. In the example above, the first 'exchange' actually begins at(4), where a nominating 'act' realises the first 'move' This could be considered a 'reframe'. The writer in fact quotes himself at (4.2.2), and it can be seen that in this way the writer is making a clarifying comment of his own previous post.
As this exchange seems to be the 'head' of the post, it might be fair to name the whole post 'Follow-up: Clarify', which seems to be a typical response type. However, because most posts do not necessarily realise only one conversational function, and indeed may realise moves or exchanges resembling conversational initiations, responses and follow-up all within the same post, it is difficult to assign any definite form-function correlation to whole posts. Determining the 'head' exchange might be feasible, in which case, whole posts could be named after the function of the 'head', which could be determined by looking at the function of the first 'Comment' after the opening sequence or Boundary Frame, as is suggested with reference to the main text example above. Within the confines of this study it will not be possible to do more than suggest that this is the case, and instead concentrate on some of the features which are relevant to an understanding of what goes on within posts through looking at a number lexicogrammatical signals.
In other words, the categories of 'exchange', 'move' and 'act' that will be introduced below are meant to form a convenient framework within which intra-message coherence may be discussed, and which may then be used in making 'predictions' about how readers interpret such possible signals. As such, these structural categories make no claim to delicacy at this point .
Criteria used by Coulthard and Brazil (1992: 71) in defining elements of exchange structure use the notion of 'predicting' and 'predicted'. They propose a type of move structurally identified by I/R which is both predicting and predicted; in other words, it fulfils two functions at once. Posts as utterances must be considered as such, although as stated previously, within the body of the post, that is, between the opening and closing sequences (or 'boundary exchanges'), several exchanges are set up by the writer through reframing moves which are typically realised by quoting the part of a previous post to which the participant wants to respond - and wants to be responded to.
Referring once again to the main text example above, at (5) the writer comments on his comment in a clarifying move, followed by an evaluation of what he did at (5.1) This is followed by a closing sequence which is realised by a type of 'hook' act (6), calling on other members of the discussion and naming them as 'Netdynamo shrinks', a somewhat derogatory means of referring to the participants who are psychologists. He then addresses several moodless questions to them. The post is closed with his signature, in this case his first name, and then a somewhat joking identifier, referring to the original opening sequence where he quotes another participant's sig file 'identifier'.
Before going back to analyse this text in more detail, it is useful here to ask by what means I determined that the closing sequence was begun at (6). In other words, how might it be determined that a 'phase shift' has occurred? Up until that stage, the writer had been operating as 'primary knower' (K1) giving the audience information as to his 'knowledge', beliefs, or experience. At (6) however, he shifts to an A1/K2 (primary action seeker) position, marked by his calling on the audience, and then asking them to provide the answer to his request.
In terms of lexicogrammar, the call was a type of summons, what Halliday calls a vocative, and the modality is evidenced by the attribute he gives those members of the audience in referring to them in a less than polite manner, thus he may be expressing his disdain for them. However, a person who was not new to this discussion would be able to retrieve something else from their mental 'identity kit' construct which they may have to hand, and understand that this is a feature of this person's 'style' of interaction, and that furthermore he is not new to psychology himself, and may actually be expressing solidarity with the members so addressed by making such comments which reflect a lack of social distance (cf Brown and Levinson 1987). The questions themselves are realised by moodless interrogatives, and this again marks a change in position from the earlier part of his post.
At this stage, under this rubric, it might be fair to posit a tentative general description of units in a rank scale down to the level 'act', which can be usefully applied to the exchanges of discussion evidenced over a one year period in this mailing list's interactions. Although tentative categories of act are named, I am not willing to make any positive identification of categories beyond the 'move' at this stage, especially since my main interest lies in the means by which participants come to understand, conceptualize, or interpret cues for structure in email list discussion, and therefore my aim in positing these ranks is rather as a framework within which further discussion can be made regarding the context of the interaction. In other words, what might constitute a typical 'speech event' in this 'speech community'?
As pointed out earlier, the 'material situation' of the speech community is conveniently defined by the Mode itself: it is the List, the automatic server which dispenses its messages to all subscribed, holds the archives and dispenses these records of list history to anyone who subscribes. It sends the welcome message, and a list of commands that the server will carry out on behalf of the subscriber. It names the 'leader', the so-called listowner, a person, who is able to unsubscribe offenders at will, and who is charged with keeping either strict order or benign commentary, as list participants decide. In the case of the list under study, they did decide, by public 'show of hands' when the job was up for grabs. The speech community past this level, consists of the messages that these subscribers care to send to each other.
It is probably fair to say that backchannel communication plays a big part of the actual speech community, but for the purposes of this study, it is the public discussion which must necessarily form the only reliable data.
Beyond the list itself, the cultural background, and the actual material situations, identities, and life trajectories of the participants must also have a bearing on the dimension 'speech community', and in fact it is this very tension of not knowing 'who' is speaking nor who one is speaking to which makes the negotiations over norms so fraught with possibilities for offence. It is therefore no wonder that aspects of interpersonal communication, role and orientation figure so obviously in discussion on this list, and that it leaks into the experiential dimension with topics of discussion centring on identity, personal 'boundaries', and group roles continually resurfacing.
How then might the 'speech event' be characterised within such a community?
It has to be said that the list history itself forms a backdrop for every message posted, but at a more practical level, the highest rank for purposes of discussion is probably 'the thread' which generally sets out the subject or topic of the discussion. There are generally several threads going at once, and means of naming threads, and changing names of threads should the topic bifurcate, are norms that are often referred to. Shank (1993) has characterised this context of multiple parallel threads as 'abductive multiloguing', which indicates the nature of the interaction. Interaction is deferred in both time and space, unlike, for example, IRC (internet relay chat) communication which is more or less conducted real time, although without benefit of the phonic channel. Therefore, messages may be composed much like a letter, but because there may be many messages addressing the same and different topics appearing at the same time, again, the notion of transitional relevance, and the means by which each post is contextualised, or 'framed' becomes an important element in message structure.
Within the thread is the 'message' or 'post'; the two terms seem interchangeable. I believe that messages themselves are of only two easily discernible types; Initiation and Response; although further analysis of the data may provide a basis upon which to make a more delicate categorization. Initiations, in the sense that the term is used here, introduce new topics but are sometimes made in response to another post. Certainly most posts hope to elicit other responses even when they are responding to some other post(s) In this way, it could be said that most messages are of a R/I type. Of course, within messages are follow-up and evaluative type comments, as well as what could be seen as eliciting moves and responses to requests. However, within the data that this particular mailing list has generated, explicit requests are not the norm; rather, participants put forward points of view, and in this way they hope to elicit responses or follow-up evaluative comments.
Sinclair's (1992) notions of prospection and encapsulation are useful here, but at this point it should be stated that the terms 'initiation' and 'response' as applied to whole posts or messages are generalisations only, and must be distinguished from initiation, response and follow-up moves as elements of the structure of the exchange, which must be viewed as taking place within and not between messages.
Sinclair (opcit: 85) goes on to point out that while third moves, or follow-up type comments are felt to be obligatory in some situations, this will depend to some degree on participant status in the group: where social distance is greater, the third move seems to become more necessary. This is almost certainly the case in classroom interaction. In the case of Netdynam, matters of social distance and solidarity are matters of both content and process, or observable on both the autonomous and interactive plane. Further study of these perspectives and their interrelationship should yield useful results.
Typically, transitional relevance is set up through the use of quotation within messages, so the concept of 'turn' and 'interruption' by another listmember is not part of this relevance due to the mode of communication. In other words, one's message may, in theory, be as long as one likes (although unwritten, yet oft-stated norms require that messages be kept shorter than around 8K), however, it may be that no one except the hardy will bother to read it. In any case, when responding to any other post, a frame will need to be made in order to (re)contextualise the post's message, and the most conversational and easiest method is to quote those parts of others' messages that one wishes to respond to. Response-type moves may be categorised into several classes, and while the following list is probably not exhaustive, it covers all of the response types I have come across to date. Within these types, there may be finer shades of delicacy at the rank 'act' than I have noted, but I do not want to argue for them at this stage. Once again, it should be noted that several of these types of 'responses' may be evident in any one post.
The main types of response are:
While 'Ignore' cannot be a response move when there has been no explicit elicitation, participants notice and talk about non response as an option. This is because most posts expect or hope for a response or follow-up of some sort, even though such expectation is not always linguistically encoded. Most posts are of the type R/I as stated above. In effect, most posts make some forward -looking or 'prospecting' informatives. The participants look for follow-up comments or acknowledgements to whatever they have written. When such prospections are made explicit, it is usually in the preclosing sequence, a few examples of which will be examined below. For now, all responses to others' posts are subsumed under the label 'Response', although it can be seen above that moves such as Evaluation and Disagree for example, might be more precisely defined as Follow-up moves, or given a broad descriptive title such as 'encapsulation' as outlined by Sinclair (opcit: 86).
Initiation-type messages are similar except that, if they are introducing a new topic, in contrast to merely responding to something in another message and signalling a change of perspective with a new subject line, some of the categories listed above change slightly (for example, there is no Comment: clarify in a totally new topic). However, as most new topic initiations also make reference to what has gone before, it would be better to look at initiating moves within posts, than to label whole posts as initiations. On the other hand, there are some posts which introduce completely new topics, or matters unrelated to the ongoing discussion. Occasionally, these posts are self-contained informatives and require no overt linguistic response, for example, announcements. Other posts seek to introduce new contentious issues, or create heated discussion, and as such may be considered 'elicitations' even though they pretend to ask nothing at all. The so-called 'troll' is a good example of this type of post. Occasionally, such posts will evoke 'responses' of "Off topic!" or arguments about the truth value of the 'information' or position of the poster, who will not respond (make a follow-up comment) at all. This again, I feel, relates to power/distance and solidarity issues. The troll is a means to subvert discussion, and by not providing follow-up the poster shows his/her distance from the group. Within this group, norms seem to have evolved which 'mark' such behaviour, although space does not allow a more complete discussion of how these norms have been discussed.
Another class of move which realises an initiation, is obviously the request, either for information or opinion, as well as offers of same. These types of move are found both within posts introducing a new topic, or which are ostensibly responding to another post.
Most messages, although they strive to keep within the topic area outlined in the subject line, actually perform several functions at once, and indeed are sometimes comprised of several 'exchanges' using quoting, and what Halliday (1994) calls 'projecting' clauses (paratactic: locution). These are marked by the formatting conventions noted above (ie, a chevron '>' at the beginning of each line of quoted material is the most usual, and is a function of the mailer each poster uses). As stated earlier, it might be fair to determine what the 'head' part of the message is and name the function of the whole post after that, but I do not think this serves any useful purpose within the bounds of this study. Therefore, from this point on, the 'types' outlined above will be used to refer to exchanges, or moves within messages. The relationship of these intra-message structures to that of the message or post as a whole, and whether such generalised descriptive categories may be linked to inter-message factors, remains to be seen and is relevant for further study.
The 'header' also forms a significant rank within message, as it gives readers a setting that, if absent would mean that the message could not be seen as such. As stated above, it sets out the list address, the listserver address, the time the message was sent, by whom, and the subject of the message. Also, it forms part of the transitional relevance by indicating the topic of the message in the subject line, important if one is interested in following a particular thread.
As stated above, the message itself, the 'body', in email metalanguage, usually starts with some type of 'opening' which I call a boundary frame. It consists of at least one act. At present I see these as:
These same moves can be found within the main body of the message as well, and as such, represent 'reFrames', and reorient to a new situation or problem that the poster intends to address. The most common is obviously reFrame: Nominate and Quote.
Comments in opening sequences seem to take on a double function: that of responding to the 'quoted' material, and providing a reframe for the body of the message and what ideational content the poster specifically wishes to address. As such it must follow another framing act; if a Comment is made without a preceding framing act, then it becomes part of the Setting: metacomment. Otherwise, Comments are not part of the opening sequence, but may be better viewed as perhaps the 'head' comment in the main body of the message
Within frames there is the 'loop' whose function is to refer back to another point in the 'conversation'. These also do not necessarily appear at the beginning of a post, but may also appear at the beginning of an exchange that needs 'reframing'.
The difference between Nominate and Greet, when names are used in a greeting, is determined by the following move, or text: If what follows is meant for everyone, and is using the named person as a contextualiser, or as someone who is then quoted, then, it is a Nominating move. If what follows is actually a comment addressed to the people named, determined by mood, personal pronouns etc, then it is a Greeting.
Setting moves may involve a number of 'acts'; those noted above are the ones observed to be the most prevalent.
What follows are a number of opening sequences showing the typical pattern outlined above, that is, showing one or more of the three framing moves outlined:
-------------------------------------------------------
| 1. | Date: Fri, 10 Nov 1995 23:44:49 -0500 |
| From: Simon | |
| Subject: Email Exegesis | |
| Folks:[Greet] | |
| We seem to have a bit of break in list weather lately, or maybe | |
| it's the eye of the storm. [Setting:loc/ feel] | |
| . . . . . . . . . . . | |
| 2. | Date: Sun, 21 Jan 1996 10:14:00 EST |
| From: Matt | |
| Subject: I'll have my cake with a little irony, please | |
| Whoah, [Greet] | |
| I'll set Bunny and Toto on to you, you keep that up... :-)[Setting:loc/ ref] | |
| The land of the people of the south wind, where the grain elevators | |
| are cathedrals and the sunflowers are heliotropes and the TV | |
| satellite dishes aren't. [reFrame:id/ loc] | |
| . . . . . . . . . . . . | |
| 3. | Date: Fri, 22 Dec 1995 00:20:33 -0800 |
| From: Jerry | |
| Subject: Re: 2nd of 3 -- Re: net transference | |
| At 6.32 PM 12/21/95, Hugh Stilley wrote: [Nom] | |
| >I want to know what "shared" fantasy means in the net context. [Quote] [A1/K2:Request for information] | |
| Shared phantasies?[reFrame:S:loop] Let me speculate where they | |
| might come from, and why we can re-cognize them. Either it's genetic, or | |
| it's cultural, or it's something else - maybe even more general than either | |
| genes or culture. [Comment/ Explain/ Opinion] | |
| . . . . . . . . . . . . | |
| 4. | Date: Fri, 22 Dec 1995 14:55:40 -0700 |
| From: Fred | |
| Subject: Purpose behind Behavior | |
| On Fri, 22 Dec 1995, Roy W---- wrote: [Nom. NB:this is a function of the mailer itself] | |
| >It seems to me that there is a pronounced tendency for folks | |
| >to assign motive to observed behavior of others. Somehow, we seem to imagine that [Quote] | |
| . . . . . . . . . . . | |
| 5. | Date: Fri, 22 Dec 1995 18:47:02 -0800 |
| From: Steve | |
| Subject: Re: transference stuff | |
| Ian wrote: [Nom] | |
| >I'll give you a small example - transference is a word brimming | |
| >with meaning for many members of this group - to me it just means | |
| >something like ascribing to others what you in some measure really think/feel | |
| [Quote] [Comment: Opinion/ Clarify: id] | |
| That's "projection", although the concepts are related. An easy way to think | |
| of transference: it's the *assumptions* we make about the emotions of | |
| people we don't know well. [Comment: Clarify/ Explain: add] | |
| . . . . . . . . . . . . . | |
| 6. | Date: Sat, 23 Dec 1995 12:54:25 -0500 |
| From: Simon | |
| Subject: soliloquy | |
| Folks: [Greet] | |
| Ian's recent comments prompted me to an emotional check-in. | |
| I do not know whether it is process or content | |
| [Setting: loop/feeling/reason] [Nom] | |
| Steve made mention of "psych-types" and the fact that no | |
| matter what road we end up following, they would always be | |
| around to look at the things which interest them. [Setting: loop] | |
| I do not consider myself to be one of the "psych-types" and during | |
| the first few weeks here found them to be rather annoying | |
| [Comment: Clarify/ Opinion: id] | |
| . . . . . . . . . . . . . | |
| 7. | Date: Sat, 17 Feb 1996 03:30:10 EDT |
| From: Frank | |
| Subject: Earburn | |
| Tick tock, tick tock, tick, talk, talk talk Someone murmur | |
| the words of power? [Summons:(marked opening) Setting: refer to in-joke/ loc] | |
| Who summons me? [answer to summons] | |
| Anyone feel the net convulse? [reFrame: loop: loc] | |
| Howaboutatremoralathespider'sweb? [reFrame:loc: refer to in-joke: id] | |
| See a pretty wave of donminoes going down?[reFrame: Setting: (marked by | |
| rhetorical interrogative and run on words) loc/ in-joke:ref to shared culture:the net] | |
| A law can be tyrranical but is not its application the true test? | |
| many a benign statute has rendered justice impotent through determined | |
| application while draconian rule does so with a feather touch...[Comment:Explain/ Opinion/ Muse] |
------------------------------------------------------------
In order to determine to what extent participants actually employed such framing moves in opening sequences, the two days of digests (from which the main text example was extracted), were analysed, and the number of such moves were counted, then looked at as a percentage of total posts made. As a matter of interest, and to discover whether the use of such moves may have changed at all during the course of a year, two days of posts from the same date in the following year were also analysed. The results are shown in Table B which can be found in Appendix 7 together with a brief commentary. At this point it is enough to note that both sets of analysis revealed that the most common type of boundary frames were Nominate (77%) and Quoting (76%). The type of move least employed was Greeting (9%). Experience suggests that the reason for this may be due to the fact that a participant feels no need to greet other participants if they are making several posts on different topics on the same day, or on a regular basis. Furthermore, that if a participant names and quotes another person's post as an opening, any greeting may be redundant, so that they are only used when the poster has not made a message for some time, or when the post begins a totally new thread or topic, realising what has been termed for this study an Initiation type post. (Appendix 5 shows an analysis of the opening sequences or framing moves from January 20-22, 1997)
Posts are sometimes closed with a short preclosing comment, the usual signoff of the poster, followed by an optional identifier as mentioned previously. In the example text quoted above, the preclosing comment, which is optional, is realised by (6), (6.1), the signoff by (7) and the identifier by (7.1) Preclosing moves (if they occur) are usually one or two of the following, and usually represent a phase shift as signalled by a change in orientation (mood) or topic change:
(Appendix 8 of the dissertation gave an analysis of the closing sections of a number of posts.)