each book deals with them only peripherally. Another reason is my interest in spoken language.
I hope this thesis will help with further investigation in the area of spoken language.
2 Previous research
In this chapter I would like to outline the research that has been done so far concerning comment clauses (CCs). The books and authors I am going to mention are quite important for my research and are recommended for further study. I have divided this chapter into three parts – one summarizes what has been written about comment clauses in monographs or publications dealing with spoken interaction, next one summarizes grammar books and finally the third part examines various articles written on this particular topic. It is also necessary to point out the fact that I took into consideration only comment clauses you know, you see and I mean. These three CCs are the most frequent and most stereotyped in the private conversation. The justification for this selection can be found in Chapter 5.
2.1 Publications about spoken interaction
Crystal and Davy (1969: 49) in their Investigating English Style do not present comment clauses but they talk about the so/called ‘clauses of parenthetic type’ – these clauses may be embedded in the main clause, or may occur in sequence with it. They also present an example: you know, that’s my sort of knitting.
In another part of the book the authors notice that words tend to be very simple in structure and therefore there is an avoidance of specialised terms and formal phraseology, and whenever they occur, their force is usually played down by the speaker, through the use of hesitation, or the use of you know, sort of, and so on (ibid.:
114).
Stenström (1994: 16) in her Introduction to Spoken Interaction speaks about interactional signals and discourse markers. The difference is that interactional signals constitute turns of their own or they link turns together but discourse markers are used as organizers and turnholders within the turn, or they are used as boundary markers. In the following extract all the underlined items belong to one of these categories:
A: ... I suppose if you got experience in American university administration you could still come back here.
B: Hm, oh, yes, certainly...well, they’re desperate for people to work in universities cos the money’s not good, so anyone...
A: Oh really, well, you know, oh, here, yeah, hm..
B: so anybody who’s soft enough to enjoy it and sort of...
If all the discourse markers are removed, what is left there would make perfect sense. So why do we actually use these markers? Stenström thinks that without them, the conversation would be much less lively and less ‘personal’ (ibid.: 17). She also indicates that interactional signals are used to start, carry on and terminate the conversation – they can either appeal for feedback or give feedback, they respond, they involve the listener in the conversation (e.g. you know), and so on (ibid.: 61). Next, she specifies discourse markers – they are used to organize and hold the turn and to mark boundaries in the discourse. They help the speaker organize the discourse – they may serve to introduce and mark the end of a topic, to introduce a digression and mark the resumption of the old topic and they can signal the end of a conversation (ibid.: 63).
As for CCs, Stenström (1994) indicates several functions which they can have in discourse:
• They can work as empathizers to engage the listener and make her/him feel part of the conversation. They often appear at the beginning and end of a turn, but also elsewhere, for instance when the speaker appeals for feedback (ibid.: 64). Example:
He’s not a relaxed lecturer but he’s a driving lecturer, you know – whereas some of them here stand up poor dears and they haven’t the first clue – they’re so nervous, you know, painful to listen to...
From the above example it can be clearly seen that if you know is removed, the facts are the same, but the effect of the utterance is not the same. What was said would have sounded more a matter of fact and probably less friendly, and the listener might have felt less involved.
• Another function is appealer. It appears in turn-final position and serves as an explicit signal to the listener that some kind of feedback would be appropriate. The “prompting force” of appealers varies from fairly weak, e.g.
when realized by you know, to fairly strong, e.g. when realized by OK (ibid.:
79). Example:
A: ...and Blundell is a rather cosy old film man, you know.
B: Yes.
• Here is another function that Stenström lists, namely inform marker. Both you see and you know can serve as inform markers, but they do it differently.
You see is typically used when A assumes that the information is new to B.
You know can be used in the same way but is more often used when B is either assumed to be somewhat familiar with the subject matter already or when A wants to create the impression that A and B share a common ground. What is more, you know1 is often used to hint at some underlying message (ibid.: 90). Example:
A: ...people come in at odd times sometimes, for instance this girl who’s working at the BBC, you see, she has all different shift hours...
B: Oh, Hart’s got this thing about status, you know, he – he doesn’t like secretaries to be merely secretaries... (shortened)
• You know, you see, I mean (and sometimes sort of) might function as verbal fillers – they are used when speaker has a difficulty in formulating the message; they have no exact meaning or purpose (ibid.: 129). Example:
I mean, she’s so little, I mean, you know, sort of one can imagine a sort of middle-aged woman with a coat that seemed, you know, sort of, just slightly exaggerated her form, you know, I mean, she could sort of slip things inside pockets...
• Finally, the comment clause I mean can work as a monitoring device. By monitoring the speaker can put things right. Sometimes the speaker needs to make a new start or rephrase what s/he was going to say in the middle of a turn, often because the listener shows that s/he cannot follow or is not convinced. In such situations the monitor I mean comes in handy (ibid.:
131). Example:
A: Have you tried at all so far, I mean, have you got round to anything?
B: No, I haven’t, I mean, I’ve done nothing except, you know, bring up this family since I left school.
1 you know is generally pronounced in a separate tone unit with varying intonation contours;
you see is more often part of a tone unit, the other items vary (Stenström 1994)
Aimer (1996) in Conversational Routines in English devotes the whole chapter to discourse markers. The title of the chapter is ‘Discourse markers as conversational routines’. She characterizes them as ‘routinized elements which contribute to the coherence of discourse in various ways’ (ibid.: 200). Further, they have the function of commenting on and organizing the message. They can be either speaker and/or hearer oriented and their function is metalinguistic (checking or controlling the communication channel) rather than referential. She recognizes their grammatical function as adverbials. As for CCs she mentions only I mean which she describes as a signal that the following unit of talk is an explanation or clarification of what has been said. It also can make the assertion less intrusive and finally, using I mean might be ‘playing for time’ under the pressure of on-line processing (ibid.: 211).
She sums up that the overriding function of discourse markers is to integrate utterances into the flow of conversation and to instruct the hearer how their interpretation is affected by the context (ibid.: 210).
Edmondson (1981: 153-156) in his book Spoken Discourse – A Model for Analysis calls clauses like you know, you see and I mean the so-called fumbles stating that ‘in performing communicative acts speakers hesitate, pause, cannot find the right word, and so on. Fumbles are conventionalized ways of plugging such potential gaps’. He distinguishes five classes of fumbles – starters, let-me-explains, underscores, cajolers and asides. The clauses you know and you see are the commonest tokens of cajolers.
Cajolers are hearer-oriented and they are used by a speaker as an appeal for understanding (compare with Stenström 1994) and they can occur in initial, medial and final position.
As for the clause I mean, Edmondson (ibid.) states that it is the most commonly occurring token of a let-me-explain, which tends to be used ‘to communicate the fact that I’m trying to communicate’. Therefore it is speaker-oriented and precedes the performance of a communicative act.
2.2 Grammar books
Comment clauses are examined in detail by Quirk et al. (1985: 1112-1118) in A Comprehensive Grammar of the English Language. They are defined as
‘parenthetical disjuncts’ which can occur initially, finally or medially and therefore have a separate tone unit. They can function either as content disjuncts to express the speakers’ comments on the content of the matrix clause, or as style disjuncts to convey the speakers’ views on the way they are speaking. It is possible to distinguish several types of CCs:
1) like the matrix clause of a main clause e.g. I believe, I hope, you know, you see
2) like an adverbial finite clause (introduced by as) e.g. as you know, as I say
3) like a nominal relative clause
e.g. what is more important, what was more upsetting 4) to-infinitive clause as style disjunct
e.g. to be honest, to be fair 5) -ing clause as style disjunct
e.g. speaking openly, speaking frankly 6) -ed clause as style disjunct
e.g. stated bluntly
The first syntactic type of comment clauses is the most important one. This can be proved by the following table:
Texts Total
Syntactic type of CCs
S. 1.3 S. 1.5 S. 1.8 No. %
like the matrix of a main clause 101 69 97 267 96
like an adverbial finite clause 4 3 2 9 3
like a nominal relative clause 0 0 0 0 0
to-infinitive 1 0 1 2 1
-ing 0 0 0 0 0
-ed 0 0 0 0 0
Total (No.) 106 72 100 278 100
Table 1: Occurrence of comment clauses in three texts (each contains 5,000 words) taken from A Corpus of English Conversation (according to Povolná 2003: 72).
Furthermore, the most typical syntactic type of CCs may have various semantic functions:
• they function as ‘hedges’, which means they express the speaker’s tentativeness over the truth value of the matrix clause. The subject is usually I and the verb is in the simple present, e.g. I think, I expect, I presume, I assume, etc.
• they can express the speaker’s certainty. Again, the subject is I and the verb is in the present simple, e.g. I know, I see, I admit, I have no doubt, etc.
• they express the speaker’s emotional attitude towards the content of the matrix clause. Subject is again I and the verb is in the present simple, e.g. I regret, I’m glad to say, I fear, I wish, etc.
• the last semantic function is used to claim the hearer’s attention. Some also call for the hearer’s agreement and, at the same time, they express the speaker’s informality and warmth towards the hearer. The subject is usually you or the implied you of the imperative, e.g. you know, you see, you realize, etc.
The second type of comment clauses (functioning like an adverbial finite clause) is introduced by as. It serves one of two syntactic functions in these clauses: as a relative or as a subordinator.
• In its relative function, as introduces a type of sentential relative clause that may precede or be inserted in its antecedent, in this case the clause or sentence to which it is attached. In its mobility, this as-clause is intermediate between the relative and adverbial constructions. Like the sentential relative which, as may function as relative pronoun, e.g. as you know, as you say, as I can see, as you may have heard, etc.
• As a subordinator, as introduces a clearly adverbial clause, and the sentential antecedent is replaced by it. It is roughly synonymous with insofar as, e.g. as it appears, as it happens, as it transpired, as I see it, etc.
The other types are marginal and therefore can be dealt with more briefly. The third type of comment clauses is introduced by what. The what-clause must be initial, e.g.
What’s more surprising, he kept the money for himself. Other examples might be what’s most significant of all, what’s very strange, etc. The remaining types of comment clauses are style disjuncts. They are non-finite clauses, thus differentiated by the form.
Leech and Svartvik (1994: 266) in their grammar book A Communicative Grammar of English define CCs as clauses which ‘comment on the truth of the sentence, the manner of saying it, or the attitude of the speaker (an emotional reaction or judgment)’
(ibid.). They state that CCs are not very much related to the rest of the main clause they belong to, they function as sentence adverbials. Usually they are marked off from the other clause by commas (in written language) or they have separate tone unit (in speech). They can occur at the beginning, in the middle, or at the end of a sentence.
Finally they name some types of CCs like I see, I think, I suppose, I’m afraid, as you see, as I said, to be frank, so to say, so to speak and state that some such items are very common as ‘discourse markers’, particularly you see, you know, I mean and I think.
Biber at al. (1999) in Longman Grammar of Spoken and Written Language examine CCs in various parts of the book. They include CCs under more general term
‘discourse markers’ and state that they are commonly used in conversation and that they are typically used to signal the pragmatic or discoursal role of the speaker’s utterance, dynamically shaping it to the ongoing exchange (ibid.: 1046). They exemplify this statement:
Yeah, well, I can understand you know, I mean…
Some other (similar) functions are following: to signal a transition in the evolving progress of conversation and to signal an interactive relationship between speaker, hearer, and message (ibid.: 1086).
The formulaic clauses I mean and you know are also called items that are ‘loosely attached to the clause and connected with ongoing interaction’ (ibid.: 1046).
Concerning the syntactic function of comment clauses (I mean, you know), they mostly look like a main clause, with the following clause functioning as a complement clause. But because these comment clauses can occur not only in initial position, there is often no clear way of choosing between the main clause analysis and the comment clause analysis for construction in this initial position (ibid.: 1076).
Generally, all the expressions (I mean, you see, you know) are interactive and cohesive. I mean can signal that a clarification is going to follow, whereas you see and see signal that what follows is an explanation of what has preceded. They also keep the same interactive function not depending whether they occur in initial, medial or final position (ibid.: 1078).
Huddleston (2002: 774) in The Cambridge Grammar of the English Language does not mention CCs or discourse markers at all. He only mentions ‘purpose adjuncts’
which can relate to the speech act or just give information about the situation described in the clause. For example To cut the long story short, he accepted her offer and left the country. Nevertheless, these adjuncts are not an object of this thesis.
Swan (1995: 151) in his Practical English Usage also does not mention CCs at all but talks about discourse markers. He describes them as words and expressions that are used to show how discourse is constructed. They can show the connection between what a speaker is saying and what has already been said or what is going to be said;
they can help to make clear the structure of what is being said; they can indicate what speakers think about what they are saying or what others have said. He also adds that there is a very large number of these discourse markers, and it is impossible to give a complete list on a few pages and therefore he mentions only a few of the most common examples. Among many categories of discourse markers there are two which should be mentioned. These are discourse markers for:
• making things clear; giving details – e.g. I mean, actually, that is to say, in other words. He claims that we use I mean when we are going to make things clearer, or give more details.
• softening and correcting – e.g. I think, I suppose, I mean, actually, etc. These can have various functions. I think/feel/reckon/guess are used to make opinions and statements sound less dogmatic – they suggest that the speaker is just giving a personal opinion, with which other people might disagree. I suppose can be used to enquire politely about something (respectfully inviting an affirmative answer).
2.3 Articles on CCs
Schiffrin (2001) in her article called ‘Discourse markers: language, meaning, and context’ defines discourse markers as “sequentially dependent elements that bracket units of talk, i.e. nonobligatory utterance-initial items that function in relation to ongoing talk and text. They can be considered as a set of linguistic expressions comprised of members of word classes as varied as conjunctions (e.g. and, but, or), interjections (oh), adverbs (now, then) and lexicalized phrases like y’know and I mean”
(ibid.: 57). She further deals with y’know stating that it is often to be found in specific discourse environments, it can conclude the argument, introduce a story preface, evoke a new referent (ibid.: 66). She concludes that “discourse markers tell us not only about the linguistic properties (e.g. semantic and pragmatic meanings, source, functions) of a set of frequently used expressions, and the organization of social interactions and situations in which they are used, but also about the cognitive expressive, social, and textual competence of those who use them (ibid.: 67).
In the article called ‘Discourse tags’ Stenström (1984) talks about the so-called D-items (discourse items), saying they have two characteristics in common: they are difficult, sometimes impossible to analyse in syntactic terms, and they contribute very little, if anything, to the propositional content of an utterance. She analysed eleven face- to-face, unplanned conversations and one preplanned monologue presented to an audience, all texts being taken from London-Lund Corpus. She identified several discourse categories like apologies (pardon, sorry), smooth-overs (don’t worry, never mind), hedges (kind of, sort of), expletives (damn, gosh, hell), greetings (hi, hello), initiators (anyway, however), etc. She included CCs into a category of ‘softeners’ and includes here the following items: I mean, mind you, you know, you see, as you know and do you see. She states that these softeners can serve as ‘communicative cues’, i.e.
explicit appeals for confirmation or agreement – especially if they carry a rising tone.
They are therefore typically turn-yielding and interpersonally oriented.
Stenström also points out the difference between you know in the dialogue and as you know in the monologue. As you know signals that the speaker presupposes that the
listeners know what he is referring to and therefore acts as a politeness device and it does not appeal for feedback, whereas you know does not necessarily presuppose previous knowledge on the part of the listener but may ‘pretend’ to do so for the sake of politeness (ibid.: 77).
Stenström (1995) analyses in detail CCs in her article ‘Some remarks on comment clauses’. She claims that CCs like I think, I mean, you know and you see are quite difficult to describe in grammatical and semantic terms alone because they simply depend on the context for their interpretation. It is also clear that they appear much more often in spoken language rather than in written language, which can be proved by the occurrence in spoken London-Lund Corpus (know, think, mean and see rank as high as 15, 24, 44 and 45) in comparison with written Lancaster-Oslo-Bergen corpora, where these expressions do not occur at all. Furthermore, verbs like mean, know, see and think are in spoken language typically used in comment clauses but they are mainly used as ordinary transitive verbs followed by an object in written language.
She mentions five main characteristics of CCs:
• they contain a transitive verb
• they resemble matrix clauses but lack complementation
• they are generally syntactically dependent
• they tend to be stereotyped
• they have a number of ‘semantic functions’ (ibid.: 291).
Next, she claims that CCs can occur in initial, medial and final position. They also occur in a separate tone unit, with the exception of I think and I mean, which often have no tone at all. There also appears to be a clear difference between I-oriented and you-oriented CCs concerning tonicity – you know and you see were quite high up the scale, whereas I think and I mean were not. This reflects a tendency on the part of the speaker to minimise attention to the self in favour of the listener. Worth mentioning is also the syntactic function of CCs – they can serve as comment clauses or matrix clauses. Stenström discovered that you see and you know served almost exclusively as a comment clause, on the other hand, I think was much more used as a matrix clause
followed by a that-clause. Concerning I mean it was not so clear – it is possibly a matrix but rather a comment clause (ibid.).
Pragmatic functions of CCs are influenced by:
• the inherent semantic content of the verbs
• position in the sentence and prosody
• the entire context of situation (ibid.: 294).
Stenström concludes the article by stating that CCs should have a specific discourse category together with other discourse phenomena, such as formulae and interjections, instead of being forced into a traditional grammatical category (ibid.: 299).
Östman (1982) in his article ‘The symbiotic relationship between pragmatic particles and impromptu speech’ analyses in detail the relationship between pragmatic particles (like you know, I guess) and impromptu speech, which is spontaneous, every- day face-to-face interaction. He examines pragmatic particles from different points of view – structure of language and discourse-functional. From the structural point of view a pragmatic particle would be short and prosodically subordinated to another word, it would resist clear lexical specification and be propositionally empty and it would tend to occur in some sense cut off from, or on a higher level than the rest of the utterance, at the same time as it tends to modify that utterance as a whole (ibid.: 149). The functional point of view examines how the language relates to other areas of human behaviour.
It is important to mention his degrees of impromptuness of a discourse which are visibility, reciprocity (interchangeability of roles), informality, spontaneity (unplanned), empathy (backchannel responses) and insonsequentiality (no import beyond immediate context of discourse). The following table shows the occurrence of the above mentioned degrees.
oral-dyad telephone lecture letter prose
visibility + - + - -
reciprocity + + - + -
informality + + - +/- -
spontaneity + + - - -
empathy + + + - -
inconsequentiality + + + +/- -
Table 2 Degree of impromptuness of a discourse (according to Östman 1982: 158).
It can be seen from the table that the more “pluses” there are, the more impromptu the discourse will be.
Probably the most complex analysis of CCs is offered by Erman (1986) in his article ‘Some pragmatic expressions in English conversation’. He claims that these phrases do not contribute much to the informational content of an utterance, but nevertheless perform several valuable functions in the speech situation. He also mentions various names that have been given to these expressions in literature, e.g.
verbal fillers, void pragmatic connectives, softeners, pause-fillers, hesitation-markers, discourse markers, pragmatic particles, etc. Erman adds another label – pragmatic expressions (PE’s). He justifies this selection by saying that he concentrates on certain linguistic expressions and the objective of his study is their use. He analyses PE’s from many different aspects (position in the utterance, syntactic environment, phonological and prosodic properties, etc.). He examined twelve 5,000-word texts, which is around 60,000 words altogether. Let me now present some of his conclusions.
Concerning the position in the utterance or turn, it was no surprise that most of the three PE’s were in medial position. Almost 83 per cent of them occurred in this position. The table below shows the total number of occurrences and percentages in initial, medial and final position in twelve conversations.
Occurrences Initial Medial Final Total
number 57 540 51 648
% 9 % 83% 8 %
Table 3 Position of CCs in the texts analysed (according to Erman 1986: 132)
Next he focuses on PE’s as connective elements. Besides usual clauses as main clause (MC) and subordinate clause (SC) he also distinguished the so-called minor sentence (MiS), which is a sentence connected by means of a PE. The following table displays the syntactic pairing of clause-types for each PE.
PE MC-MC MC-SC MC-MiS Others Total
N % N % N % N %
you know 48 52 % 17 18 % 10 11 % 18 19 % 93
you see 39 61 % 9 14 % - - 16 25 % 64
I mean 59 68 % 4 5 % 6 7 % 18 20 % 87
Total 146 60 % 30 12 % 16 7 % 52 21 % 244
Table 4 Syntactic pairing of clause-types (according to Erman 1986: 133)
As can be seen from the table the predominant type of clause-pair is MC to MC for all three PE’s.
Another aspect of Erman’s research is the semantic relationship between propositions connected by means of PE’s. He claims that two neighbouring clauses may be grammatically unlinked, which does not mean there is no connection of meaning between then – it means, rather, that the connection is implicit. In informal speech it is often the case that clauses are juxtaposed (next to each other) without explicit cohesive signals. Therefore the semantic relationship is not shown by any conjunction or an adverbial, but has to be inferred from the content and context of the discourse. He then analyses one by one each pragmatic expression.
• you know – a causal relationship between the clauses can be seen although the causality is not manifested by a logical marker:
...Tim would know you know we ought to get him down here...
This sentence could be paraphrased like this:
...Tim would know and therefore we ought to get him down here...
• you see – in the middle of an utterance you see is used between clauses (as a connective element2) rather than between constituents (as and intrusive element).
In the following example the speaker amplifies the information given in the first proposition by telling exactly when the finalists stopped attending seminars:
...but the thing is for instance they’ve stopped you see you know they did eighteenth and nineteenth century and stopped...
2 When a PE is used in the middle of an utterance it can either occur between or within constituents – then it is called intrusive element (e.g. Ingeborg works with her you know on nineteenth century...) or between clauses – then it is called connective element (e.g. Tim would know you know we ought to get him down here).
• I mean – this PE requires two arguments, which means that it necessarily both refers back to a previous proposition and points forward to an upcoming argument. In most of the examples from the texts examined the speaker in the second clause takes back, explains or gives support to an argument brought out in the previous clause (ibid.: 137). It can be clearly illustrated by the following example:
...but I’m sure it’ll take me all my time to do it in three weeks I mean I’ve seen what it’s been like for you...
Another aspect is phonological and prosodic properties of PE’s as connective elements. Erman claims that all the three PE’s have different phonological properties depending upon their role in the tone unit. For example, if we take you know as a connective element, then know carries nucleus with rising tone. Nevertheless, in half of the examples you know also formed a separate tone unit. On the other hand, you see often appears as ‘nuclear tail’ and in this position clearly belongs to the preceding prosodic syntagm, which it finishes off. This is one of the reasons why the link between clauses connected by you see in tail position is less obvious. And finally I mean that appears in all the examples in prenuclear position either in preonset position or with onset on mean (ibid.: 144).
The last aspect of PE’s are their pragmatic functions. Erman here summarises the main functions of PE’s as connective elements. He thinks that the main function is that of organizing the discourse, PE’s are used in the structuring of the text and of the information and therefore appear at strategic points in the production of discourse (ibid.).
Povolná (2005a) in her article ‘On some interactive discourse items in English telephone conversation’ analyses six texts taken from the London-Lund Corpus. The research focuses on some features typical of spoken language, especially authentic telephone conversation. It is important to realize that telephone conversation is quite unique because the involved participants are not visible to each other. All the conversations examined were recorded surreptitiously, i.e. without the prior knowledge of the speakers that their conversation was being recorded. It is necessary to realize that there are three ways the current speaker can appeal to the current hearer to produce
some kind of reaction: declarative questions, question tags and comment clauses (Urbanová 2002: 17). Povolná focuses on two of them, namely question tags and comment clauses you know and you see. She states that it is important to distinguish two types of verbal reactions – CCs can prompt from the listener a proper verbal response or a backchannel, which does not involve any shift of the speaker and is just the hearers’
signal that they are still paying attention. Thus, discourse items can prompt three types of reactions:
• a verbal response (a shift of speaker is necessary)
• a backchannel signal (no shift of speaker)
• no verbal reaction at all
Concerning question tags, there are basically two types:
• tags with a rising tone (for verification)
• tags with a falling tone (for confirmation)
The occurrence of discourse items in telephone and face-to-face conversation summarises the following table:
Text types Telephone conversation Face-to-face conversation
Discourse item No. % No. %
You know 67 62.5 102 49.8
You see 15 14 49 23.9
Question tags 25 23.4 54 26.3
Total (No.) 107 100 205 100
Table 5 Interactive discourse items in telephone and face-to-face conversation (according to Povolná 2005a: 144).
It is evident that the number of interactive discourse items is relatively low in telephone conversation – only 107 cases, in comparison with private face-to-face conversation with 205 cases.
As for reactions to the use of interactive discourse items, the most frequent type of response in telephone conversation is a proper verbal response (46 per cent), followed by no verbal reaction (38 per cent) and a backchannel (16 per cent). In face-to-face conversation, the situation is different – the majority of discourse items prompt no verbal reaction (41 per cent), followed by verbal response (36 per cent) and a backchannel (23 per cent).
It can now be elicited that in telephone conversation the number of all the three interactive discourse items is lower with any kind of reaction. The conclusion of the research shows that the use of non-verbal reactions on their own, such as head-nods or facial gestures, on the part of the current hearer can often represent sufficient support for the current speaker to go on speaking because they are in face-to-face contact.
However, when such contact is not available, which is the case of telephone conversation, the current hearer tends to produce much more frequently some kind of feedback in the form of a verbal response rather than a backchannel signal (ibid.: 151).
‘Clausal forms you know, you see, I mean, I think etc. in the language of public radio discussions’ is another article by Povolná (2004). She compares the usage of these clausal forms in public radio discussions with face-to-face private conversations.
She states that the language of public radio discussions is different in many aspects from that used in authentic face-to-face conversations. The main difference is that public radio discussions are planned in advance and the discussed topics are highly predictable, which results in better preparations of participants’ contributions before the actual beginning of a given radio discussion. Another difference is the length of speakers’ turns. In public radio discussions turns are usually longer than those in face- to-face conversations because they are usually not interrupted by the other participants.
Next, there is less cooperation between all interlocutors and there is hardly any simultaneous speech present in public radio discussions because all of the participants know that they will be addressed one after another by the main speaker, which is usually the radio reporter. Speakers also have to be as explicit as possible because the public cannot ask any questions, when they misunderstand something (ibid.).
As the public radio discussion is not the genre that will be dealt with in detail, the results are shortened. They show that a much higher number of CCs is used in conversational texts (277 occurrences) than in radio discussions (only 103 occurrences) in spite of the fact that the length of the two texts examined is identical. The conclusion is that participants in face-to-face private conversations tend to use comment clauses much more frequently than participants in public radio discussions (ibid.).
Povolná (2005b) tries to contribute to the study of some interactive discourse items, specifically the clausal forms you know, you see, and I mean in her article ‘On the role of some interactive D-items in different genres of spoken English’. These discourse items are studied with regard to what they prompt from the current hearer, which can be a verbal response, a backchannel signal or no verbal reaction at all. The examined material consists of nine texts taken from the London-Lund Corpus, namely three private face-to-face conversation, three telephone conversations and three public radio discussions. These genres of spoken English have been chosen for the analysis because they have both similar and different characteristics which can have some influence on the role of the D-items under investigation (ibid.: 170). The conclusions of the research can be seen in the following tables:
Text types Face-to-face conversation
Telephone conversation
Public radio discussions
Discourse item No. No. No.
You know 102 67 14
You see 49 15 9
I mean 37 34 26
Total (No.) 188 116 49
Table 6 Occurrence of D-items in spoken English (according to Povolná 2005b: 171).
The above table shows that public radio discussions tend to be the least interactive genre of all.
Reaction Total
you know Response Backchannel No verbal r. No.
Appealer 8, 1, - 1, 2, - 0, 0, - 9, 3, -
Inform marker 15, 11, 1 16, 3, 0 34, 14, 10 65, 28, 11
Empathizer 5, 10, 1 14, 7, 0 8, 16, 2 27, 33, 3
Monitor 0, 1, - 0, 1, - 1, 1, - 1, 3, -
Total 28, 23, 2 31, 13, 0 43, 31, 12 102, 67, 14
you see Response Backchannel No verbal r. No.
Appealer 3,1, - 0, 1, - 0, 0, - 3, 2, -
Inform marker 4, 1, 1 4, 1, 0 25, 6, 6 33, 8, 7
Empathizer 2, 2, 0 4, 2, 1 6, 1, 0 12, 5, 1
Monitor 0, -, 0 0, -, 0 1, -, 1 1, -, 1
Total 9, 4, 1 8, 4, 1 32, 7, 7 49, 15, 9
I mean Response Backchannel No verbal r. No.
Total 7, 21, 3 14, 9, 4 16, 4, 19 37, 34, 26
Table 7 Reactions to you know, you see and I mean in face-to-face conversation, telephone conversation and public radio discussions – (according to Povolná 2005b: 178ff), slightly simplified
This table shows the overall results. It is evident that speakers in face-to-face conversation often produce a verbal response to the use of any of the three D-items analysed, it tends to be a backchannel signal rather than a verbal response because such a signal is usually sufficient in these kinds of situations. In telephone conversations, a verbal response rather than a backchannel signal is a more common kind of reaction to the use of D-items because of the lack of face-to-face contact between speakers. As for the public radio discussions the number of all D-items is very low here and what is more, if any D-item occurs, there is hardly ever any kind of verbal reaction on the part of the current hearer (ibid.: 180).
Povolná concludes the research by stating that ‘the kind of reaction the individual interactive D-items prompt depends on their particular discourse function and above all on the overall character of the speech situation in the given genre’ (ibid.: 180).
3 Spoken language
3.1 Written vs. spoken language
According to Geaney (1996: 24) 99 per cent of English use is speech. Why is it so?
The reason for this might be that people like conveying information among themselves especially with friends and relatives they feel comfortable with. Another reason might be that conversation can be about nothing in particular, whereas written language usually conveys some information.
The above mentioned Geaney’s assertion is further supported by Leech (1982: 134) who claims that in the history of the human race, spoken language definitely came before writing and that there is no clear evidence of the existence of a writing system of any kind before about 3500 BC, whereas it is presupposed that spoken language existed well before then. Also in the history of several societies, spoken language appeared earlier than written language, and many languages do not have any written form. All in all, children also learn to speak before they learn to write. On the other hand, Leech (ibid.) adds that written language is quite important from an educational and social point of view. Our society is based on the advantage that writing gives it. He concludes that we cannot simply say that written language is superior to spoken language and vice versa because they are simply different.
Leech and Svartvik (1994: 12) talk about two channels in English: speech and writing. Each of them has a different transmission system. Speech is transmitted by sound-waves, which originate in speaking and are received in hearing. Writing, on the other hand, is transmitted by letters and other visible marks, produced in writing and received in reading. They also state that spoken and written English do not have different grammars – they are only used differently on these two channels.
3.2 Differences between spoken and written language
Written language is different from spoken language mainly because of the absence of such features as intonation, stress and gestures. In written language more stress is put on selection of vocabulary, syntax and punctuation. Written language is also more planned and organized. Another characteristic might be self-sufficiency, which means that the writer somehow takes into account that the reader will be far away from him (the writer can be also dead) and therefore everything that is necessary for understanding of the written text must be inside that particular document. This does not apply for spoken language – when there is something not understood, it is quite simple to ask for the information again (Geaney 1996).
Impersonality is also typical of written language. A piece of writing can be composed with no specific reader in mind; therefore the writer has to be absolutely sure that it will be understood by everybody, which is achieved by using an impersonal style.
There will not be any personal reference such as personal pronouns I and you – these are used mainly in spoken language. Another feature is lexical density, which means, that in written English there are much more content words (words that bear lexical meaning like nouns, verbs and adjectives) than grammatical or function words (words that have only grammatical meaning like determiners, prepositions, pronouns and auxiliary verbs). From this reason the information load of written language is higher.
Furthermore, the vocabulary of written language is much more extensive than that of spoken language (ibid.).
Spoken language, on the other hand, is typical for its inexplicitness, which means that it is not necessary to express every single piece of information by speech because parts of the information are conveyed via ‘body language’ (facial expressions, gestures), shared knowledge3 of the participants and the opportunity of feedback. There is also the lack of clear sentence boundaries. In speech there is the absence of clearly defined units.
In written language there are sentences which begin with a capital letter and end with
3 Also known as ‘mutual knowledge’ is the knowledge that a speaker and addressee are seen as jointly taking for granted at a given point in an interchange. E.g. if A says to B Jones is coming, one condition for successful communication is that A and B should share the knowledge of who Jones is (Matthews 1997).
a full stop, but in spontaneous speech it is often very difficult to recognize the beginning and end of a sentence. Concerning the structure of sentences, speech is much more simple than writing mainly because of the lack of planning – there is simply no time for making complex and flowery sentences. Another feature of spoken language is repetitiveness. Important information has to be repeated because the addressee cannot refer back to what has gone before. It is especially noticeable in television advertisements. Speech is also characterized by normal non-fluency. This non-fluency is caused by several phenomena: hesitation, unintended repetitions (e.g. you, you...), slips of tongue, overlapping or simultaneous speech, false starts (breaking of a sentence in the middle as a result of a change of mind; e.g. You should really – I don’t think it’s a good idea), fillers (e.g. er, uhm), grammatical blends (sentence begins in one way and ends in another; e.g. Would you mind telling me what’s the time? – this sentence begins as an indirect question, but ends as a direct question) and unfinished sentences. The most important feature of spoken language for the present research are monitoring and interaction features. These appear in speech in face-to-face conversation rather than in monologue. Monitoring features indicate the speaker’s awareness of the addressee’s presence and reactions (e.g. well, I mean, sort of, you know) and interaction features invite the active participation of the addressee (questions, imperatives, etc.). The last characteristic of spoken language is informality because speech is generally less formal then writing (Leech 1982). The table below summarises the differences between spoken and written language.
Spoken language Written language
Inexplicitness Explicitness
Lack of clear sentence boundaries Clear sentence boundaries
Simple structure More complex structure
Repetitiveness Non-repetitiveness
Normal non-fluency Fluency
Monitoring features No monitoring features
Interaction features No interaction features
Informality Formality
Table 8 Differences between spoken and written language (according to Leech 1982: 139).
3.3 Functions of spoken and written language
The main advantage of written language is its relative permanence, which allows keeping records of anything. Next, it is possible to use it for communication over a great distance (e.g. by letters, newspapers) and address large audience simultaneously (e.g. by books and all kinds of publications). Nevertheless, relatively recent inventions (radio, tape-recorder, telephone) helped to overcome the limitations of spoken language concerning time, distance and numbers. Unlike speech, written language is not only permanent but also visible. It can be carefully planned and revised by the writer. Parts of writing can be omitted or read again. On the other hand, speech enables quick and direct communication with immediate feedback from the addressee. Speech has also a socializing function; it integrates us into the society, it is an everyday activity for everyone and everywhere. All in all, both speech and writing are complementary in function; it is necessary to use them both appropriately (Leech et al. 1982).
3.4 Face-to-face and telephone conversation
Telephone conversation is quite unique because it is a type of a conversation in which the participants do not see each other. This lack of visual contact results in certain differences between this kind of conversation and usual face-to-face conversation (Crystal & Davy 1969). The following table summarizes the main differences between face-to-face and telephone conversation:
face-to-face conversation telephone conversation face-to-face contact lack of face-to-face contact
visible feedback lack of visible feedback
implicitness explicitness
lack of planning partly planned
topics unpredictable topics partly predictable varying length of turns shorter turns
pauses hardly any pauses
simultaneous speech rare simultaneous speech
Table 9 Differences between face-to-face and telephone conversation (according to Povolná 2005b: 171), slightly modified
The above table shows that the telephone conversation requires a much higher level of explicitness than face-to-face communication. For example, it is not possible to “rely on the extralinguistic context to resolve ambiguities in speech; visual feedback being absent, auditory cues become all-important, and in view of the diminished quality of the voice over the telephone, there develops a greater uncertainty and confusion in maintaining the ‘give and take’ of the dialogue” (Crystal & Davy 1969: 119). What is more, speakers cannot see the facial reactions of their hearers and therefore have to ask them frequently whether they have understood everything what has been said or not (Povolná 2005b). Another problem is the quality of voice while talking via telephone – the quality is quite low; moreover, there are many distortions of certain sounds.
There is also a tendency to avoid long utterances without introducing pauses. These pauses allow the listener to confirm his interest (e.g. by backchannelling) and his
‘presence’ in conversation. This backchannelling (e.g. by words like hm, yes, I see, yeah) is very important in telephone conversation because without it, the speaker may think that the connection was lost(Crystal & Davy 1969). There is also another reason for avoiding long utterances and obvious inclination to speak rather in shorter turns – usually the speaker has to pay every single minute of a telephone conversation and thus s/he wants to say as much as possible in the shortest period of time.
4 Material under investigation
For my analysis I have used texts from London-Lund Corpus (for text classification, see Svartvik 1990), namely two face-to-face conversations (S.1.4 and S.1.13) and two telephone conversations (S.8.1 and S.8.3). Each of these texts consists of 5,000 words and they represent spontaneous conversation among educated people. The speakers are British and the recordings were made surreptitiously, i.e. without the prior knowledge of being recorded (with some exceptions that will be mentioned later).
These two genres of spoken English have been chosen because they have similar and different characteristics (see chapter 3.4 above) – both of them represent spoken interaction and have the form of a dialogue, but they differ in the lack versus presence of face-to-face contact between participants.
The first text (S.1.4) is a face-to-face conversation recorded in 1969 and includes these two participants:
A: male academic, age: 48 B: male academic, age: 48
Both men were recorded surreptitiously and the conversation deals with many topics. It was quite difficult to guess what they are talking about without knowing the context;
nevertheless they are probably teachers of ancient languages (Greek, Latin) because they talk about reading Aristophanes and Seneca. They intercut this conversation by the preparation of coffee and the amount of sugar lumps they want to put in. Then they talk about paintings by various artists and their description and they end the dialogue about teaching in general.
The second text (S.1.13) is also a face-to-face conversation recorded in 1975 and the participants are:
a: female academic, age: 40
B: male retired charity commissioner, age: 61 C: female housewife, age: 60 (wife of B)
The lower-case a indicates that this participant knew about the recording and her task was to keep conversation going. Moreover, her contributions to the conversation were not included in the total 5,000 words of the text. Concerning the content of
conversation, the female academic gradually asks about the B’s and C’s relatives. The first relative is a woman-dentist and they discuss in general the position of women- dentists in society. Next relative is ‘poor old Ben’, who suffers from schizophrenia.
Another topic is prices of petrol in Great Britain followed by prices of heating oil.
Among other topics I can name for example the standard of living in GB, real estates, cousins, etc.
The third text (S.8.1) is an array of telephone conversations from 1975. This array consists of 16 telephone conversations and there are various participants – e.g. a broker, a university lecturer, a secretary, an estate agent, a sewing machine salesman, a clerk, etc. Let me now describe some of the telephone conversations. The first one is between a male estate agent and a female university lecturer – the lecturer wants to buy a house and she lays down the conditions (price, size, place, etc.). Another conversation is between a sewing machine salesman and a female university researcher. The researcher probably wants to buy a sewing machine but she cannot decide on the brand and type, so she is constantly asking the salesman to recommend her some for a good value.
The last text (S.8.3) is again an array of 13 telephone conversations recorded in 1975 as well. The participants are similar to S.8.1, there are again secretaries, university lecturers, officials, an electrician, university researchers, etc. One of the conversations is between a female university lecturer and a male mortgage consultant and they are discussing the terms under which it is possible to take a mortgage on a house. Another conversation is between a male electrician and a male research assistant. The assistant has some problems with his tape recorder and wants to have it fixed.
5 Classification of comment clauses
Taking into consideration the classification of CCs by individual authors mentioned above, this part of the thesis tries to study different functions of CCs according to these authors with slight modifications.
First of all, let me justify the selection of the term ‘comment clauses’. As explained in the theoretical part of my thesis there are many titles used for expressions like you know, you see and I mean. Erman (1986) calls them pragmatic expressions, Stenström (1994) interactional signals, Quirk et al. (1981) defines them as parenthetical disjuncts.
Other authors’ terms used in literature are e.g. verbal fillers, void pragmatic connectives, softeners, pause-fillers, hesitation-markers, discourse items, fumbles, cajolers, let-me-explains and pragmatic particles. On the other hand, the term ‘comment clause’ has been used by authors like Quirk and Greenbaum (1973), Leech and Svartvik (1994), Crystal (1995) and Biber et al. (1999). There is also a general term ‘discourse markers’ used for example by Swan (1980) but this term is too broad and includes all kinds of adverbial phrases used e.g. for structuring (firstly, secondly), adding (moreover, furthermore), generalising (in general), contradicting (on the contrary) and many others. Also other terms mentioned earlier comprise other categories than just what can be labelled comment clauses. Therefore the term ‘comment clause’ seems the most appropriate for the present research.
Secondly, in my research I have included only the comment clauses you know, you see and I mean. This is because they are the most frequent and most stereotyped in spoken language. There are of course other types of CCs like I suppose, I think, I believe, etc., but these are not so frequent and occur only marginally in the texts analysed.
Another problem worth explaining is also a situation about the comment clause I mean. For example Quirk et al. (1985) do not include I mean into comment clauses but rather speak about it as a device for mistake editing used in order to correct a phonological or semantic mistake (which is very common for informal speech).
Edmondson (1981) considers I mean as the so-called let-me-explain which
‘communicate the fact that I’m trying to communicate’. Nevertheless, I mean seems to
share enough features with you know and you see (e.g. they all take a direct object in the form of a nominal that-clause, they all belong to the “private” type of factual verbs and they help the smooth flow of conversation) that it is possible to include it into the category of comment clauses. Erman (1986: 137) also adds that I mean is the only comment clause that requires two arguments, i.e. it necessarily both refers back to a previous proposition and points forward to an upcoming argument.
I have analysed CCc according to three different criteria and these are pragmatic (semantic) functions of CCs, their position within the turn and the type of response they prompt from a current hearer.
Concerning the pragmatic functions, I have used the classification by Stenström (1994). She recognizes five semantic functions of CCs, namely empathizers, appealers, inform markers, verbal fillers and monitors.
EMPATHIZERS – they draw the listener in a conversation. They make him/her feel part of the conversation. They occur usually at the beginning or end of a turn, but they can occur in the middle of a turn as well. By empathizing the speaker also intensifies the relationship with the listener. However, the speaker does not usually wait for the listener to react but goes on talking without paying attention to the listener’s appreciative comment (Stenström 1994). It is also necessary to point out that as I was obviously not present in the conversations analysed, it was impossible to include silent feedbacks like head-nods or facial gestures. According to Quirk et al. (1985) empathizers can also ask for listener’s understanding. By using them, the speaker invites the listener to take an active part in conversation. They frequently prompt some feedback in the form of a backchannel signal or they do not prompt any immediate verbal reaction at all.
APPEALERS – these CCs usually appear in the final position of a turn. They serve as an explicit signal to the listener that some kind of feedback would be appropriate.
The appealer is strongest when uttered after a silent pause, where it has a questioning effect. Because of this questioning effect they mostly occur in a separate tone unit with a rising tone.
INFORM MARKERS – concerning inform markers Stenström (1984) states that they are used when the speaker assumes that the information is new to the listener (this is often achieved by the use of you see), or the listener is somewhat familiar with the subject matter and the speaker wants to make impression on the listener that they both
share a common ground. Povolná (2003) adds that CCs as inform markers can also indicate to the listener that he/she should pay attention either to some completely new piece of information or its new aspect.
VERBAL FILLERS – CCs functioning as verbal fillers are used when the speaker has problems in formulating the message. They come in handy when the speaker takes turn and is not prepared properly. Furthermore, they show that the speaker wants to say something but needs more time to put it into words. They usually occur at the beginning of the turn, where the rough planning of the entire utterance takes place (ibid.) and they have no exact meaning or purpose.
MONITORS / ERROR CORRECTIONS - this function is mainly realized by the comment clause I mean. With the help of I mean the speaker can start the sentence again, rephrase it or put things right. Schiffrin (1987) adds that I mean can also expand the ideas or explain the intentions. And, as already mentioned, I mean requires two arguments – one left behind and one upcoming, for this reason I mean mostly occurs in medial position.
The next classification depends on the position of a comment clause in a turn. On this occasion, let me explain the turntaking system. The turntaking system consists of three basic strategies – taking the turn, holding the turn and yielding the turn. These strategies seem very clear where one person speaks at a time, while the other person patiently waits for her or his turn. But in reality, it is not that simple. The listener may
‘jump in’ the conversation without waiting for the current speaker to finish, which results in simultaneous speech and interruptions, or the speaker may forget what s/he wanted to say, which results in unwanted silence if the listener is not prepared to fill the gap. Turntaking also presumes a shift of speakers. An utterance that is produced while the other participant of conversation goes on speaking cannot be regarded as a turn.
Turn in everything the current speaker says before the next speaker takes over. Thus, backchannels cannot be considered as proper turns but they are important for the smooth flow of communication because they indicate to the current speaker to go on speaking. Turns can be either quite short consisting only of a single words or they can be also very long and resemble short monologues (Stenström 1995). In agreement with Erman (1986) and Stenström (1995) there can be three positions in the turn – at the
beginning of the turn (initial), within the turn (medial) and at the end of the turn (final).
Position in the turn greatly influences the pragmatic functions of CCs.
The last criterion for classification concerns the type of response to the use of CCs.
As Biber et al. (1999: 1091) state, ‘it might be supposed that, in general, statements (which impart rather than elicit information) do not need a response from the hearer.
However, in practice responses to statements are very frequent. Functionally, these inserts can be classified as backchannels because of their role in signalling feedback to the speaker that the message is being understood and accepted. Given the interactive nature of conversation, backchannels are important in indicating that speaker and hearer are keeping in touch with one another, and that the communication is still in progress’.
It is important to distinguish a proper turn from a backchannel because CCs can prompt from the current hearer: verbal response, backchannel signal or no verbal reaction at all (Povolná 2005b). Again, it was impossible to conclude from the texts analysed how often the participants prompt some silent feedback in the form of head-nods or facial gestures; therefore the current hearer’s non-verbal reactions are treated together with no verbal reaction at all.
Quirk et al. (see chapter 2.2) also deal with the function of CCs in the structure of a sentence – syntactic function. They distinguish type 1 CCs, where they look like the matrix clause of a main clause, type 2, where CCs function like an adverbial finite clause (introduced by as) and other marginal types (nominal relative clause, to-infinitive clause as style disjunct, -ing clause as style disjunct, -ed clause as style disjunct). As my research covers only syntactic type 1, I do not deal with other syntactic types; CCs functioning as adverbial finite clauses occurred only four times in the present research, namely as you know twice and as you see twice as well. As stated above, it is possible to distinguish six syntactic types – first three types (matrix clause of a main clause, adverbial finite clause and nominal relative clause) function as content disjuncts, which means that they express the speakers’ comments on the content of the matrix clause and the other three types function as style disjuncts, i.e. they convey the speakers’ views on the way they are speaking.
Let me now exemplify the classification with concrete examples taken from the London-Lund Corpus.
PRAGMATIC FUNCTIONS
1) EMPATHIZER – CCs that function as empathizers engage the listener and make him/her feel part of the conversation, without them the listener may feel less involved in conversation.
Example 14:
A: you say this is who is *this Velasquez*
B: *it’s a copy* of a Velasquez that and I don’t . copies are not such fun I mean it
*((syll))*
A: *but* still
A: I don’t know whether I want to live with it though
B: I don’t think I would want to live with a copy ((which)) you know it’s like the original and not like it from what I remember I’ve seen . you know coloured reproductions of the original in art book I know it is Innocent the fourth - I’m sure because I ((you know)) I’ve seen the portrait in lectures on Velasquez ((4 sylls illustrate))
(S.1.4.586-608)
In the first example two male academics are talking about art, specifically about some painting by Velasquez. Speaker B in the last turn is quite enthusiastic about the originals and with the help of the comment clause you know wants the listener A to be involved in conversation and wants A to listen to him. In this case, you know definitely functions as an empathizer.
Example 2:
B: Pope Innocent the fourth – ((a)) seventeenth century pope . ((of)) some . some ilk or other --- ((about 18 sylls untranslatable)) I don’t know Leslie’s views and I said to him you know one of the things that’d ((it)) seems to me it would be convenient if we ((could)) all . if we could you know set more or less agree together as ((to)) when we stop lecturing this term *cos*
A: *if* we could all ((what))
(S.1.4.611-622)
In Example 2 the same male academics as in Example 1 continue in their conversation. In this case, both tokens of you know function again as empathizers. It can
4 The transcription of examples has been simplified. Symbols showing the stress or the intonation have been left out. The list of symbols can be found at the beginning of the thesis.
