'Given the text, we construct the situation from it' (Halliday and Hasan. 1985:
38). If the 'we' is taken, for argument's sake, to include the 'discourse analyst',
in what ways might we attempt to reconstruct the context of situation from a
text? A sample of CMC text, specifically, a multilogue taken from a BBS (bulletin
board service) will be examined using three different approaches in order to
illustrate the possible strengths and weaknesses in each analytical stance.
The first will look at features of lexicogrammar which may construe the context
of situation in terms of the interpersonal and textual metafunctions. In order
to play with the concept of the relationship between lexis, text and context,
the text was altered by removing lexical items that were deemed to give away
the topic of conversation, so that the field of discourse was unable to be retrieved.
Formatting inherent in the mode of interaction was also stripped from the contributions
to create a text similar to that of 'normal conversation'. Given these constraints,
how far is it possible to reconstruct the context of situation with what remains
of the text? One area associated with defining the channel and medium of interaction
will be discussed: the significance of the technological channel in re-constructing
the situation, especially when looking at the continuum writtenness/spokenness,
and the evident degree of process sharing.
The second approach will look at the usefulness of the basic structural model
of conversation as outlined by John Sinclair (in Coulthard (ed)1992: 86-88),
and a preliminary double tagging of the contributions is suggested which takes
into account both the illocutionary and perlocutionary force of each 'utterance'
in the context of both the preceding and the following contributions, thus adding
a dynamic dimension to the model which may be lacking in a purely lexicogrammatical
view. A third area of discussion will attempt to show some implications of Grice's
notion of the 'co-operative principle', dependent on assumed extra-linguistic
knowledge, in accounting for participant responses in the context of the logogenetic
development of the text. It is suggested that this mode of interaction may promote
deliberate 'flouting' of the maxims of conversation in order to position other
participants in a way that ensures responses.
In analyzing texts generated by two or more participants in CMC, it is contended
that several approaches can help analysts link texts to their context of culture
and provide means of interpreting, as well as making descriptions of the sociohistorical
practices underlying most conversation, including those generated in a purely
textual environment.
-------
"Given the text, we construct the situation from it." (Halliday, 1985: 38)
It can be seen that, without the content-full lexical items, it is difficult
to determine what the participants are talking about: lexis has a relationship
to experience, so it seems that the actual field is not available for retrieval.
However, the texture of the text is still available if the removed lexical items
are tagged for reiteration. For example, the ostensible topic is "A and B",
but after contribution 5) the topic switches to L(1&2 - indicate that this
lexical item is comprised of two words). What that is, does not seem to matter
for purposes of interpreting aspects of tenor and mode.
In looking at the tenor, or indicators of the interpersonal metafunctional aspects
of the text, the fundamental area of examination is the mood of the contributions.
The first move by S1 will show that what passes for a question as topic opener,
functions as a negative evaluation of the topic to be discussed, and the mood
is actually declarative rather than interrogative. The structure 'What has ---
got to do with -----' is almost by convention recognised as a negative evaluation
of such a relationship.
In the next question in the sequence which is made by S1 at 5), the modality
evidenced by the word 'can' seems to indicate a genuine request for permission
to do something - or to be more precise - to have something done to him. However,
one of the content words (here represented by 'L') which has been deleted from
this question actually marks a new topic (which continues for the rest of this
text), and seems to be evidence that S1's attitudes toward the discussion, and
the other participants, may be one of power-seeking: this participant seeks
to 'police' the topic boundaries through introducing a somewhat 'irrelevant'
response in this instance.
The second question (by S2 at 2) in the sequence functions as a reply to S1's
ostensible opener, and tries to resolve the negative evaluation displayed. However,
S1 undermines the move completely (at 3), by referring to the participant as
'the man', while at the same time focusing attention not on the idea being discussed,
but the person making the comment. Halliday and Hasan (1976: 280) have pointed
to this function of the class 'general nouns' as carrying a connotation of familiarity,
and thus realizing part of the interpersonal metafunction by highlighting a
participant/speaker's attitude. (Indeed, the content word deleted from this
utterance (3) is also one of negative evaluation).
Further on in the text, another question is made by S5 which seems to be seeking
information (10a) but is actually an offer. She asks the participants at large
to respond, not to a request, but to an offer. In this manner, this participant's
offer could be seen as an attempt to draw the other participants together, as
well as deflect their attention to her role in the group as perhaps 'thoughtful
helper'. It thus functions on the interpersonal level of the text, as a proposal,
rather than a proposition. S5's offer is made concerning the topic introduced
by S1 earlier (at 5), and in this case S1 quickly responds supportively with
an contribution in the imperative, in effect, 'telling' her what to do, thus
underlining his own role as self-appointed 'leader'. S5, although not a native
speaker of English, as evidenced by other 'grammatical' errors in her contributions,
nevertheless has a firm grasp of what is going on at the interpersonal level:
she names him as 'Mr --W--' (an epithet concerning his marital status: Mr Newly
Wed) displaying her own attitude of familiarity towards her interlocutor, and
her rejection of his position as leader, or at least, his position with respect
to her. A modal expression of certainty further emphasises her position in the
use of 'Of course I will '.
There are other features of S1's attitude towards his own role, most notably
in his rejection of another's contribution at 7) where he responds 'Nonsense'
to a comment by S3 whose tentative comment is signaled by the use of modal 'could'.
Yet another participant whose words show an attitude of 'superiority' is evidenced
at 9) where S4 announces that others have discovered something which he presumably
already knew. S4 then goes on to make a contribution in the imperative, as if
to emphasise this role.
From the preceding analysis, it may be observed that in some cases, at least
the tenor of a text may be easily grasped by an outsider having little information
regarding the shared knowledge of the participants beforehand, and with almost
complete ignorance of the field of discourse. A more detailed grammatical analysis
of mood and modal features of the text would of course reveal such patterns
of tenor more clearly, but here my main point has to highlight the interpersonal
metafunctional aspects of what was originally a completely text-based interaction,
in other words, a form of conversation in the written mode.
The experiential and textual metafunctions are closely linked, however, so,
to continue the thought experiment a little further, does this mean that texture
is lost in such an altered text? Hasan (1985: 94) goes so far as to say that
the cohesive harmony of a text depends on the harmony between these two metafunctions.
So that, if most of the lexical items which are concerned with the field of
discourse and thus help to realize the experiential metafunction, are unavailable,
perhaps the textual metafunction cannot be adequately described.
A brief consideration of lexical density will show that a large proportion of
the words concerned with subject matter, deleted from the text, were reiterations,
so the lexical density is low. According to Ure's (1971 cited in) work on lexical
density, lower lexical density is a feature of spoken text, so in the matter
of medium one might be inclined to retrieve a spoken context of situation, even
when presented with the text in the form shown above. Even a quite liberal calculation
of lexical density in the case of this text reveals a percentage of around 20.
As for collocation, there seems to be plenty, but mostly between the lexically
'full' words that have been removed. One intuitive example might be the three
word set at 12a) - 'add some more', but this forms a very local example of the
way the text hangs together at the clause level, rather than at the overall
textual cohesion level. Reference and general nouns are a feature of the text,
however. The main ones seem to be endophoric references : 'one of the many',
'the man', 'the very person', 'itself', 'one', '(that's)what'; while a couple
rely on referents outside the text: 'here', 'people in here', 'out there'. Merely
enumerating these does not provide satisfaction, because it tells only about
patterns within the text itself, but fails to reveal anything substantial about
the context of situation, except that it could be said that these people seem
to exist in a place separate from 'out there'. In other words, it gives a very
small clue to the 'material context of situation' as distinct from the textual
(Hasan.op cit: 99). The removed items are tagged so that the patterning of the
text can still be viewed, and it appears to maintain some type of cohesion,
even though coherence may not seem its greatest feature in its present form.
Apart from lexical density, what other textual clues as to the nature of medium
can be pinpointed? Montgomery (1986: 108-111) outlines a number of features
of spontaneous speech. The first of these is pauses, especially within the 'turn'.
It seems relevant here to mention the significance of pause in the relationship
between speech functions, such as offers and invitations, and rejections or
refusals which are marked by pauses, as well as various kinds of structural
complexity (Levinson. 1983: 307). The text in question manifests few of these
types of markers, except for the curious inclusion of 'trailing off' indicators
(...) at the end of sentences, which, I believe is actually one of the most
significant features of texts of this kind, and indicates that the text has
been constructed in the written medium. (A comparison with other texts made
in the same medium and channel would serve to underscore this feature, and would
show similar examples of what I like to call 'mode-bleeding'.) There are also
no 'ah's, 'erm's or 'um's. Furthermore, there are no 'incomplete' sentences,
no interruptions or overlapping comments from other participants, no interjections
or 'expressions of attention', such as 'mmmm', 'yeah' or 'that's right', and
only one of what Montgomery (op cit; 110) calls 'markers of sympathetic circularity',
where the speaker seeks the continued attention of the interlocutor by appealing
to shared understanding, and which are realized by expressions such as 'you
know', 'sort of' and 'and that'. This again would indicate, without other markers,
that the text was created in the written medium.
Is it possible then, to determine from the text itself, what are the channel
and medium of interaction? It could be said that the interaction was not carried
out in the phonic channel, or that visual, immediate contact was not available,
due to the absence of some of the features of 'process sharing', most particularly
those pertaining to interruptions, overlappings, or incomplete or grammatically
complex sentences mentioned above.
Hasan (1985: 58) distinguishes between channel and medium by saying that while
the channel may be phonic or graphic, the medium is located on a continuum between
spoken and written modalities. She makes the interesting point that the concept
of 'process-sharing' is closely linked with the channel, as it is said to relate
to the degree to which participants can be said to share in the creation of
the text. Here, Hasan relates the channel of communication to the concept of
process-sharing:
..the physical presence of the addressee impinges on the textual
processes in a way that the writer's own awareness of the
needs of the addressee can hardly ever do: for one thing, in the
phonic channel both the speaker and the addressee hear (and
often see) the same thing at the same time. This is obviously
not possible when the channel is graphic.
It seems obvious that a high degree of process-sharing would be evident in the
phonic channel, and that this would also positively correlate with the spoken
medium. Telephone conversations, for example, are conducted in the phonic channel
and the spoken medium, but without any face to face cues. Nevertheless, in the
case of telephone interaction, the degree of process-sharing can be seen to
be quite high, so that textual indicators, even within a short one-person transcript
of a telephone conversation, should be able to alert any would-be analyst to
the medium and channel of interaction.
In this case the problem is to discover what indicators there are in the text
which show in what manner the text was actually created or produced as what
now becomes an object of study - as data. Because the text seems to be comprised
of grammatically simple, yet lexically light short turns, it has some similarity
to conversation, with a higher than normal degree of process sharing than would
normally be expected in the graphic channel - yet obviously has been created
in the written medium - one which is written as if spoken - one where the graphic
channel and the written medium do not result in an interaction so low in 'process
sharing' as might be imagined, and a textual realization which resembles conversation
in many ways, yet lacks obvious features expected of it. Indeed, it might be
traditionally supposed that the graphic channel and the written medium would
preclude any interaction that would result in anything resembling conversation
due to the concomitant modal factor of low contact - a deferral in time and
space - between participants.
However, the actual mode of interaction here is complicated by its technological
mediation, and therefore, although the context of situation as a set of discourse
features may be retrieved through a consideration of field and tenor, the enactment
of the interaction itself is dependent on features of the material context of
situation, which are very difficult to retrieve from a text of this nature unless
a) participants make overt reference to it in the text, b) the formatting and
technological framing markers that accompany the receipt of each post is left
intact in the transcript of the interaction, c) the reader of such a text is
familiar with such interaction through having personal experience of such a
mode of communication, or of course, d) one is actually a participant in the
interaction, either as audience or contributor.
Such interaction is carried out wholly via text sent through the interfaces
of personal computer and software to a local BBS, where the messages can be
seen, read and responded to by anyone else having access to the same service.
This means that, while participants may respond to only one other participant
at a time, in practice, each person's message is addressed to a large unseen
audience of readers, which I note here only as a means of defining more delicately
the mode of interaction as a case of one-to-many, rather than what might sometimes
appear in text as one-to-one. The whole text is therefore 'marked' to the extent
that its production and reception do not follow previously expectable correlations
between features of the textual metafunction and that of the context of situation
as a whole. It is also obvious that because of this (and the text being used
here as exemplar shows one possible occurrence - cf below), contributions to
the multilogue are not necessarily sequential with respect to what might be
said to 'initiate the response'. In the case of the headers in a BBS, this is
made clear by sequential numbering, so that the subject line will read "Re:
<name-of-thread> (2) [..3,..4 &c]" (whereas, in the altered text here,
such indicators no longer exist as a textual feature).
To refer again to the matter of lexical density and its relationship to observed
ratios found at the ends of the written-spoken continuum, it is interesting
to note that in a study undertaken by Yates (1996) for example, it was found
that within his selected data, a comparison between i/ CMC, ii/ 'written', and
iii/ 'spoken' corpora, revealed that along the dimensions type/token ratio,
lexical density, and modality, CMC data was found to be comparable or closer
to that of speech, rather than writing. In the case of the use of modal auxiliaries,
it was found that in CMC, the frequency was higher than in speech or writing
(op cit: 44). While it would be enticing to draw conclusions regarding the reasons
for these findings, it must be said that such conclusions would be based on
intuitive and experiential knowledge only. That being said, however, from my
experience it seems that people who communicate in a computer-mediated environment
find it necessary to make their strategies for being 'polite', and their need
to express solidarity /power distance attitudes much more explicit in such a
medium than they would need to do in normal conversational settings.
Because of these technological, material-context constraints, I would like to
propose a dimension 'involvement' (pace Biber), or, to use Ure's (1971) scale:
most active - most reflective (where spoken text is more spontaneous and thus
more active and less reflective than written language, which in turn is reflected
in several factors one of which is lexical density) as a descriptive device,
which would use the dimension time taken to respond to contributions in CMC
interactive texts. In the case of email, for example, information in the header,
certainly part of the text, shows at what time each message was sent. The text
under consideration has been stripped of these textual indicators in order to
highlight the conversational nature of the discourse. However, the source, being
a BBS, allows one to more delicately investigate the degree of 'involvement'
of the participants because the interface makes available the time the message
was sent, and the time the message was read by whom, as well as the time a response
was begun, and sent - allowing a calculation to be performed of the time taken
to actually write a response after an initial reading.
Therefore the notion of process-sharing in the context of CMC-generated texts
does not need to be considered difficult to apply due to the problematic nature
of the written/spoken and graphic/phonic interactive dimensions of mode, if
the dimension of 'involvement' or time taken for a response to be made, is also
used in this analytic context. It is posited, for example, that the time taken
to respond will have some bearing on the features of process-sharing and the
textual metafunction of the texts so produced. Not only this, of course. As
was pointed out earlier, the textual and the experiential are closely linked,
and it appears from personal research in the area of email discussion lists
that the norms of interaction, and the field of discussion itself will also
have a bearing on the textual metafunctional realization of register in this
mode.
If we now try to characterize the text, the 'conversation', as a type of discourse
structure showing 'higher patterns of language' perhaps above the level of the
sentence in a hierarchical framework, is it possible to adopt a framework developed
in a context completely different from that of CMC and yet find comparable structural
similarities? The model developed by Sinclair, Coulthard and Brazil for the
analysis of classroom interaction hopes to provide a framework flexible enough
to be adapted for other contexts, and so I would like to briefly discuss the
descriptive usefulness of the model in the context of the text under consideration.
Coulthard & Brazil (1992:64) say of such a model "...the structural framework
operates by classifying each successive discourse event in the light of the
immediately preceding one." , and Sinclair proclaims that "The need for a level
of discourse, where the higher patterns of language can be described without
reference to any particular social use, is fairly obvious." (1992:88)
The following are the elements of structure in the model:
Initiation [I], which 'prospects' a response [R];
[I*] which is a move not prospecting [R];
Challenge [C] which diverts [R]; and
Follow-up [F] which acknowledges, or 'encapsulates' an [I] move.
Possible sequences:
I R (F)
I*(F)
IC = I
I*C = I
IRC = I
In this model prospecting moves are said to contribute to discourse management
and operate on the interactive plane, showing the process of the unfolding discourse.
Therefore Initiation moves have as their motivation a need for some type of
response - either activity or verbal answer. In the case of CMC, it can already
be seen that silence as a phrase can take on new meaning when any move seems
to prospect a non verbal response (for example a call to do something rather
than verbally respond). The autonomous plane on the other hand shows the product
of the discourse, the meaning it has related to the previous contributions,
so that responses will serve to show the perlocutionary force of any previous
utterance in context. A [C]hallenge move breaks presuppositions and initiates
a new exchange, which thus "cancels the interactive value of the previous move,
leaving only its contribution on the autonomous plane...[in this way]...the
subject matter becomes the discourse itself." (Sinclair.1992: 87)
What becomes the 'meaning' or 'implication' that each contribution really might
have in the context of the interaction is sometimes difficult to decide. For
example, what might appear on the surface to be a request for information, may
in the context of the situation imply that the speaker is demanding some form
of behaviour, or expressing disapproval, or merely making an observation, and
that it is therefore no easy matter to determine whether an interlocutor has
made an initiation, a challenge, a response, or a follow-up move. In other words,
the autonomous plane does seem to be an important means of understanding the
management of the discourse interaction: the experiential and the textual metafunctions
are closely related.
The problem with this model for me is that, even though it hopes to introduce
criteria that are predictive or 'prospecting', if each contribution is only
classed by reference to the utterance which preceded it (cf above) one significant
area of meaning seems left out of the picture. In this sense, the model hopes
to use a dynamic perspective, by classing each move according to its place in
the conversation at the time of utterance - as a response in context. But at
the same time, I felt compelled to use also a process of 'back analysis' in
order to tag the sequences of moves - as an acknowledgement that responses made
by participants themselves show how each previous contribution/utterance has
been taken up within the context of the conversation itself, in 'real time'.
In other words, each move in the exchange(s) is also tagged according to the
move which follows it, relying on the perlocutionary force of the contribution
rather than the analyst's perception of illocutionary force (to use what I consider
useful conceptual distinctions from the field of pragmatics). This also helps
to bring more of a participatory dimension back into play in the analysis, rather
than relying on entirely synoptic perspectives (although it can be seen that
tagging moves according to whether they are deemed to be prospecting a response,
also uses a dynamic view): the reader-response, if you'll allow, determines
the previous contribution's meaning in context.
The text appears below with series of tags appended to each move in the conversational
exchange. In many cases, the contributions which follow have been looked at
to determine what the respondee, or the next participant's response, shows has
been taken as the meaning of the contribution to which it is a response. This
means that the tagging appears not as one possibility, but often as two, labeled
here a), b) and c) to show what I have decided might be the flow of the conversational
structure. It should also be noted here once more, that contributions do not
always appear in sequence in CMC conversations: the response to a contribution
might follow two or more contributions later, sometimes even longer, and it
must be realized that the time taken for the participants to read and respond,
and the ease of responding to anything any time makes the concept of interruption,
prevalent in Conversational Analysis, irrelevant in most sequences. It also
makes it difficult sometimes to decide the sequence of moves if the headers
and subject lines have been removed, as these usually set up the relevance of
the post by contextualizing it in a named thread and a numbered post to which
it is responding. For example, in the text below it seems that contribution
9)-9a) is actually a follow-up move to contribution 7). It is also obvious that
I have considered some participants to have made 'rhetorical moves' for which
they intend to answer or give response to themselves.
S1: 1)What has love got to do with relationships?
[I] a)b)
S2: 2)Isn't love just one of the many relationships we are seeking by being
here in Japan? [C]a) / [R]b)
S1: 3)Gracious - how the man does boast.... [R]a) / [F]b) / [C]c?)
S3: 4)It's getting nasty in here.... [C]a) / [I*]b) The very person --S2--
has to show up here to smooth it out! [C]a) / [F]b)/c?)
S1: 5)Mmm yummy - being smoothed by --S2-- [R]a) / [I]b) - what a treat
- [F]a)b)
Can I be smoothed with Maple Syrup? [I]a)b)
S3: 6) Maple Syrup could be too sticky, though. [C]a)b)
S1: 7)Nonsense! You just have to lick hard..... [R]a)a)1 / [C]b)
S3: 8) Maple Syrup tastes always good with pancakes, not only itself....
[F]a) / [R]b)
8a)That's what --S2-- is looking for out there, right, --S2--? [I]a)b)
S4: 9)You've discovered one of Canada's greatest secret uses of maple
syrup.
Bon Apetit! [R]a)b)
9a)(and lick liberally! )[F]a)a)1b)
S5: 10)I'm going to Canada next month. [I*]
10a)How many people in here would love to get Canadian Maple Syrup
and try out the secret? [I]a)b)
S1: 11)Count me in... [R]a)b)c)
S5: 12)Of course I will, Mr.Newly Wed. [C]a) / [I*]b) / [F]c)
12a)I hope Canadian Maple Syrup will add some more sweetness in your
marriage life. [R]a) / [F]b) / [I*]c)
My goal here has been to test the model against data to see in what ways the
model is an adequate analogue for descriptive purposes, and in what ways it
fails to account for observed linguistic behaviour. From my point of view, it
would seem that such a structural model may be a little too parsimonious for
texts of this nature, and that analysis of conversation and conversation-like
interaction needs to incorporate more than that offered by a structural account
of the relationship between instances of classes of moves within larger structures
of exchanges between human participants.
On the other hand, I have not analysed the data to any great delicacy. If the
main obligatory moves/sequences in this hierarchical model are able to be consistently
conflated with classes of metafunctional realizations over a wider set of data,
such that, for example, mood and modality features indicating distinctions between
propositions and proposals regularly construe certain classes of move or act,
then I feel the model may have descriptive and interpretative relevance for
this type of interaction (see some preliminary musings in this direction in
my masters dissertation). However, it is easy to see that every realization
in the mood system; statement, question, offer and command; would be able to
realize either an [I]nitiation or a [R]esponse move, for example, so it would
not seem that any one-to-one functional correlations of a purely grammatical
or linguistic nature can be postulated, while it may be possible to outline
patterns of discourse structure (commonly realized by certain linguistic features)
which seem to regularly recur in such contexts.
The structure of the text is closely bound up with the notion of coherence,
so that any analyst or participant reacts to utterances according to the apparent
meaning in context, in the sequential unfolding of the text: cohesion is said
to contribute to texture, which in turn is manifested by certain kinds of semantic
relation, and this in turn may help any text to be perceived as coherent. On
the other hand, 'the property of texture is related to listener perception of
coherence' (Halliday & Hasan. 1985:72). So while the cohesive ties which
bind the text together are to be found intra-textually, and can be pointed out,
one must also look outside the text, to the context of situation itself in order
to decide whether such intra textual cohesion is perceived as such by other
participants: listeners and/or readers...
Furthermore, it becomes obvious that the would-be analyst will need to have
had -experience- in such contexts of situation in order to be able to retrieve
the 'meaning' from the lexicogrammatical features, cohesive ties and all, of
any text. To make things even more 'delicate', experience of one context of
situation will not necessarily be relevant in another 'culture', as meanings
are locally-determined. As Hasan (1985:101) points out, 'a given context of
situation - a CC - has meaning only within a culture.' Put another way, Shiro
(in Coulthard (ed).1994: 174) cites Morgan and Sellner who maintain that 'formal
cohesion is a natural effect of textual coherence, rather than the cause [of
textual coherence]'. In any case, it seems that cohesive ties have a descriptive
usefulness in telling us what devices participants use for ensuring that 'what
is said is relevant to the discourse, and [that it] relates to its context'
(Halliday & Hasan 1985: 45), my point being that such devices in this context
must be linguistically encoded, rather than, for example, signaled by such things
as rhythm, intonation and gesture, and that therefore these linguistic elements
may be likely means for discovering how the discourse is perceived as structurally
coherent and therefore relevant to the ongoing interaction.
One approach which might be applicable, if we want to consider by what means
any interaction can be considered as coherent and in what way each contribution
is deemed relevant by the participants, are notions of conversational implicature,
especially the maxims proposed by H. Grice (see Levinson. 1983). These maxims
are not prescriptive or structural, but said to be used somewhat predictively
from the point of view of participants. This does not imply that participants
apply any conscious rules to conversation, nor that there are any cognitive
constructions in the 'mind' which account for conversational behaviour on the
semantic level, but that participants generally assume that any conversation
is a goal-directed activity, and that meanings are being made or negotiated
within such interactions. Participants are therefore likely to 'read' any response
or contribution as co-operative at some level even if appearances may suggest
otherwise.
These maxims may be briefly described as:
i/ Quality, ie, one should be sincere, as truthful as possible
ii/ Quantity, ie, one should say (only) as much as what is required at that
juncture
iii/ Relevance
iv/ Manner, ie, one should avoid ambiguity and obscurity, try to speak clearly.
However, for me, all of these may be subsumed under the notion of relevance,
which seems to account for all conversational moves which can in any sense be
seen to be related to any previous contribution in an interaction, especially
where notions of structure above the level of the sentence are invoked. The
means by which relevance is perceived by participants is open to conjecture,
but the means by which moves are made or relevance implicated (through reference
and cohesive devices for example) are open to observation and description. Here
I would like to suggest that so-called Gricean maxims are a useful conjectural
gap-filler which proposes how perceptions of coherence and relevance are related
to the linguistic realizations of actual conversations, whether written or otherwise.
Deliberate flouting of conversational conventions, or expectations, for example,
will account for the reading of irony or sarcasm, or some interpersonal positioning
move which may imply social distance, or a power relationship with respect to
the interlocutor(s) (see Brown & Levinson 1987, Hodge and Kress 1988, and
Fairclough 1992: 123 here). This may be seen in the response S1 makes at 3)
when he refers to the previous contribution as a 'boasting' move, when it appears
to have been made as a neutral observation pertaining to the ostensible topic
of the conversation. This means that, although the comment made at 3) might
be seen to be irrelevant in the context of the conversation, the next participant
seems to interpret it as being relevant at another level, which appears to him
to be the level of interpersonal attack. Overt personal comments are generally
an indication of lack of distance, ie, eschewing normal western conventions
of politeness, where such conventions might be relevant between participants
who are not familiar to each other. On the other hand, personal comments of
this nature, while acceptable between equals, convey different implications
where social distance is assumed: the next participant then characterises the
interaction as a 'space', and quips at 4) that it is 'getting nasty in here'
thus seeming to reject any claims of solidarity S1 may be making with respect
to S2 , and then goes on to mention the participant (S2) by name saying that
he has 'showed up to smooth things out.' which again may be seen as a distancing
move with respect to S1. That S1 is aware of this, ie, interprets it as relevant
to the conversation while certainly not at all co-operative, is shown in S1's
next move which takes up the metaphorical allusion to things 'being smoothed'
out, concretizes it as an activity and links it back to the named participant
S2, in a superficially positive way. He also makes a seemingly irrelevant introduction
of the thing with which he would enjoy being smoothed with, which in turn is
taken up as the topic of conversation thereafter.
With some experience of these types of contexts of interaction, specifically
CMC, it is my contention that some contributions in this mode are made with
the expectation of a negative response - what I would call a 'hook' move/initiation/response,
and many of S1's contributions can be seen in this light. In some ways this
would seem to fly in the face of Grice's 'co-operative principle', but only
if one interprets 'co-operation' as having a limited set of realizations in
interactional activity. For example, in order to co-operate with my interlocutor
if we have decided to debate an issue, or even play a game of cards competitively,
it may be necessary for me to flout the maxim of quality, manner, or relevance
in order to continue the game at all, or in other words to set up the conditions
for a response, even if this is an angry or superficially 'uncooperative' response:
silence may be a phrase and be read as having some meaning in context, but for
interaction to continue at all there often needs to be some conflict. Certainly
in CMC, 'silence' will denote that there is no interaction in progress, and
so, in order to 'co-operate' in this mode, interactants may need to be a little
contentious.
The point might be that the medium and channel of interaction in this case promotes
the deliberate flouting of such cooperative principles while maintaining a superficial
textual similarity to conversational patterning, and mainly because the process-sharing,
while necessarily part of the mode of text creation, IS deferred in time and
space. What this means is that because of the attenuated process-sharing interaction
available to participants, usual means of eliciting a response being impossible,
deliberate flouting of such maxims, especially in expectations of Manner, actually
promote misunderstandings, or multivalent interpretations, which in turn promote
responses in the effort to clarify co-operative goals (see for example G.Kress
1985, on matters of conflict and silence).
BIBER, D. & E. FINEGAN (eds) 1994. Sociolinguistic Perspectives on Register.
OUP.
BLOOR, T. & BLOOR. 1995. The Functional Analysis of English. Arnold.
BROWN, P. and LEVINSON, S. 1987. Politeness: Some universals in language usage.
Cambridge University Press.
COULTHARD, M. (ed) 1992. Advances in Spoken Discourse Analysis. Routledge.
COULTHARD, M. (ed) 1994. Advances In Written Text Analysis. Routledge.
DON, A. 1997.The Context of Mailing List Interaction. http://www.loris.net/lexie.html. See also >>
EGGINS, S. & D. SLADE. 1997. Analysing Casual Conversation. Cassell.
FAIRCLOUGH, N. 1992. Discourse and Social Change. Cambridge: Polity Press
FERRARA, K. et al.1991. Interactive Written Discourse as an Emergent Register.
Written Communication. vol.8:1: 8-34.
GHADESSY, M. (ed) 1993. Register Analysis. Pinter
HALLIDAY, M. A. K. 1994. An Introduction to Functional Grammar. Edward Arnold.
HALLIDAY, M. & R. HASAN. 1985. Language, context and text: aspects of language
in a social-semiotic perspective. OUP.
HARRÉ, R. & L. VAN LANGENHOVE. (eds) 1999. Positioning Theory. Blackwell.
HODGE, R. and G. KRESS. 1988. Social Semiotics. Cornell University Press.
HOEY, M. 1991. Patterns of Lexis in Text. OUP.
KRESS, G. 1985. Linguistic processes in sociocultural practice. Deakin University
Press.
LEVINSON, S. 1983. Pragmatics
MARTIN, J.R. 1992. English Text. John Benjamins
McELHEARN, K. 1996. Writing Conversation: An Analysis of Speech Events in E-mail
Mailing Lists. http://www.mcelhearn.com (a typology of email lists)
MONTGOMERY, M. 1986. An Introduction to Language and Society. Routledge.
RUESCH, J. and G. BATESON. 1987 (1951). Communication: The Social Matrix of
Psychiatry. New York: W.W. Norton.
YATES, S. J. 1996. 'Oral and written linguistic aspects of computer conferencing'
in Susan Herring (ed). Computer-Mediated Communication. Benjamins.