IPRA PAPERS IN PRAGMATICSVol. 1, No.1.July 1987

TABLE OF CONTENTS

Opening Remarks .......................................................i

Table of Contents .................................................. iii

CHARLES GOODWIN 6 MARJORIE HARNESS GOODWIN

Concurrent operations on talk: Notes on the

interactive organization of assessments ............................. 1

DONALD BRENNEIS

The Fiji Indian pancayat as therapeutic discourse .................. 55

SANDRA A. THOMPSON and WILLIAM C. MANN

Rhetorical structure theory: A framework for the

analysis of texts .................................................. 79

KATHRYN WOOLARD

Codeswitching anal comedy in Catalonia ............................ 106

Notes on Contributors .............................................. 123

Style Sheet ........................................................ 125

iii

Concurrent Operations on Talk: Notes on the Interactive Organization of Assessments

CHARLES GOODWIN MARJORIE HARNESS GOODWIN*

The analysis of conversation has a strong relevance to the study ofpragmatics. Thus in introducing the scope of pragmatics Levinson (1983:284) notes that

It is not hard to see why one should look to conversationfor insight into pragmatic phenomena, for conversation isclearly the prototypical kind of language usage, the formin which we are all first exposed to language - the matrixfor language acquisition.

The field of study that has provided the most extensive analysis of thepragmatic organization of conversation is the line of inquiry initiated bythe late Harvey Sacks and his colleagues.1 Indeed

* We are very deeply indebted to Alessandro Duranti, William Hanks,Gail Jefferson and Emanuel Schegloff for insightful comments andsuggestions on an earlier version of this analysis. This paper wasinitially presented at the Annual Meeting of the American SociologicalAssociation, San Francisco, September 9, 1982.

1. For a detailed study of both work in conversation analysis, and howthat work is relevant to pragmatics in general, see Levinson (1983).See Heritage (1984a) for more extensive treatment of the relationshipbetween conversation analysis and the ethnomethodological tradition itemerged from within sociology, and Heritage (1985) for a detailedsummary of work within the field. For collections of specific analysissee for example Atkinson and Heritage (1984), Button and Lee (inpress), Schenkein (1978), and Zimmerman and West (1980). C.Goodwin (1981) and Heath (1986) examine in detail the

IPRA Papers in Pragmatics 1, No.1 (1987), 1-54

2

Levinson (1983: 285) observes that

if, as

we

shall argue,

the proper

way

to study

conversational organization

is through

empirical

techniques, this suggests that the largely philosophicaltraditions that have given rise to pragm

atics may have to

yield in

the future

to m

ore em

pirical kinds

ofinvestigation of language use.

Though starting from

an ethnomethodological, rather than linguistic or

philosophical, tradition the basic goals of conversation analysis are quitecom

patible with those of pragm

atics. Thus H

eritage and Atkinson (1984:

1) note that

The central goal of conversation analytic research is the

description and explication of the competences that

ordinary speakers use and rely on in participating inintelligible, socially organized interaction. A

t its most

basic, this objective is one of describing the proceduresby w

hich conversationalists produce their own behavior

and understand and deal with the behavior of others. A

basic assumption throughout is G

arfinkel's (1967:1)proposal that these activities - producing conduct andunderstanding and dealing w

ith it - are accomplished as

the accountable products of comm

on sets of procedures.

Much research w

ithin conversation analysis has investigated howsubsequent utterances display an analysis of prior ones, and how

suchsequential organization is a basic resource utilized by participants for theproduction and understanding of action, and the talk that em

bodies it.A

nalysis in the present paper will focus

----------pragm

atic organization of nonvocal phenomena, and processes of

interaction between speakers and hearers that occur w

ithin individualturns at talk, a topic that is quite relevant to the analysis beingdeveloped w

ithin the present paper.

3

instead on

how

individual utterances

and single

turns at

talk are

themselves constituted through an ongoing process of interaction betw

eenspeaker and recipient. T

o do this we w

ill examine the process of assessing

or evaluating entities that are being talked about. Study of this process will

provide an opportunity to investigate within a coherent fram

ework of

action a range of phenomena that are typically studied in isolation from

each other, including:

-H

ow participants achieve, and display to each other, congruent

understanding of the events they are talking about;

-T

he dynamic achievem

ent of social organization within the

turn at talk;-

The interactive organization of affect and em

otion;

-H

ow participants attend in detail to structure in the stream

ofspeech

as a

resource for

the organization

of their

ongoinginteraction.

-T

he integration of speech and body movem

ent within

coherent activity systems.

-T

he way in w

hich activity systems provide organization for

both interaction and the talk occurring within it.

In brief, the analysis of assessments w

ill permit us to analyze the

pragmatic organization of a range of social, linguistic, and cognitive

phenomena, as they are displayed and utilized by participants in the details

of their actual talk.

1 Interaction Within the U

tterance

One very productive strategy for uncovering the interactive

organization of talk has focused on ways in w

hich subsequent utterancesdisplay an analysis of prior ones (S

acks, Schegloff and Jefferson 1974:

728). How

ever, despite the great power of this

54

methodology, and in particular its ability to reveal how

participantsthem

selves analyze prior talk in a way relevant to the activities they are

engaged in, there are limitations to it. For exam

ple, with it it is difficult to

determine precisely how

participants attend to utterances as they are beingspoken. T

he treatment that a bit of talk gets in a next utterance m

ay be quitedifferent from

the way in w

hich it was heard and dealt w

ith as it was

spoken; indeed, rather than presenting a naked analysis of the prior talknext utterances characteristically transform

that talk in some fashion - deal

with it not in its ow

n terms but rather in the w

ay in which it is relevant to

the projects of subsequent speaker. Thus w

hile subsequent utterances canreveal crucial features of the analysis participants are m

aking of prior talkthey do not show

how participants hear the talk as it is em

erging in the firstplace, w

hat they make of it then, and w

hat consequences this has for theiractions, not in a next turn, but w

ithin the current turn. From

anotherperspective it can be noted that the stream

of speech is highly organized insyntactic and other w

ays. What, if any, consequences does such structure

have for the organization of action within a turn of talk; for exam

ple when

the utterance manifestations of a noun phrase em

erge within the stream

ofspeech, can the distinctive properties of such a structure (including thesyntactic fram

ework it displays, and the ordering of elem

ents within it) be

used by participants as a resource for the organization of their interactionw

ith each other?In brief it w

ould be valuable to begin to uncover the types oforganization that a strip of talk provides, not sim

ply for subsequent talk,but for the organization of action as it is being spoken.

2 Data and T

ranscription

We w

ill investigate in some detail sequences of conversation recorded on

audio and videotape. The tapes are from

a larger sample of data recorded

in a range of natural settings. The data to be exam

ined here are drawn

largely from a fam

ily dinner, a

backyard picnic, and a telephone call between tw

o college students. 2 Talk is

transcribed through use of the Jefferson transcription system (S

acks,S

chegloff and Jefferson 1974:731-733). The follow

ing are the featuresm

ost relevant to the present analysis:

-Italics indicate som

e form of em

phasis, which m

ay besignaled by changes in pitch and/or am

plitude.

- A

Left B

racket connecting talk on separate lines marks the point at

which one speaker's talk overlaps the talk of another.

-C

olons indicate that the sound just before the colon hasbeen noticeably lengthened.

- A

dash marks a sudden cut-off of the current sound.

- Intonation

: Punctuation symbols are used to m

ark

intonation changes rather than as gramm

atical symbols:

∞ A

period marks a falling contour.

∞ A

question mark indicates a raising

contour.∞

A com

ma indicates a falling-raising

contour.

-N

umbers in parentheses m

arks silences in seconds andtenths of second's.

-A

series of `h"s preceded by a dot marks an Inbreath.

-D

ouble parentheses enclose material that is not part of the

talk being transcribed, for example a com

ment by the transcriber if

the talk was spoken in som

e special way.

-A

degree sign ϒ ϒ indicates that the talk following it is

spoken with noticeably low

ered volume.

----------

2. For more detailed analysis of the data and the m

eans usedto obtain it see C

. Goodw

in (1981).

3 Assessm

ents

One activity that both speakers and

recipients perform w

ithin theturn at talk is evaluating in som

e fashion persons and events being

described

within their talk. T

he following provid

e examples of such

assessmen

ts: 3

(1) G. I 26:22:40

(2) 6.84:10:30

Eileen

:

Cu

rt:->

This guy had

, a beautiful, thirty two O

:lds.

In both cases speakers p

reface descrip

tive nouns w

ith the word

"beautifu

l" and thu

s evaluate the p

henomena referenced

by thosenou

ns (i.e., in #1 E

ileen assesses the "Irish Setter" she is talkingabout by d

escribing it as "beautiful").

The w

ord "assessm

ent" can in fact be used

to refer to a range ofevents that exist on analytically d

istinct levels of organization. Inview

of this some d

efinitional issues arise:

1.T

he term

can be u

sed to d

escribe a structu

ral un

it that

occurs at a sp

ecific place in

the stream

of speech

, forexam

ple the ad

jective "beautifu

l." For clarity this sense ofthe term

, which is u

sed to d

esignate a specific, segm

entalu

nit in the stream of sp

eech can be called an assessm

entsegm

ent. Though w

e will quickly see that not all assessm

ent

----------

3. For other relevant analysis of how assessm

ents are organizedw

ithin

conversation

see C. G

oodw

in (1986), M

.H. G

oodw

in(1980), and

Pomerantz (1978, 1984a).

signals are limited

to specific segm

ental phenom

ena in thisw

ay (and m

oreover that segments that p

recede the exp

licitassessm

ent term, for exam

ple intensifiers, m

ight also be part

of the activity of assessment), being able to talk abou

t anassessm

ent occu

rring at a p

articular p

lace offers greatad

vantages for startin

g analysis of th

e larger activity ofperform

ing assessments - e.g., once an assessm

ent segment is

located an analyst can look in d

etail at the different typ

es ofaction that not only co-occu

r with this event bu

t also preced

eand

follow it. M

oreover, particip

ants themselves attend

toth

e d

istinctiv

eness

and

salien

ce of

such

seg

men

talp

hen

omen

a; for examp

le they d

istingu

ish an

assessmen

tsegm

ent from events that p

recede it, and

treat it as a place

for

heig

hten

ed

mu

tual

orien

tation

an

d

action

(a

phenomenon to explored

is detail later in this paper).

2.In

ad

dition

to

usin

g p

hen

omen

a th

at can

be

neatly

segmen

ted in

the stream

of speech

, such

as assessmen

tad

jectives, particip

ants can also disp

lay their involvement in

an assessment thou

gh nonsegmental p

henomena su

ch asintonation, and

also through recognizable nonsocial d

isplays

(M.H

. Good

win 1980). Ind

eed it som

etimes becom

es quite

difficu

lt to

precisely

d

elimit

the

boun

daries

of an

assessment." A

s a function of langu

age (in the Pragu

e senseof th

at word

) rather th

an a sp

ecific act, the activity of

assessment is not lim

ited to w

ord or syntactic level objects,

but rather, like p

rosody in an u

tterance, runs over syntactic

units. In this sense it acts much like intonation (w

hich is

----------

4. Frequen

tly the left bou

nd

ary of an assessm

ent is esp

eciallyd

ifficult to p

recisely delim

it. Note, for exam

ple, not only the

analysis later in

this p

aper of h

ow p

articipan

ts attend

tointensifiers, etc. that p

recede assessm

ent segments as d

isplaying

emerging involvem

ent in the activity of assessment, bu

t also thew

ay in which p

rior talk can "seed" a su

bsequent assessm

ent byforesh

adow

ing

that

an

evalu

ation

is abou

t to

occur

(ap

hen

omen

on th

at is being in

vestigated in

work cu

rrently in

progress).

67

indeed one principal resource for displaying evaluation) 5 vis-a-vissegm

ental phonology. 6 A display show

ing a party's involvement in

an assessment can be called an assessm

ent signal. Assessm

entsegm

ents constitute a particular subset of assessment signals. It is

however quite relevant to distinguish assessm

ent segments from

the larger class of assessment signals since they have the special,

and quite useful, property of being precisely delimited in the

stream of speech.

3.T

he term "assessm

ent" can also be used to designate a particulartype of speech act. T

his sense of the term differs from

the first two

in that emphasis is placed on an action being perform

ed by anactor, rather than on the speech signal used to em

body that action,or the particular place w

here it occurs in the stream of speech. A

nassessm

ent in this sense of the term can be called an assessm

entaction.

----------

5. With respect to the close ties betw

een evaluation and intonation notethat P

ike, in his seminal study of E

nglish intonation (Pike 1945),

argued that the principal function of intonation was to show

the attitudeof the speaker tow

ard what he w

as saying. While such a view

of thefunction of intonation is clearly inadequate as a general analysis of thew

ork that intonation does, it does capture and highlight the way in

which intonation can tie together phenom

ena being talked about, with

the speaker's alignment to, and experience of, those phenom

ena. Suchanalysis

of the

way

in w

hich intonation

can display

speaker'sevaluation of the talk being produced is m

ost relevant to the structureand organization of assessm

ent actions.

6. In his analysis of narrative Labov (1972) classifies evaluation as one

distinct element of narrative structure, but also notes that unlike other

features of narrative which occur at specific places w

ithin the overallstructure of a narrative (for exam

ple the coda occurs at the end)evaluation can pervade the narrative. S

uch analysis supports theargum

ent about the distribution of assessment signals that is being

made here.

9

Several issues relevant to the analysis of assessments on this level

of organization can be briefly noted. First, w

hile most analysis of

speech acts

has focused

on actions

embodied

by com

pletesentences or turns, assessm

ents constitute a type of speech actthat can occur in the m

idst of an utterance. Subsequent analysisin this paper w

ill investigate some of the consequences of this.

Second, a crucial feature of assessment actions is the w

ay in which

they involve an actor taking up a position toward the phenom

enabeing assessed. For exam

ple in assessing something as "beautiful"

a party publicly comm

its themself to a particular evaluation of

what they have w

itnessed. By virtue of the public character of this

display others can judge the competence of the assessor to properly

evaluate the events they encounter (such a process is clearly centralto the interactive organization of culture), and assessors can be heldresponsible for the positions they state. T

hird, in so far asassessm

ents make visible an agent evaluating an event in his or her

phenomenal w

orld, they display that agent's experience of theevent, including their affective in

volvemen

t in the referent beingassessed. A

ffect displays are not only pervasive in the production ofassessm

ents, but also quite central to their organization. Moreover,

public structures such as this, that display the experience of oneparticipant, also provide resources for the interactive organizationof c

o-experience,

a process

that can

be accom

plished and

negotiated in fine detail within assessm

ents.

4.A

ssessment actions are produced by single individuals. H

owever

(as will be investigated in som

e detail in this paper) assessments

can be organized as an interactive activity that not only includesm

ultiple participants, but also encompasses types of action that are

not themselves assessm

ents. This can be called an assessm

entactivity.

Within

this activity

individuals not

only produce

assessment actions of their ow

n but also monitor the assessm

entrelevant actions of others (M

.H. G

oodwin 1980), and indeed

dynamically m

odify their own behavior in term

s of both

98

5.

what they see others are doing, and the recognizable structure of

the emerging assessm

ent activity itself (a topic to be explored indetail later in this paper).

Finally the w

ord assessable will be used to refer to the entity

being evaluated by an assessment.

In subsequent analysis the context in which the w

ord "assessment" is

being used will usually indicate w

hich of the several senses of the termnoted above is relevant at that point. T

herefore these distinctions will not

be marked in the text unless necessary.

4 Assessm

ents that Precede A

ssessables

What consequences does the fact that a speaker doesn't just describe

something, but also does an assessm

ent of it, have for how that tallk is to

be heard and dealt with by recipients? T

o start to investigate this issue we

will look at #1 in som

e detail. For com

pleteness a full transcript of thissequence w

ill now be provided. H

owever to m

ake the presentation of theanalysis as clear as possible sim

plified extracts from this transcript w

illthen bw

used to illustrate specific phenomena.

(1) G.126:22:40

4.1 Using an A

ssessment to Secure R

ecipient Co-P

articipation

Returning to the question of how

speaker's assessment m

ight beconsequential for recipients' action it can be noted that in #1 just after thenoun phrase containing the assessm

ent, one of Eileen's recipients, D

ebbie,responds to w

hat has just been said with an elaborated "A

h::".

(1) G.126:22:40

By placing an assessm

ent in her talk speaker secures an imm

ediatesubsequent assessm

ent from a recipient. M

oreover, though the way in

which she pronounces her "A

h: ::" Debbie coparticipates in the evaluative

loading of Eileen's talk, and indeed m

atches the affect display contained inE

ileen's assessment w

ith a reciprocal affect d

isplay. T

he talk marked

with the assessm

ent is thus not treated simply as a description, but rather

as something that can be responded to, and participated in, in a special

way.Further insight into w

hat this might m

ean from an organizational point

of view can be gained by exam

ining the sequential structure of this talkin m

ore detail. It can be noted, first, that recipient's action does not occurat the end of speaker's current turn-constructional unit, the characteristicplace for recipient response, but rather at a point w

here her currentsentence has recognizably not reached com

pletion. Structurally, the

1110

assessments of both speaker and recipient are placed in the m

idst of aturn-constructional unit. 7

4.2D

ifferential Treatment of Talk as it E

merges and W

hen itR

eaches Com

pletion

The issue arises as to w

hat relevance such sequential placement has for the

organization of action within the turn. For exam

ple does access to multiple

places to operate on the same strip of talk provide participants w

ithresources for the organization of their action that they w

ould not otherwise

have, and if so how do they m

ake use of these resources? One w

ay toinvestigate this issue is to look at how

this talk is treated when it does

eventually come to com

pletion. Looking again at the data it can be seen

that at its completion E

ileen's talk is not dealt with as an assessable but

rather as a laughable. Moreover such treatm

ent of this talk was in fact

projected for it before it began (arrows m

ark points of laughter in thepreface, clim

ax and response sequences):

----------

7. For more detailed analysis of how

assessments contrast w

ith continuersin term

s of their precise placement relative to the talk of another see C

.G

oodwin (1986).

13

Com

ponents of this sentence are thus dealt with In one w

ay as it emerges

through time, w

hile the sentence as a whole Is treated in a different fashion

when

it reaches

completion.

Schegloff

(1980) has

argued that

onesystem

atic issue posed for recipients of extended sequences of talk isw

hether to operate on a current piece of talk in its own right or treat it as a

preliminary to som

ething else. Here w

e find the participants able to dealw

ith a single piece of talk in both ways. B

y marking the description of the

dog as an assessable speaker was able to extract it from

its embedded

position within the story as a w

hole for treatment on its ow

n terms.

How

ever in that that description occurred at a point where speaker's

sentence was recognizably incom

plete, the not-yet-actualized tying of thistalk to relevant further talk. is also an operative feature of its structure, w

iththe effect that the larger sentence rem

ains something to be returned to after

the assessment activity has been brought to com

pletion. Within this single

utterance the participants are thus able to perform a range of

12

----------

different interactive activities, and deal with the talk that it contains in

distinctive, separable ways.

4.3 Pre-P

ositioned Assessm

ent Adjectives as G

uides for Hearers

Let us now

examine in m

ore detail the interactive organization of the nounphrase itself, the w

ay in which its com

ponents might be attended to as it

emerges

through tim

e. It

can be

observed that

within

it speaker's

assessment term

occurs in a particular position relative to the object beingassessed, i.e., it occurs before that object. T

hus by the time the object itself

emerges recipients have been alerted to hear it in a particular w

ay. The

issue arises as to whether recipients do in fact track the em

erging structureof a noun phrase on this level of detail. Is it the case that at the com

pletionof the w

ord "beautiful" a recipient will deal w

ith the next words to be

spoken in a different way than he w

ould have before hearing this term?

Features of these data not yet examined provide som

e evidence that indeedrecipients do deal w

ith the interactive import of em

erging talk on this levelof detail. Just after saying "beautiful" speaker hesitates. P

aul, the partyw

ho experienced with teller the events being described, appears to

interpret this hesitation as the beginning of a word search; just after it he

provides the projected next item in speaker's talk, the w

ords "Irish setter,"beginning an instant before speaker herself says this. H

owever, Paul does

not simply speak these w

ords; rather through his actions while speaking he

makes visible an alignm

ent tow

ard them that is congruent w

ith theassessm

ent just made by the speaker. H

is talk is produced in a lowered

"reverent" tone and while speaking Paul perform

s a prototypical nonvocalassessm

ent marker, a lateral head shake. 8 Indeed this action is escalated

during Debbie's receipt of the assessm

ent when he closes his eyes and

performs an even larger head shake over her "A

h:::,":

8. For m

ore detailed analysis of the way in w

hich such a headshake isused as an assessm

ent marker see M

.H. G

oodwin (1980).

Thus in the very next m

oment after E

ileen says "beautiful" Paul treats as anassessable w

hat is about to be described in the not-yet-completed noun

phrase. Moreover this m

arks a definite change in his alignment to that

phenomenon. W

hen, in asking Eileen to tell the story, he first m

adereference to "the dog" he did not orient to it as an assessable.

What happens here is also relevant to the analysis of affect as an

Interactive phenomenon. It w

as noted earlier that Debbie reciprocated the

affect display made available by E

ileen's assessment. W

e now find that

Paul does this as w

ell. Eileen's assessm

ent thus leads to a sequence ofaction in w

hich three separate parties coparticipate in the experienceoffered by the assessm

ent through an exchange of affect displays. These

data also demonstrate how

evaluative loading is not restricted to specificsegm

ents within the stream

of speech, but instead can accrue to a sequenceof rather heterogeneous phenom

ena (for example the noun "Irish S

etter"and the non-lexical "A

h:::") and can even bridge actions performed by

separate speakers (i.e., Paul's head

1415

shake encompasses not only the joint production of "Irish Setter" but also

Debbie's subsequent "A

h:::"). 9 In brief,

while

on the

one hand

assessments constitute a m

ode of interaction that can occur within

utterances, indeed within subcom

ponents of utterances, on the other handthey also provide an exam

ple of an activity structure that can seamlessly

span multiple utterances, and even utterance: by different speakers.

The follow

ing provides further information about how

the activity ofassessing w

hat is being said might provide organization for the interaction

of participants within relevant descriptive units, such as the utterance

manifestations of noun phases. H

ere, even though the original descriptionof the ice cream

is responded to as an assessable (lines 1 and 2), when

speaker, after describing the machine used to m

ake it, returns to the icecream

itself in line 9, recipient does not display any heightened alignment

to it. Speaker then interrupts the noun phrase in progress before it has

reached a recognizable completion and redoes it, only this tim

e placing thew

ord "homem

ade" before the type of ice cream. Just after this w

ord, overthe second production of "peach" recipient begins to treat the talk inprogress of as an assessable:

----------

9. For other analysis of actions spanning m

ultiple speakers see Ochs,

Schieffelin and Platt (1979).

The second version of "peach" is treated by recipient in a w

ay that the firstw

asn't, and this change in alignment appears to be responsive to the details

of the way in w

hich speaker organizes her emerging description. First, by

interrupting that talk before it has reached a point of recognizablecom

pletion speaker shows recipient that for som

e reason it is no longerappropriate for that talk to continue m

oving towards com

pletion. What

speaker does next, in part by virtue of its status as a repair of the talk justm

arked as flawed, provides som

e information about w

hat she found to beproblem

atic with the earlier talk. Insofar as the second version differs from

the first primarily through the addition of the w

ord "homem

ade" that termis m

arked as in some sense essential for proper understanding of the

description in progress. How

ever, recipient has already been told in line 1that the ice cream

was hom

emade. T

hus speaker is not telling recipientsom

ething new, but instead inform

ing her that something that she already

knows has not yet been taken proper account of. B

y taking up the same

alignment to this new

version of the description that she gave to its firstproduction in "hom

emade," speaker attends to the repair as having

precisely this import. In brief it w

ould appear that the problem being

remedied w

ith the repair lies not so much in the talk itself as in the w

ay inw

hich recipient is visibly dealing

1617

19

with it. M

oreover speaker is able not only to see this problem but to

initiate action leading to a remedy of it w

hile the description itself is stillin progress. S

uch events enable us to see in greater detail some of the

ways in w

hich concurrent operations on talk ate sustained and shown to be

relevant through an active processes of interaction between speaker and

recipient as the talk is being spoken.

5 Post-P

ositloned Assessm

ents

In the data so far examined the assessm

ent term and the phenom

enonbeing assessed have been packaged together w

ithin a single unit, forexam

ple within a single noun phrase. It is, how

ever, , possible to performthese activities separately. F

or example, in the follow

ing "asparagus pie"is introduced in a first sentence and then it is assessed in a second: 10

----------

10. Constructions such as this, in w

hich an entity is introduced in a firststructure, and then com

mented on in a second, have been the subject

of extensive analysis from a num

ber of different perspectives. Thus

linguists have studied such structures both in terms of syntactic

processes such as left dislocation (Gundel 1975; R

oss 1967), and interm

s of how topics, and com

ments on those topics, are organized w

ithrespect to the contrast betw

een "given" and "new" inform

ation (Chafe

1976; Li and T

hompson 1976). It should how

ever be noted that while

organizing information is a very im

portant aspect of the discourseorganization of such constructions, inform

ation managem

ent isnonetheless only one of a range of functions that such structures canperform

(Silverstein 1976). Thus, as w

ill be seen later in this paper, inm

any cases recipient collaborates in the assessment, operating on it

even before speaker has explicitly stated her position. In addition tom

arking the salience of different types of information, such a

structure thus invokes a framew

ork for

----------

heightened mutual focus on, and coparticipation in, the tal

containing the

assessment.

Though

the current

analysiem

phasizes the organization of participation structures, rather thathe transfer of inform

ation, it seems quite com

patible with th

emphasis in previous analysis on foregrounding the m

aterial in thcom

ment or proposition. R

ecently students of discourse (Duran

and Ochs 1979; O

chs and Schieffelin 1983a, 1983b) have begun t.investigate

the pragm

atic organization

of left-dislocation

examining phenom

ena such as how the `R

eferent + P

ropositionstructure can be used to organize and focus recipient's attentionthe w

ay in which such structures m

ight be fruitfully investigateas discourses (i.e., sequences of com

municative acts), rather tha

as single syntactically bound units (Ochs and S

chieffelin 1983aand

how,

as a

multi-purpose

gramm

atical construction

left-dislocation can be used to provide a warrant for a speaker

current claim to the floor (D

urand and Ochs 1979). S

uch pragm

atic focus

is quite

consistent w

ith the

analysis bein

developed here. The present data provide an opportunity to expan

the dimensions and fram

es of reference that have so far beeem

ployed to study structures of this type. On a m

ore general levew

e think that it is quite important that study of the functiona

organization of linguistic and discourse structure not be restricteto

issues of

information

managem

ent, but

also include

thm

ultifaceted activities, pragmatic functions and participatio

structures that are invoked through talk.

18

20Hare the assessm

ent occurs after the assessable has been made

available11 and is the only activity done in the speaker's second sentence.

The ability to perform

assessments in this fashion is useful to participants

in a number of different w

ays. For exam

ple, with such a structure

participants are able to assess phenomena that w

ould not fit neatly within

a single

unit. In

the follow

ing speaker

has provided

an extended

description of a movie she has seem

5.1P

ost-Positioned A

ssessments as Techniques for D

isplayingC

losure

A

first observation

that can

be m

ade about

such post-positioned

assessments is that, by m

oving to the assessment, speaker show

s thatthough her talk is continuing, a m

arked structural change has occurred init. L

ooking again at #5 it can be observed that when speaker begins the

assessment she is no longer describing events (here incidents in the

movie), but instead com

menting on the description already given:

The issue arises as to how

actions such as these are perceived,attended to and participated in by recipients.

----------

11. Where the assessm

ent occurs in the stream of speech relative to the

assessable is attended to in the fine detail within these utterances.

Thus in #l, in w

hich the assessment preceded the assessable, the

clause containing the assessment w

as introduced with "this" (i.e.,

"this beautiful Irish Setter"), which established its upcom

ing referentas an available object for com

mentary, w

hile in #4 the anaphoric term"it" presupposes the prior establishm

ent of the referent as availablew

ithin the discourse.

Such a shift from

Description to A

ssessment of D

escribed Events in fact

constitutes one of the characteristic ways that speakers begin to exit from

astory. H

ere Hyla does not end her story but instead begins to tell N

ancym

ore about what happened in the m

ovie. How

ever the way in w

hich sheresum

es the telling in fact supports the possibility that participants doattend to assessm

ents as marking a m

ove toward closure. A

fter Nancy

produces her own assessm

ent Hyla does not, as she had after earlier

continuers and brief assessments, produce a next event in the story. Instead

she follows recipient's assessm

ent with another one of her ow

n. Hyla then

interrupts this assessment before it reaches com

pletion and marks her

return to the description of the movie w

ith a misplacem

ent marker, "oh."

12

Thus the resum

ption of the telling

----------

12. See H

eritage (1984b) for more detailed analysis of how

the particle`oh' functions w

ithin interaction.

21

2223

is shown to be a m

isplaced activity, rather than one that would follow

unproblematically from

the assessment activity then being engaged in.

6 Perform

ing an Assessm

ent as a Structured Interactive Activity

Looking

now

at the

structure of

the sentences

used to

constructpost-positioned assessm

ents in #4 and #5, it can be noted that despitedifferences

in the

words

used a

similar

format

is found

in both

assessments:

[it] + [copula] + [adverbial intensifier] + [assessment term

]

A first observation that can be m

ade about this format is that it seem

sto reflect a division of activity w

ithin the utterance, with the first part of

the sentence being occupied with referencing the assessable and the

second, specifically the material after the verb, w

ith the activity ofassessm

ent itself. Moreover the w

ay in which each utterance is spoken is

consistent with such a possibility. In both cases the speech quality of the

assessment term

itself is heightened through noticeable lengthening ofsounds w

ithin it. Such enhancement of the talk is absent from

the first partof the utterance but begins to em

erge at the beginning of the adverbialintensifier, w

hich in both cases receives additional stress in addition tolengthening of sounds w

ithin it. In brief both the semantic organization of

these sentences, and the way in w

hich they are spoken, seem to reveal a

movem

ent toward heightened participation in the activity of assessing by

speaker, as the sentence unfolds.

Looking at these data from

a slightly different perspective it can alsobe noted that speaker's heightened participation in the activity ofassessm

ent begins before the assessment term

itself,

with the intensifier .13 E

arlier it was seen that im

mediately upon the

occurrence of an assessment adjective recipients could begin to treat the

talk to follow as an assessm

ent. This raises the possibility that by attending

to the pre-positioned intensifier recipients of sentences of the type nowbeing exam

ined might be able to align them

selves to the emerging talk as

an assessment before the assessm

ent term itself is actually produced.

Indeed when the actions of hearers to these utterances are exam

ined it isfound that

----------

13. The intensifier is clearly part of the assessm

ent activity and it would

be wrong to suggest that the assessm

ent does not begin until theadjective explicitly states an evaluation. It is how

ever quite useful todistinguish the intensifier from

the assessment adjective in order to

demonstrate how

participants collaboratively work tow

ard achievingheightened

mutual

focus over

the assessm

ent adjective.

The

distinctions at the beginning of the paper between a

ssessmen

tsegm

ents and assessment activity w

ere drawn precisely to deal w

ithsituations such as this. T

he intensifier is an assessment segm

ent in itsow

n right,

but one

that can

be clearly

distinguished from

the

assessment adjective that follow

s it.

24

in both cases recipients start to produce an assessment of their ow

n just asthe intensifier com

es to completion:

Thus at the point w

here speaker actually produces her assessment term

recipient is simultaneously providing her ow

n assessment of the sam

em

aterial. Such activity has a num

ber of consequences for the presentanalysis. First, it provides a clear dem

onstration of how the production of

an assessment can constitute a social activity involving the collaborative

action of multiple participants. Second, the placem

ent of recipient's actionsupports the possibility that she is tracking in rather fine detail both theem

erging structure of speaker's sentence, and the activity that speaker isprogressively entering. It w

ould thus appear that subcomponents of

speaker's utterance, such as the intensifier, as well as the details of its

sound production, contribute to the interactive organization of the actionsof speaker and hearer in the activity they are jointly engaged in. In thissense the em

erging structure of speaker's utterance, and the details of thew

ay in which it is spoken, constitute one aspect of the context that

recipients are actively attending to within the turn as consequential for the

organization of their own actions. M

oreover that context, and the utteranceitself,

are intrinsically

dynamic,

and are

attended to

as such

byparticipants.

By

making

projections about

the future

course of

anutterance, recipients dem

onstrate that they are not dealing with it as a

monolithic w

hole, or simply as a static string of sym

bolic components tied

together through

25

syntax, but rather as a process that emerges through tim

e and carries with It

an expanding horizon of projective possibilities that are relevant to theactions that recipient m

ight engage in while acting as a hearer to the

utterance. 14

6.1 Extended O

verlap

The assessm

ents produced by recipients in these data take the form of

complete substantial sentences in their ow

n right. In that they are placednot after speaker's action has com

e to completion, but w

hile speaker'sassessm

ent is also in progress a state of extended simultaneous talk by

different participants results (i.e., in length and structure something m

orethan overlap of ongoing talk by continuers or brief assessm

ent tokens suchas "oh w

ow".) 15

----------

14. For other analysis of how the w

ay in which recipient projections about

the future course of a sentence are relevant to the organization of theirinteraction w

ith speaker see Jefferson (1973) and Sacks, Schegloff andJefferson (1974). F

or analysis of how deictic term

s dynamically

modify em

erging context as an utterance unfolds see Hanks (1986).

15. For m

ore detailed analysis of the interactive organization of briefassessm

ents such as "oh wow

", and the way in w

hich they contrastw

ith continuers such as "uh huh" see C. G

oodwin (1986). A

ssessments

by recipient can range from fully referential and predicational ones

down to relatively desem

anticized displays of empathy, etc., that lack

an explicit

referent and

evaluation, but

do display

affectiveinvolvem

ent in principal speaker's statement. Indeed som

e evidencesuggests that this division of labor, w

ith principal speaker providingthe referent that recipient is responding to, m

ay be attended to in finedetail in the sequential organization of concurrent assessm

ents. Thus,

when recipients produce brief concurrent assessm

ents, speakers may

delay entry into subsequent units of talk until the assessment has run

its course, so that the a new

26

This is not, how

ever, treated as a situation requiring a remedy; 16 for

example,

neither party's

talk contains

restarts, hitches,

or other

perturbations, or indeed any displays that problems exist w

ith the currentstate of talk. M

oreover, if the analysis developed above is correct, thissim

ultaneous talk is not the result of an accidental failure to achieve propercoordination but rather som

ething that the participants have systematically

achieved through close attention to the emerging structure of the talk and

activity in progress. What happens here thus provides further support for

the possibility that assessments do indeed constitute w

ays of analyzing andoperating on talk that can be perform

ed while that talk is still in progress.

Indeed it appears that constraints which elsew

here exert quite powerful

influence on the sequential organization of talk, for example an orientation

to one but only one party speaking at a time, can be relaxed for

assessments. It w

ould thus appear that, in a number of different w

ays, theactivity of assessing som

ething provides participants with resources for

performing concurrent operations on talk that has not yet com

e tocom

pletion.

6.2D

ifferential Access as an O

rganizing Feature of C

oncurrentA

ssessments

Though the talk of both speaker and recipient in #4 and #5 is assessing the

same m

aterial each party in fact says rather different things. Is suchvariation sim

ply haphazard or does it reveal further aspects of thephenom

ena the participants are orienting to as relevant for the organizationof their activity? L

ooking more carefully at precisely w

hat is said it can benoted that in its details the talk of each party attends to the access each hasto the phenom

ena being assessed. For example, H

yla with her initial

----------

referent does not occur while the assessm

ent is still in progress (C.

Goodw

in 1986).

16. For analysis of how

participants can negotiate speakership within

overlap see Jefferson (1973) and Schegloff (in press).

27

"it's" makes reference to an actual m

ovie she has seen, and assesses it inunequivocal term

s. Nancy, how

ever, by saying "That sounds so: goo::d?"

attends to what she is assessing as being available only through H

yla'scurrent description of it. S

imilarly in #4 D

ianne, who depicts herself as

having directly experienced in the past the pie she is now describing, m

akesreference to that specific pie. H

owever, C

lacia, by putting her assessment in

present tense, deals not with the specifics of that particular pie, but rather

with it as a class of phenom

ena that the pie currently being describedinstances. A

mom

ent later, after Dianne has described the pie in m

ore detail,C

lacia says "Oh: G

o: d that'd be fantastic." Here by constructing her

assessment in conditional tense, she again m

akes visible in her talk thelim

ited access17 she has to the phenom

ena she is assessing. Thus one of the

reasons that the assessments of the separate participants differ from

eachother is that each has different access to and experience of the event beingassessed. T

his feature provides organization for a range of phenomena

implicated in the construction of each utterance, such as the choice of

particular words and verb tenses. B

y constructing their assessments in this

fashion participants also attend in detail to how they have been organized

relative to each other by the telling in progress. For exam

ple the differentpositions of describer and describee are show

n to remain relevant even

when both are assessing in a sim

ilar fashion the events which have been

described. In brief, despite their apparent simplicity, assessm

ents show a

view of the assessable as som

ething perceived by an actor who both takes

up a particular alignment to it and sees the assessable from

a particularperspective, one that m

ay be quite different from that of a coparticipant w

hois sim

ultaneously assessing the same event.

It can also be noted that insofar as both the assessable, and the activityof assessing it, em

erge as talk unfolds through time,

----------

17. For other analysis of how the structure of talk displays the type of

knowledge that speaker has of the event being talked about see M

.H.

Goodw

in (1980), Hanks (1986) and Pom

erantz (1980).

28

differences in participant perspective have a temporal organization as

well. T

o note one simple exam

ple, in #1 Paul, the party w

ho saw w

ithspeaker

the dog

being assessed,

was

able to

act just

after the

pre-positioned assessment adjective w

as spoken. How

ever at that pointthe assessable itself w

as not yet available to speaker's addressed recipient,D

ebbie, and indeed her response occurred only after the assessable hadbeen described. Issues of em

erging perspective within the activity of

assessing are clearly relevant to other processes as well, such as w

ays inw

hich recipients project from an intensifier that an assessm

ent is about tooccur.

6.3 Making V

isible Congruent U

nderstanding

Though the talk of the separate parties show

s that each is viewing the

assessable from a different perspective, in other w

ays the assessments

produced by each seem to have an underlying sim

ilarity. For example in

#4 both speaker and recipient assess asparagus pie positively. Thus w

iththeir assessm

ents the participants are able to display to each other thatthey evaluate the phenom

ena being assessed in a similar w

ay. Moreover

by virtue of the way in w

hich each assessment takes into account the

distinctive position of the party making it, these sim

ilar evaluation: areshow

n to result from independent appraisals of the phenom

ena beingassessed. In essence w

ith the assessments the participants show

each otherthat, on this issue at least, their m

inds are together, they evaluate thephenom

ena being discussed is a similar w

ay.

Assessm

ent: reveal not just neutral objects In the world, but an

alignment taken up tow

ard phenomena by a particular actor. M

oreoverthis alignm

ent can be of some m

oment in revealing such significant

attributes of the actor as their taste, and the way in w

hich they evaluatethe phenom

ena they perceive. It is therefore not surprising that displayingcongruent understanding can be an

29

issue of some im

portance to the participants. 18 Further support for activeattention to such an issue is found w

hen a visual record of the actions ofthe participants in #4 is exam

ined. As C

lacia produces her assessment she

nods toward D

ianne:

With her nods C

lacia proposes that the talk she is producing, and theposition taken up through that talk, is in agreem

ent with D

ianne's. Indeed,taken as a w

hole the actions she performs here provide a strong display of

agreement. First, w

ith the content of her utterance she states a view of the

assessable that is compatible w

ith Dianne's. S

econd, with her nods she

marks that talk nonvocally as an agreem

ent. Third, she perform

s this actionnot after hearing D

ianne's assessment but at the very m

oment it is being

spoken. It is of course true that the talk so far produced provides materials

(for example the intensifier) that strongly suggest, and perhaps actually

project, a favorable assessment. N

evertheless at the point where C

laciaacts, D

ianne has not officially stated a position. By placing her talk w

hereshe does C

lacia argues that her way of view

ing the assessable is so attunedto D

ianne's that she is prepared to both comm

it herself to a position, andcategorize that position as an agreem

ent without

----------

18. For other analysis of displaying congruent understanding see C

.G

oodwin (1981:114-116) and Jefferson (1983).

3130 .

actually hearing Dianne's. 19 T

hus with the content of her talk, nonvocal

displays about it, and its sequential placement, C

lacia argues strongly thather view

of the assessable is congruent with D

ianne's.

It is being argued that recipients produce concurrent assessments by

making projections about events w

hich have not yet occurred. If this isindeed the case then it w

ould be expected that on some occasions the

projections made by recipients w

ould turn out to be inaccurate. Rather than

providing evidence against the position being argued in this paper, such anevent w

ould constitute strong evidence that recipients are in fact engagedin the activity of anticipating future events on the basis of the lim

itedInform

ation currently available to them. T

he following provides an

example of how

a recipient's projection of an emerging assessm

ent can beerroneous, w

ith the effect that the concurrent appreciation being displayedby recipient is quite inappropriate to w

hat speaker turns out to in fact besaying:

----------

19. It may be noted that the placem

ent of this strong agreement is alm

ostthe m

irror image of one of the w

ays in which im

pending disagreement

is displayed sequentially. Pomerantz (1984a) describes how

recipientsprepared to disagree frequently delay a response to w

hat has just beensaid.

In the beginning of this sequence Em

ma describes a "D

AR

LIN

G D

RE

SS"that she has m

ade and Nancy replies to her description w

ith concurrentassessm

ents in lines 5, 13 and 20. In lines 19 and 21 Em

ma starts to m

ovetow

ard a recognizable assessment, follow

ing `was' w

ith the intensifier 'so'.R

ight after this happens Nancy in line 22 starts to coparticipate in the

assessment by producing an elaborated, appreciative "A

h : : : : : :". The

positive affect displayed by Nancy is quite congruent w

ith the favorablew

ay that the dress has been described in the sequence until this point.H

owever it turns out that E

mm

a is now m

oving her talk to a negativedescription of the w

eather on her trip, i.e. it "wz so h

o:t there" that theydidn't even stay for dinner. B

y relying on cues of the type being analyzed inthe present paper N

ancy has attempted to align herself to an assessm

entbefore it is

32

actually produced, but the talk has progressed in ways quite incom

patiblew

ith her

projection of

it, w

ith the

effect that

she is

respondinginappropriately to w

hat Em

ma is saying. S

uch data provide a strongdem

onstration of how projecting w

hat another is about to say so as toconcurrently coparticipate in it constitutes a contingent accom

plishment.

Fortunately the em

erging structure of interaction provides resources form

oving past, and attempting to recover from

, such a faux pas, and in line33 w

e find Nancy once again producing a concurrent assessm

ent toE

mm

a's description of the weather, only this tim

e her response is quiteappropriate.

Returning now

to example #4 w

e find that Dianne also perform

s anum

ber of relevant nonvocal actions. As she produces the assessm

entterm

she lowers her head into a nod w

hile simultaneously lifting her

brows into a m

arked eyebrow flash. T

hese actions are preceded bym

ovement of her head and upper body in a w

ay that shows heightened

orientation toward recipient over the intensifier:

Dianne's nonvocal behavior like her talk seem

s to display aprogression tow

ard heightened involvement in the assessm

ent as

33

her utterance unfolds. 20 These actions becom

e most intense over the

assessment term

itself, and indeed at this point in the talk quite a range ofboth vocal and nonvocal action is occurring. T

he ensemble of things done

over the assessment does not, how

ever, seem a collection of separate

actions, but rather integrated elements of a single interactive activity of

assessment. M

oreover the visible behavior of speaker, as well as the

unfolding structure of her talk and recipient's participation in that talk,seem

to demonstrate system

atic movem

ent toward this point through tim

e.In essence one seem

s to find here an organized activity that participantsrecognize and system

atically bring to a recognizable climax.

6.4 Bringing A

ssessment A

ctivity to a Close

Having seen how

participants attend to the structure of assessments as an

activity so as to collaboratively bring that activity to a recognizable peak orclim

ax, we w

ill now look at som

e of the ways in w

hich movem

ent away

from such a point m

ight be accomplished. O

ne way to approach this issue

is to

ask "W

hat can

participants do

next?" S

ome

actions w

ithinconversation have the property of being nonrepeatable (see for exam

ple theanalysis of sum

mons-answ

er sequences in Schegloff (1968)), i.e.,

----------

20. In that recipient's nod begin after speaker's body displays heightenedorientation tow

ard her over the intensifier, one might be tem

pted toargue that the nods are solicited or at least triggered by the bodym

ovement speaker has just m

ade. How

ever it seems m

ore accurate tosay that recipient is responding to the em

erging activity of assessment,

something visible in a range of different w

ays, e.g., the intensifieritself, its placem

ent in the talk so far produced, the way in w

hich it isarticulated, the visible actions of speaker's body relevant to it, etc.A

rbitrarily segregating interactive events in terms of w

hether they areproduced vocally or nonvocally seem

s neither helpful analytically, norto accurately reflect w

hat the participants are doing.

once they have been validly performed they cannot be im

mediately

redone. Assessm

ents, however, are repeatable. M

oreover while som

erepeatable actions are used to progressively operate on new

material, for

example a series of questions in a m

edical interview, so that each instance

of a similar action actually deals w

ith separate phenomena, a participant

can make continuing assessm

ents of the same assessable. F

or example

after producing "Irish setter" as an assessment P

aul continues to displayinvolvem

ent in the activity of appreciating it. First, he coparticipates in theassessm

ent made by D

ebbie's "Ah:::," by producing an assessm

ent headshake in tim

e with it. T

hen, as his eyes return to Debbie, he uses an

assessment form

at similar to that found in #4 and #5 to extract the

assessment from

the embedded

3435

position it occupies in Eileen's sentence, and m

ake it the exclusive focus ofa new

sentence of his own:

Paul's utterance is also accompanied by assessm

ent head shakes. Thus as

Eileen returns to the substance of her story Paul m

akes

36

use of both vocal and nonvocal phenomena to produce repeated

assessments of the assessable. 21

----------

21. Paul's continued assessm

ents co-occur with E

ileen's return to herstory.

Itcan be noted that P

aul's talk is produced with noticeably low

eredvolum

e and that he does not orient toward the sam

e recipients Eileen

is then gazing at. He thus seem

s to produce minim

al intrusion intoE

ileen's talk. For exam

ple with the contrast in volum

e between his

talk and Eileen's he show

s others present that though that talkoverlaps E

ileen's, it should not be heard as competitive w

ith hers, andindeed

she does

not treat

it this

way.

Moreover,

even w

hilecontinuing in the assessm

ent, Paul seem

s to remain aw

are of theem

erging structure of Eileen's talk, and to organize at least som

efeatures of his actions in term

s of it. For example as she com

es to thecom

pletion of the background material in her story he brings his

assessment activity to a close and returns his gaze to her as the clim

axsegm

ent of the story is entered (for more detailed analysis of how

participants might attend to the em

erging structure of a story toorganize even actions unrelated to the story see C

. Goodw

in (1984)).Finally, it can be noted that insofar as Paul actually participated in theevents being described by E

ileen the issue arises as to how he can act

as a recipient to her current talk (for more detailed consideration of

this issue, and the interaction it engenders, see Sacks 10/19/71 and C.

Goodw

in (1981, chapter 5). By perform

ing the extended assessment

Paul finds a way to deal w

ith the events being described in the currenttalk in a w

ay that is appropriate to his experience of them.

37

Returning to #4 it is found that just after the assessm

ent producedconcurrently w

ith Dianne's, C

lacia repeats that assessment:

How

ever during this second assessment she acts quite differently than

she had during the first. Thus the subsequent assessm

ent is spoken with

markedly low

ered volume (this is indicated in the transcript by the sm

allertype). M

oreover while speaking C

lacia actually withdraw

s from her

coparticipant.

Thus, w

hile the initial concurrent assessment w

as produced within a state

of heightened orientation toward coparticipant and the talk in progress, this

second assessm

ent is

done w

hile C

lacia is

displaying dim

inishedparticipation in the activity, and indeed seem

s to be withdraw

ing from it

It is thus found that a single assessm

ent activity can

encom

pass a

range of different types of participation. The sequencing of participation

in these data — collaborative orientation tow

ard the emergence of the

assessment, elaborated participation in it as it is actually produced and

finally a trailing off of involvement in it —

is consistent with the possibility

that

38

what is occurring here are successive stages of a single natural activity

that emerges, com

es to a climax and is then w

ithdrawn from

.

6.5 Assessm

ents as Resources for C

losing Topics

Instead of

just analyzing

these different

participation structures

assuccessive stages of an unfolding activity it is also useful to exam

ine inm

ore detail how the possibility of investing assessm

ents with different

kinds of participation might provide participants w

ith resources for theorganization of their activity. F

or example, assessm

ents are one of thecharacteristic activities used to exit from

larger sequential units in talk suchas stories and topics. Indeed one frequently finds strings of assessm

ents atsuch places. W

hen one examines precisely how

such assessments are

spoken it is found that frequently they are operating not only to exit fromw

hat was being talked about in the story to topic, but that in addition the

different participation

possibilities provided

by assessm

ents are

systematically being used to bring the heightened m

utual orientation thatsuch a focused activity has engendered to a close. A

simple exam

ple isfound shortly after the sequences analyzed in #4. In the intervening talkD

ianne has described in greater detail the asparagus pie that Jeff made:

39

As D

ianne moves from

a description of the pie to an assessment of it,

she noticeably reduces the volume of her talk w

hile simultaneously

withdraw

ing her gaze from C

lacia. Thus she has not only m

oved into adifferent kind of talk (e.g. from

description to assessment) but also

changed the nature of her involvement in that talk and the structure of her

orientation to coparticipant. Despite the apparent sim

plicity of what

Dianne has done, the changes produced are in fact rather intricate. T

hussom

e of what happens —

the move from

description to assessment, the

reduction in volume and the w

ithdrawal of gaze from

recipient — indicate

that she is proposing topic closure. How

ever even as she does this she isdisplaying heightened involvem

ent in the substance of her talk. The

assessment itself w

ith its "savoring" voice quality (achieved in partthrough the sam

e lowering of volum

e that might otherw

ise indicate move

toward closure of the sequence) and actions of her body during it, such as

the assessment head shakes, all display elaborated appreciation of w

hatshe has been talking about. In essence the actions D

ianne performs seem

both to foreshadow topic closure and to show

heightened involvement in

the topic.A

t first glance such a combination m

ight appear inconsistent or evencontradictory. H

owever to see this m

ixture of phenomena in

40

such a way is to im

plicitly assume that topics run out only because

participants lose interest in them. If a topic has in fact engrossed the

attention of those talking, this would be a very poor w

ay to end it. On the

other hand, one would not w

ant to talk about that topic forever. Thus one

might w

ant to look for ways of dealing w

ith talk in progress that showheightened appreciation of it, w

ithout however proposing that others need

continue talking about it forever. Dianne's assessm

ent has precisely theseproperties. S

he is able to show coparticipant (for exam

ple with her gaze

withdraw

al) that

she is

not aw

aiting further

talk from

her,

while

simultaneously appreciating w

hat has just been said. Indeed one of thereasons w

hy assessments m

ight be so extensively used to close stories andtopics is that they provide this m

ixture of participation possibilities fororganizing the interaction then in progress.

Some dem

onstration that the participants themselves m

ight analyze anassessm

ent such as Dianne's as including an ensem

ble of activity of thetype just described is provided by the talk C

lacia produces next. In itsproductional features this talk responds to the various elem

ents of Dianne's

talk, while ratifying the change in participation status she has proposed:

First, as C

lacia begins to speak she too withdraw

s her gaze from her

coparticipant. Second, her talk is produced w

ith not simply low

eredvolum

e but drastically reduced volume (indicated in the transcript by the

two degree signs before it.) T

he talk itself is, however. a m

arked upgradeof the assessm

ent Dianne just m

ade:

41

The exchange of affect provided by the exchange of assessm

ents givesthe w

ithdrawal the intim

acy of a parting touch, in which the character of

the apparent referent of the assessment becom

es far less important than the

shared affect and coexperience the participants display to each other. Inthese data speaker and recipient, through the details of the w

ays in which

they performed their assessm

ents, have moved aw

ay from the substance of

the topic

in progress

while

simultaneously

showing

their ongoing

appreciation of it. At the sam

e time they have dism

antled the facingform

ation that had been sustained through that talk. Insofar as no new topic

is yet on the floor, the state of disengagement w

hich has thus beencollaboratively entered

42

through this process of phased withdraw

al 22 is quite appropriate to theircurrent actions.

7 Refusal by R

ecipient to Coparticipate in the A

ssessment

In the data so far examined recipients have accepted speaker's proposals

about how the entity being assessed should be evaluated. H

owever not all

assessments are responded to in such a felicitous fashion. R

ecipients canrefuse to treat as an assessable som

ething that speaker proposes should beso treated, and in so doing call into question a speaker's com

petence toproperly evaluate the phenom

enon being assessed. Exam

ple #2, which has

not yet been examined in detail, provides data in w

hich this happens. By

looking at it we w

ill be able to investigate some of the consequences that

producing something as an assessable has for both the party m

aking theassessm

ent, and the talk in progress.T

his utterance was produced as speaker w

as beginning an extendedstory. In form

it is quite similar to # 1:

With the w

ord "beautiful" speaker marks the talk to follow

as a descriptionof an assessable. Indeed both the w

ord "beautiful" and the talk after it aregiven special salience through the com

ma intonation around "beautiful."

Moreover this talk is accom

panied by relevant nonvocal actions, includinggestural intensifiers and head m

ovements by speaker, that seem

to bothenhance

the assessable

character of

his talk

and invite

recipientparticipation in

----------

22. For m

ore extended analysis of the organization of engagement

displays and entry into disengagement see C

. Goodw

in (1981, chapter3).

43

it (for purposes of the present analysis it is not necessary to examine these

actions in detail).It w

as seen in the data examined earlier that after hearing an assessable

recipients are not only able to respond to such action but participate in it ina variety of different w

ays. Indeed less than two m

inutes earlier in thissam

e conversation Mike produced a description of a car that C

urt assessedin rather elaborate fashion:

How

ever, in #2, despite the explicit assessment term

before Curt's

description of the "thirty two O

:lds", and the intonational and nonvocalem

phasis given it, Mike does not respond to w

hat Curt has said in any

way. A

fter leaving a full half second of silence that not only provides Mike

time for response, 23 but also m

akes

----------

23. For m

ore extended analysis of how speakers analyze the absence of

response to their talk and use further talk to pursue such response seePom

erantz (1984b).

44

visible interactively the absence of such a response, Curt produces further

talk:

By providing further inform

ation about the car being described Curt show

sthat he is still aw

aiting a response to his earlier talk. Moreover the w

ordchosen

Is inform

ative about

the type

of response

he is seeking.

Specifically this term provides recipient w

ith further grounds for treatingw

hat has just been described as something to be assessed. Indeed

"original" was the very first attribute used by C

urt to assess the Cords tw

om

inutes earlier (c.f. #8).A

t this point Mike does provide a response:

Mike's nod receipts C

urt's talk but in no way assesses it. R

ather the nodseem

s to constitute a type of continuer, an action which deals w

ith the talkw

hich has just been heard as preliminary to further talk, rather than as

something to be appreciated in its ow

n right (C. G

oodwin 1986; Schegloff

1980). Insofar as Curt's talk is recognizably one of the early stages of a

story it is technically possible to analyze it in this way. H

owever, as #1

demonstrated, it is also possible to deal w

ith such talk in its own term

s,and indeed C

urt hiss formulated this description as an assessm

ent, anaction recipients can and do participate in. T

hus by responding in the way

that he does Mike show

s that he has dealt with w

hat Curt has said, w

ithouthow

ever treating it in the way that C

urt proposed it

45

should be treated. Rather by responding w

ith a continuer Mike has m

ade acounterproposal - that C

urt should move forw

ard with the story.

Such a sequence of action makes it relevant to exam

ine what

happens next. After M

ike's action Curt produces further description of the

car, and then reveals this to be not a next event in the story but additionaldem

onstration of how "original" the car w

as. When this assessm

entextends into yet another turn constructional unit M

ike turns away from

Curt and begins to search for a cigarette.

Mike does not return his gaze to C

urt for over 33 seconds, moving back

into orientation toward him

at the point where the story approaches its

climax.

These data provide som

e demonstration of how

establishing theassessable character of an object is not som

ething done by speaker alone,but rather an interactive event. T

he participation possibilities provided byassessm

ents enable participants to negotiate both the status of a proposedassessable, and the w

ay in which the talk containing it w

ill be attended to.

46

8 Assessing P

henomena E

xperienced Only T

hrough Talk

One m

ight wonder how

Mike, or any recipient w

ho hasn't actually himself

experienced the assessable being described, could be expected to evaluateit. Q

uite clearly, as the data examined earlier in this paper dem

onstrate,recipients do assess phenom

ena available to them only through a speaker's

talk. What is involved in such a process? S

ome issues relevant to this

question will be briefly noted. First, as has already been seen, recipients do

organize their assessment w

ith attention to ways in w

hich their access tothe assessable differs from

speaker's. Second, recipients m

ay choose totrust the com

petence of speaker to properly evaluate what she is treating as

an assessable. Third, it w

ould appear that the assessable character of atleast som

e phenomena can be adequately established entirely through an

appropriate description of them. F

or example neither "hom

emade ice

cream" nor "C

ord" is preceded by an explicit assessment term

(such as"beautiful”) but recipients receipt both w

ith assessments. T

his suggeststhat independent of the specifics of the particular entity being described,its m

embership in the classes of phenom

ena identified by those terms is

itself adequate grounds for finding it to be an assessable. As C

urt sayselsew

here "Any C

ord is nice."

47

to be taken for granted, but rather something to be interactively achieved.

Indeed it would appear that m

aking such assessments Is one of the

places where a participant's fine grained com

petence In a particulardom

ain of culture is not only displayed to coparticipants but challenged orvalidated by them

. Thus in these data w

e find that Mike is som

eone who

will refuse to give a "thirty tw

o O:lds" the kind of evaluation that C

urtw

ould give both it and a "Cord." Indeed, unlike C

urt, Mike is som

eonew

ho makes distinctions betw

een Cords. In essence it w

ould appear thatw

hat is at issue in showing com

petence in making such distinctions is not

simply a cognitive phenom

enon (though the processes involved may be

central to the construction and operation of "domains of know

ledge" thatanthropologists and other students of the cognitive organization of culturehave studied as static phenom

ena, and analyzed in isolation from the

detailed interaction within w

hich they become visible) but a social and

interactive process, and indeed one that can have real consequences forthe standing participants achieve vis-à-vis one another.

Looking back at #1 in light of these considerations it can be seen that

in unproblematically accepting E

ileen's assessment as som

ething that shew

ill participate in, Debbie validates E

ileen's competence to properly

evaluate the phenomena she encounters. T

hough this might seem

sounrem

arkable as to escape notice, it is quite a bit more than M

ike givesC

urt.

9 Initial Alignm

ent and Subsequent Understanding

Do any system

atic reasons exist for Curt to pursue his assessm

ents of thecar w

ith such tenacity? Curt's activity of assessing the car occurs in a

particular sequential position, in an initial "background" segment of a m

oreextended story. In #1 it w

as seen that assessments m

ade in this positionm

ight treat phenomena quite differently from

the way in w

hich they aredealt w

ith later in the story. This does not how

ever exclude the possibilitythat on

How

ever, as Curt finds out w

hen he offers a "thirty two O

lds" as is itselfan assessable, the status of any particular descriptor is not

48

some occasions assessm

ents performed on phenom

ena in such a positionm

ight nonetheless also be relevant to the larger organizational structure ofthe em

erging story. Indeed attending to the kind of work that assessm

entscould do here w

ould help us to uncover in more detail the range of

activities that participants are engaged in while attending the initiation of a

story. In Curt's story it is eventually revealed that the ex-w

ife of the owner

of this car caused its engine to blow up by stuffing a rag in its radiator

hose. After the story reaches its clim

ax the participants deal with it by

debating what w

ould be proper punishment for the ex-w

ife -- the mildest

(and only printable) suggestion being Curt's "I'd kill y'know

that'd beenough t'go after a shotgun w

ith." Quite clearly punishm

ent like thisw

ould not be appropriate for someone w

ho damaged the engine of just any

car. Rather to understand the events in the story in the w

ay thatparticipants show

that they understood them, one m

ust conceptualize thecar as an extrem

ely special, very highly valued object, one whose

destruction merits extraordinary punishm

ent. Thus, w

hen Curt introduces

the car early in the story he is faced with the task of aligning his recipients

to it in a particular way. A

process well suited to not only displaying

alignment, but securing it from

others, is the activity of assessments.

When used to introduce entities that w

ill figure prominently in a story a

noun phrase containing an assessment adjective, such as "a beautiful thirty

two O

:Ids", contains within its structure elem

ents capable of performing

two of the central tasks posed during story initiation: m

aking phenomena

available for subsequent reference24 and aligning participants to those

phenomena in an appropriate fashion. F

rom such a perspective C

urt'sattem

pt to have the car evaluated in a particular way w

ould appear to beneither idiosyncratic, nor sim

ply an attempt to rem

edy an affront to hisjudgm

ent, but rather a systematic part of the w

ork he is faced with in

beginning a story: preparing his

----------

24. For more extensive consideration of how

this issue is relevant to theorganization of stories and other m

ulti-unit turns see Schegloff

(1980:114-115).

49

recipients to understand what he is to tell them

in an appropriate way, or

at least the way that he w

ants them to understand it.

Assessm

ents are found to occur in a diverse range of sequentialpositions w

ithin talk, for example, as subordinate parts of sentences

dealing primarily w

ith other matters, in the background segm

ents ofstories, and as extended sequences w

hen stories and topics are brought tocom

pletion. The phenom

ena just noted would suggest that the assessm

entsin these apparently heterogeneous positions m

ight in fact be related to eachother. For exam

ple, the understanding of a story displayed in a sequence ofassessm

ents at

its conclusion

is intim

ately tied

to w

ays in

which

participants were led to see characters and events in the story w

hen theyw

ere first introduced. Assessm

ents thus constitute a most Im

portantresource for collaboratively building w

ithin the talk itself an interpretivecontext that w

ill utilized for the analysis of subsequent talk and action. Inbrief, despite their apparent sim

plicity assessments constitute one central

resource available to participants for organizing the perception andinterpretation of w

hat is being talked about, providing them w

ith the abilityto not sim

ply display alignment to ongoing talk, but establish and negotiate

that alignment through a system

atic process of interaction while the talk

being aligned to is still in progress.T

he data which have been investigated here have enabled us to

investigate a range of issues relevant to how assessm

ents are organized asan activity w

ithin the turn at talk. One of the very interesting things about

assessments is the w

ay in which they integrate a range of phenom

enaoccurring w

ithin the turn that are frequently studied quite separately. In sofar as assessm

ents are achieved through the collaborative action of multiple

participants they provide an elementary exam

ple of social organizationw

ithin the boundaries of the turn. At the sam

e time they constitute a key

locus for the display and achievement of congruent understanding, and thus

are quite relevant to the study of cognition as a practical, everyday activity.In addition they provide an exam

ple of how affect and the display of

emotion are organized as interactive phenom

ena. In accomplishing this

activity participants must pay close attention to w

hat other participants aredoing, the details of

what is happening in the stream

of speech, and the recognizable structureof the activity itself. T

he study of assessments thus perm

its analysis in anintegrated fashion of a range of phenom

ena relevant to the organization oflanguage, culture, cognition and em

otion in the midst of actual interaction.

References C

ited

Atkinson, J. M

axwell and John H

eritage (Eds.).

1984 Structures of Social Action. C

ambridge: C

ambridge U

niversityPress.

Button, G

raham and John R

. Lee (E

ds.).in press Talk and Social O

rganisation. Clevedon, E

ngland: Multilingual

Matters.

Chafe. W

allace L.

1976 Givenness, C

ontrastiveness, Definiteness, S

ubjects, Topics and

Point of V

iew. In Subject and T

opic. C. L

i, ed. New

York:

Academ

ic Press.

Durand, A

lessandro and Elinor O

chs1979 L

eft-Dislocation in Italian C

onversation. In Syntax and Semantics,

vol. 11, Discourse and Syntax. T

almy G

ivon, ed. New

York:

Academ

ic Press.

Garfinkel, H

arold1967

Studies in

Ethnom

ethodology. E

nglewood

Cliffs,

N.J.:

Prentice-Hall.

Goodw

in, Charles

1981 Conversational O

rganization: Interaction Betw

een Speakers andH

earers. New

York: A

cademic Press.

5051

Goodw

in, Charles

1984 Notes on Story Structure and the O

rganization of Participation. InStructures of Social A

ction. Max A

tkinson and John Heritage, eds.

Pp. 225-246. Cam

bridge: Cam

bridge University Press.

Goodw

in, Charles

1986 Betw

een and Within: A

lternative Treatm

ents of Continuers and

Assessm

ents. Hum

an Studies 9:205-217.

Goodw

in, Marjorie H

arness1980 P

rocesses of Mutual M

onitoring Implicated in the P

roduction ofD

escription Sequences. Sociological Inquiry 50:303-317.

Gundel, J.1975 L

eft Dislocation and the R

ole of Topic-C

omm

ent Structure in

Linguistic T

heory. Ohio State W

orking Papers in L

inguistics18:72-132.

Hanks, W

illiam F.

1986 The Interactive Structure of Indexical R

eference. Paper presentedat the Invited Session on `R

ethinking Context' at the 1986 A

nnualM

eeting of the Am

erican Anthropological A

ssociation.

Heath, C

hristian1986 B

ody Movem

ent and Speech in Medical Interaction. C

ambridge:

Cam

bridge University Press.

Heritage, John1984a G

arfinkel and Ethnom

ethodology. Cam

bridge: Polity Press.

Heritage, John1984b

A

Change-of-S

tate T

oken and

Aspects

of Its

Sequential

Placem

ent. In Structures of Social Action. J. M

axwell A

tkinsonand John H

eritage, eds. Pp. 299-345. C

ambridge: C

ambridge

University Press.

52

Heritage, John C

.

1985 Recent D

evelopments in C

onversation Analysis.

Sociolinguistics New

sletter 15, No.1:1-19.

Heritage, John and J. M

axwell A

tkinson1984 Introduction. In Structures of Social A

ction. J. Maxw

ell Atkinson

and John

Heritage,

eds. Pp.

1-16. C

ambridge:

Cam

bridgeU

niversity Press.

Jefferson, Gail

1973 A

C

ase of

Precision T

iming

in O

rdinary C

onversation:O

verlapped Tag-P

ositioned Address T

erms in C

losing Sequences.Sem

iotica 9:47-96.

Jefferson, Gail

1983 C

aveat Speaker.

Preliminary

Notes

on R

ecipient T

opicShiftIm

plicature. Tilburg P

apers in Language and L

iterature 30.

Labov, W

illiam1972 T

he Transform

ation of Experience in N

arrative Syntax. InLanguage in the Inner C

ity: Studies in the Black E

nglishV

ernacular. ,

ed. Pp.

354-396. Philadelphia:

University

ofP

ennsylvania Press.

Levinson, Stephen C

.1983 Pragm

atics. Cam

bridge: Cam

bridge University Press.

Li, C

harles and S. Thom

pson1976 Subject and T

opic: A N

ew T

ypology of Language. In Subject and

Topic. Charles L

i, ed. New

York: A

cademic P

ress.

Ochs. E

linor and Bam

bi B. Schieffelin

1983a Foregrounding Referents: A

Reconsideration of L

eft Dislocation

in Discourse. In A

cquiring Conversational C

ompetence. E

linorO

chs and

Bam

bi B

. Schieffelin,

eds. Pp.

158-174. B

oston:R

ougledge & K

egan Paul.

53

Ochs, E

linor and Bam

bi B. Schieffelin

1983b Topic as a D

iscourse Notion: A

Study of Topic in the

Conversations

of C

hildren and

Adults.

In A

cquiringC

onversational C

ompetence.

Elinor

Ochs

and B

ambi

B.

Schieffelin, eds. Pp. 66-113. Boston: R

ougledge & K

egan Paul.

Ochs, E

linor, Bam

bi B. Schieffelin and M

artha L. P

latt1979 P

ropositions across Utterances and Speakers. In D

evelopmental

Pragm

atics. E

linor O

chs, and

Bam

bi B

. Schieffelin,

eds. P

p.251-268. N

ew Y

ork: Academ

ic Press.

Pike, K

enneth L1945

The Intonation of Am

erican English. A

nn Arbor. U

niversity ofM

ichigan Press.

Pom

erantz, Anita

1978 Com

pliment R

esponses: Notes on the C

o-operation of Multiple

Constraints.

In Studies

in the

Organization

of C

onversationalInteraction. Jim

Schenkein, ed. Pp. 79-112. New

York: A

cademic

Press.

Pom

erantz, Anita M

.1980

Telling

My

Side: 'L

imited

Access'

as a

`Fishing' D

evice.Sociological Inquiry 50:186-98.

Pom

erantz, Anita

1984a Agreeing and D

isagreeing with A

ssessments: Som

eFeatures of P

referred/Dispreferred T

urn Shapes. InStructures of Social A

ction. J. Maxw

ell Atkinson and John

Heritage, eds. P

p. 57-101. Cam

bridge: Cam

bridge

Pom

erantz, Anita

1984b Pursuing a Response. In Structures of Social A

ction. J.M

axwell

Atkinson

and John

Heritage,

eds. Pp.

152-164.C

ambridge: C

ambridge U

niversity Press.

54

Ross, J.1967 C

onstraints on Variables

in Syntax.

Ph.D.

Dissertation,

MIT

,C

ambridge, M

ass..

Sacks, Harvey, E

manuel A

. Schegloff and Gail Jefferson

1974 A Sim

plest Systematics for the O

rganization of TurnT

aking forC

onversation. Language 50:696-735.

Schegloff, Em

anuel A.

1968 Sequencing in Conversational O

penings. Am

erican Anthropologist

70:1075-1095.

Schegloff, Em

anuel A.

1980 Prelim

inaries to

Preliminaries:

'Can

I A

sk Y

ou a

Question'.

Sociological Inquiry 50:104-152.

Schegloff, Em

anuel A.

in press Recycled T

urn Beginnings: A

Precise R

epair Mechanism

inC

onversation's Turn-taking O

ragnisation. In Talk acrd SocialO

rganisation. G

raham

Button

and John

R.E

. L

ee, eds.

Clevedon, E

ngland: Multilingual M

atters, Ltd.

Schenkein, Jim1978

Studies in the Organization of C

onversational Organization. N

ewY

ork: Academ

ic Press.

Silverstein, Michael

1976 Shifters,

Linguistic

Categories,

and C

ultural D

escription. In

Meaning in A

nthropology. Keith H

. Basso and H

enry A. Selby,

eds. Pp. 11-56. Albuquerque: U

niversity of New

Mexico Press.

Zim

merm

an, Don H

. and Candace W

est (Eds.).

1980Sociological Inquiry: Special D

ouble Issue on Language andSocial Interaction.