Chapter 1 (a part of)
Intersecting machineries
> Shows many important features of conversation
Any bit of talk is the product of several “organizations” which operate concurrently and intersect
in the utterance, thereby giving it a highly specific, indeed unique, character.
Harvey Sacks: interactions being spewed out by machinery, the machinery being what we’re
trying to find; where, in order to find it we’ve got to get a whole bunch of its products.
> This is a highly decentralized or distributed view of human action that places the emphasis not
on the internal cognitive representations of individuals or on their “external” attributes (doctor,
woman, etc.) but on the structures of activity within which they are embedded.
For a given occasion, there are specific places within it at which point particular actions are
relevantly done. E.g. greetings are properly done at the beginning of an encounter. When an
action is done outside of its proper place in conversation it is typically marked as such (with
“misplacement markers” like “by the way).
When Ann finishes her utterance - “That was fun” - she may relevantly expect Jeff to say
something by virtue of the way turn-taking in conversation is organized. So we have two more
organizations - the organized sets of practices involved in both the construction and the
distribution of turns - implicated in the production of this fragment of conversation.
Responses to assessments and other sequence-initiating actions (what we will call “first pair
parts” like questions, requests, invitations etc.) can be divided into preferred and dispreferred
types. Where agreement is relevant, a kind of “with me or against me” principle operates such
that anything other than agreement is a disagreement (with the Ann and Jeff example).
We can see this addition of “-ish” to Ann’s utterance as a form of self-repair. So we have another
organization of practices - the organization of repair - implicated in this short fragment of
consideration.
Individualist view: sees the utterance as the product of a single, isolated individual speaker.
Externalist view: sees the utterance as the product of intersecting, external forces such as the
speaker’s (or the recipient’s) gender, ethnic background, age, class or whatever else.
There is one more organization of practices that should be mentioned here - those involved in
selecting the particular words used to construct the turn.
A central principle of conversation which Sacks and his colleagues termed “recipient design”:
“the multitude of respects in which the talk by a party in a conversation is constructed or
designed in ways which display an orientation and sensitivity to the particular other(s) who are
the co-participants.
> Speakers design their talk in such a way as to make it appropriate and relevant for the persons
they are addressing. E.g. with an expression like “that” in “that was fun” the speaker clearly
presumes that the recipient will know what she means to refer to in using it.
,Anything that happens in conversation happens within some particular, ultimately unique,
context. As it turns out, although the structures that organize conversation are context-free in
certain basic and crucial respects, they are at the same time capable of extraordinary context-
sensitivity.
Chapter 3 – Turn taking
Turns and turn-taking provide the underlying framework of conversation. Turns are the basic
unit of conversation.
There is a widely held view that turn-taking is characteristic of “polite” conversation. The second
widely held view is that certain groups of people (e.g. West Indians, Jews, Native Americans)
simply do not take turns in conversation.
The one at the time rule
Sacks: ‘’the machinery of turn-taking is organized so as to minimize both gaps in which no one is
talking and overlaps in which more than one person is talking at the same time.’’ Of course, there
are exceptions. E.g. entering a room, its normal that multiple people greet you at the same time.
Laughing is also something that people do together.
“One-party-talking-at-a-time” cannot, be achieved by each participant waiting for the other to
finish. Several reasons why this won’t work.
- Even in the simplest situation where just two people are talking, waiting would result in
the production of a gap between the end of one turn and the beginning of the next.
- Were conversation organized this way, some kind of unambiguous “turn-completion
signal’’ (e.g. saying ‘over’ with a walkie talkie) would be required.
Grossly apparent facts of conversation (don’t have to learn them)
Turn-taking system is:
- Locally managed. It organizes only current and next turn and not, what will happen in
thirty seconds or tomorrow.
, - Party-administered. There is no “referee” to determine who should speak next and for
how long. The participants work this out.
Important: What is a turn-at-talk (a turn)?
Example 1 utterance of tourist would not be hearable by the co-participants as constituting a
turn. Example 2 would be a hearable turn.
Every language of course has grammatical rules and these appear to shape what will and what
will not count as a turn-at-talk. So only saying ‘has’ would not count as a turn, some word has to
follow. Except when you put it in a different context >
Intonation is also really important.
Turns are constructed out of units which Sacks, Schegloff and Jefferson called “turn-
constructional units” (TCU). A single turn-at-talk may be built out of several TCUs.
Various kinds of evidence converge to indicate that the system allocates rights to produce one
turn-constructional unit at a time – be it lexical, phrasal, clausal or sentential. At the completion
of each unit, transition to a next speaker may occur. Thus, unless special provisions have been
made, at the possible completion of a current turn unit, transition to a next speaker is relevant. >
a transition relevance place (TRP).
People monitor talk not only to find possible points of completion (voltooiing) but to project and
anticipate them before they actually occur. A point of possible unit completion is a place for
possible speaker transition > TRP.
Speaker transition at such transition-relevant places is organized by a set of rules.
The rules are ordered and this ordering is crucial to the way in which they organize the
distribution of turns-at-talk. E.g. a party wishing to invoke rule (b) must start early, before rule
(c) is invoked.
, With more than two parties, there are additional motivations for an early start since more than
one potential next speaker may target a given transition relevance place. This explains why next
speakers often start before the actual completion of a turn. Such early starts result in short
segments of overlapping talk. Early starters may begin “at the final sound(s) of the last word of
what constitutes a possibly complete utterance” (terminal overlap) or at a “recognition point”
where although the utterance has not yet been quite completed, “that which is being said within
and through it has been made available”
Selecting a next speaker can be done by e.g. an addressed question. Like questions, other
sequence-initiating actions may select a next speaker when combined with some form of
address. E.g. complaints or requests.
Sequence-initiating actions
How does a speaker show that they are addressing what they say to some specific co-
participant?
- The use of an address term. Saying someones name. Effective but not very common.
- Address is often signaled by gaze
- More tacit and context-tied methods. E.g. when it is made clear in the prior talk that only
one person is going to an event, so it is adressed tot his person.
- Select a next speaker by eliminating all other participants.
Transition space
The transition space can be extended and exaggerated so as to invite or re-invite speaker
transfer. It has boundaries that can be manipulated. It can be “opened early” or “closed late”. The
length of the transition space may also be either extended or compressed.
Conversationalists talk in ways that obscure, eliminate or highlight the possible completion of a
turn so as to compress or extend the transition space. A point of possible completion is
something that a speaker constructs and prepares for the recipients as a discrete place within
the ongoing course of talk.
Overlap and interruption
Overlapping talk, tends to occur in a highly restricted set of places in conversation. Most overlap
appears to be a product, rather than a violation. Thus, overlapping talk is typically not the
product of conversationalists “not listening to one another”. On the contrary, it provides
orderliness in conversation > evidence that participants attend to one another’s talk.
Many cases of overlapping talk occur when the beginning of a next turn starts just before the
prior has come to completion.
Three types of overlap which account for the vast majority of cases in conversation: “turn-
terminal”, “turn-initial” and “recognitional”. None of these are really interruptive.
Very common: overlapped talks get repaired. Often, an overlapped fragment will be repeated so
that it is eventually produced clearly > a repair mechanism.
Speakers do monitor another’s talk even if it overlaps with their own. A common place for
overlap is at and around possible turn completion.
Post-overlap hitches: e.g. sting of repeated I’s, stuttering. They may defend the beginning of a
turn against further interruption.
