Modes of Meaning in a Science Activity

Modes of Meaning in aScience Activity

Gordon WellsUniversity of Toronto, Toronto, Canada

In this article I explore the interrelated contributions of different modes of meaning

making to the interaction between an adult and two young students around a science

experiment involving refraction, henceforth referred to as the `̀ shoebox'' experiment.

The event, which I shall describe in more detail below, was videorecorded in 1992 in a

classroom of grade three and four students in an elementary school serving a

multiethnic population in downtown Toronto. I was both the camera operator and the

visiting adult.Some years later, I selected the videotape of the shoebox experiment as the basis

for practical work in my graduate class on discourse analysis. There were two reasons

for choosing this material. The first was because of its completeness as the record of

the various stages of an activity, from initial selection by the two children to the

writing of the texts in which they reported what they had done and learned; the second

was that, as I had been the `̀ visiting teacher,'' I would be able to offer a participant's

perspective and also to encourage the sort of critical analysis of the interaction that

would have been ethically less acceptable if another teacher had been involved. As it

turned out, both these expectations were fulfilled, though the second not exactly as I

had anticipatedÐas I shall explain below.

`̀ DOES LIGHT ALWAYS SHINE STRAIGHTFORWARD?''

The episodes to be investigated occurred in the course of a unit on light and

color. On this occasion, the class teacher had asked the children to work in

groups of two or three and to find a question that they would be interested in

exploring, using the variety of materials that she had made available. Along with

mirrors, batteries and bulbs, colored acetate and paints, etc., the teacher had also

brought in a number of books, some of which included suggestions for

experiments to carry out. It was from one of these books that Jasmin and Alex

had chosen their topic and during the course of a whole morning they worked on

the suggested experiment. Briefly, this required them to take a cardboard box, for

Direct all correspondence to: Dr. Gordon Wells, Ontario Institute for Studies in Education,

University of Toronto, 252 Bloor St., W. Toronto, Canada M5S 1V6.

Linguistics and Education 10(3): 307±334 All rights of reproduction in any form reserved.

Copyright D 2000 by Elsevier Science Inc. ISSN: 0898±5898

example a shoebox, and cut two narrow vertical slits about 2 cm apart in the

middle of one end. Having placed a jam-jar of water in the middle, they were to

direct a flashlight horizontally through the slits so that the beams would pass

through the jar of water. This, they were informed, would cause the beams to

cross, a phenomenon that they should be able to observe by looking at the

bottom of the box between the jar and the end of the box opposite the flashlight.

These instructions were accompanied by a schematic representation of what they

should see.

In the first part of the morning, Jasmin and Alex had assembled the materials

and, in order to `̀ see the light criss-cross,'' they had taken the experimental set-up

under the teacher's desk (having first draped coats at back and front to keep out

the natural light). Then, after having satisfied themselves that they were able to

reproduce the criss-cross effect, they were anxious to share their success with

others. This was the stage at which, after morning recess, I joined them.

The foregoing account of the children's activity is based on a number of

sequences recorded on the videotape. During the first part of the morning, I set

the camera up to record what took place in various locations but, while it

recorded, I spent time with different groups, observing and occasionally joining

in their discussion. Although one of the sequences recorded had captured the

experiment under the desk, I had neither observed it nor seen the tape when I

placed the camera close to where they were based and deliberately went to talk

with them.

What followed can be divided into six episodes. In the first, I asked the two

children to tell me about the question they had been interested in and what they

had found out. This they did in considerable detail, showing and describing what

they had done. In the second episode, which was a continuation of the first, I

began to challenge their description by asking them how they `̀ knew'' that the

beams of light crossed. They claimed that the two vertical bars of light on the far

end of the box were the result of the beams of light crossing after they passed

through the jar. Could they convince me by explaining how that happened or, at

least, by demonstrating to me that they actually crossed? This led to the joint

construction of the `̀ finger test'': by covering one of the slits with a finger on

the inside of the box, the diagonally opposite bar could be made to disappear.

The successful completion of this test then led to a further question: Was it the

water in the glass jar or the glass jar alone that produced this result? Since

Jasmin thought it was the water and Alex the jar, a further experiment was

required. An empty jar was substituted for the full jar and the finger test

repeated. This clearly `̀ showed'' that it was the water and not the jar that caused

the beams of light to bend.

In the third episode, the two children enthusiastically explained what they had

been doing to the Vice Principal, who happened to have come into the classroom.

They also shared their work with their class teacher, though much more briefly. In

308 WELLS

the fourth episode, which occurred shortly after the preceding one, I rejoined the

children as they were preparing to write their report, having first placed the

camera to record our interaction. During this episode, we discussed who the

report was to be written forÐ`̀ people who don't know,'' Jasmin suggestedÐand

what information these readers would want to know. This discussion led to the

joint construction of a genre structure appropriate for their writing and a rehearsal

of potential `̀ content.'' Just before this was completed, however, we were

interrupted by the class teacher's request to the class to `̀ clear up'' and by an

invitation to Jasmin and Alex to tell the assembled class about what they had

done and learned. Their `̀ show and tell'' constitutes episode five. The final

episode took place the next morning, as the two children worked together on their

written reports.

A PERSONAL INVESTIGATION

I have to confess that the theme of this articleÐthe inter-relatedness of different

modes of meaning makingÐwas not the reason for starting this investigation. In

class, we were initially concerned with different approaches to the analysis of

discourse and with the different possible units of analysis that might be

employedÐmoves and exchanges (Berry, 1981; Sinclair & Coulthard, 1975),

turns and adjacency pairs (Sacks, Schegloff, & Jefferson, 1974), cohesion chains

(Halliday & Hasan, 1976), thematic units (Lemke, 1990); we were also making

connections to Vygotsky's (Vygotsky, 1981) ideas about semiotic mediation and

Bakhtin's (Bakhtin, 1986) work on speech genres. However, as we viewed,

reviewed and discussed the tape analytically, it was the nature of my own

participation that became problematic. In particular, I was disturbed by the

differential manner in which I had interacted with the two childrenÐan issue that

was raised, very tactfully, by several members of the class. What was it, I

wondered, that had led me to give so much more of my attention to Jasmin than

to Alex?

At first I thought it was just that Jasmin was more interested than Alex in my

attempt to get them to see their efforts to replicate the phenomenon of `̀ bending

light'' as an opportunity to engage in the more scientific activity of formulating

and testing an explanatory hypothesis. Certainly, on a purely verbal level, Jasmin

was much more willing than Alex to engage in my reconstruing of the activity in

more synoptic terms (Halliday, 1993), and she also seemed to be the main

beneficiary of my efforts. In fact, from the transcript made of the recording, Alex

hardly seemed to participate in the discussion at all. However, as became

apparent when we paid close attention to the videotape, Alex was just as

interested as Jasmin in the phenomenon itself. It was he who manipulated the

flashlight in such a way that the crossing beams of light could be clearly

observed to `̀ criss-cross'' on the bottom of the box. Was it, then, that his

MODES OF MEANING IN A SCIENCE ACTIVITY 309

command of English was more limited and that, for this reason, he left the

speaking to Jasmin?1

Comparison of the two written reports that they subsequently produced

certainly suggested that Jasmin was more fluent than Alex and that, to a much

greater extent than he, she was writing to communicate rather than simply to

complete a required task. However, her assumption that she should do most of the

speaking seemed to have a different basis, for it was not only when interacting

with adults that Jasmin monopolized the interaction. As I discovered later, when I

viewed the episodes from earlier in the morning in which they looked through the

book to select an experiment, and again when they carried it out in the dark under

the teacher's desk, it was clearly Jasmin who was in charge. She made all the

decisions, told Alex what to do, and monitored his actionsÐto the extent of

taking over from him when she was not satisfied with his performance. However,

when I first interacted with the two children I was not aware of this power

imbalance; if I had been, I might have made a greater effort to ensure that I

gave Alex's contributions as much attention as Jasmin's.

On still further viewing of the tape, a different and perhaps more plausibleÐ

though less pedagogically acceptableÐexplanation for my concentration on

Jasmin occurred to me, namely that she was socially more adept than Alex.

Enjoying adult attention, she was more skillful in securing and responding to an

unknown adult's interest and, without being aware of it, I responded by giving

most of my attention to her. There certainly seemed to be grounds for accepting

this interpretation. For example, where Alex was apparently absorbed in his own

goal of repeating the phenomenon of the `̀ criss-cross'' beams of light, Jasmin

was willing to play up to my agenda and engage in the attempt to convince the

skeptical adult that their claim to have seen the beams criss-cross could be

justified by empirically supported explanation.

It was thus in order to check out these hypotheses that I first began to look

more closely at, and listen to, the extra-linguistic dimensions of the interaction.

Up until that point, although having viewed the videotape many times, I had been

satisfied, for the purposes of analysis, to work from the written transcript that had

been prepared from the audio track of the recording. When I began to pay closer

attention to all the information available in the videotape, however, I came to see

how inadequate the transcript was as a record of what had been going on.

ATTENDING TO NON-VERBAL MEANING IN DISCOURSE

Face-to-face interaction involves many dimensions of meaning, only some of

which are represented in a written transcription of `̀ what was said.'' In

particular, a written transcript fails in general to capture meanings that are

conveyed by such non-verbal means as intonation, facial expression, gesture,

and participants' spatial orientation to each other and to the material artifacts

310 WELLS

involved in their activity.2 However, this is hardly surprising, since for those

who grow up in highly literate cultures, meaning is often equated almost

exclusively with linguistic meaning, that is to say, with the semantic distinctions

that are realized lexico-grammatically through words and the syntactic relation-

ships between them.

To understand the probable reasons for this `̀ logocentrism'' it is necessary to

consider the different ways in which meaning is realized in observable behavior

in the different semiotic modalities and how, in the historical development of

writing, only some modes of meaning making came to be represented directly in

the visuographic display.

A central feature of language is its tendency towards categoriality. For

example, despite the continuous variation of the sound that is produced in

speaking, the stream of speech is treated as consisting of a succession of tokens of

discrete phonemic categories, and particular sequential patterns of phonemes are

construed as particular morphemes or lexemes. So although, in fluent speech,

sounds may be produced that are intermediate between, for example, `̀ p'' or `̀ b''

or `̀ r'' and `̀ w,'' they are nevertheless heard, in context, as instances of one or the

other phoneme, and the larger segments of which they are a part are heard as

discrete lexemesÐfor example, either `̀ pie'' or `̀ by.'' There are no intermediate

possibilities.3 And the same is true of the meanings that words and syntactic

structures encode; for the most part, they are treated as equally categorical. (But

see Lemke (in press) for a discussion of the realization of graded meanings.)

This is not the case, by contrast, for intonation or gesture, where parameters

such as amplitude, rate or directionality can vary continuously, with correspon-

ding continuous gradation of the meanings expressed. For example, within the

tone group (i.e., a segment of speech containing a tonic syllable on which there is

a terminal pitch movement), the greater the amplitude on the tonic or any other

stressed syllable, the more the meaning of the word or phrase in which that

syllable occurs will be perceived as being emphasized and therefore in contrast

with alternatives that potentially might have occurred in that position (including

the unstressed, and therefore unmarked, occurrence of the same item). Similarly,

with gesture, the orientation of the palms towards or away from the speaker, for

example, can signify varying degrees of commitment, on the part of the speaker,

to the point of view being expressed.

The strong tendency for linguistic forms and meanings to be treated as

categorical may also explain why the (largely non-categorical) non-verbal

dimensions of meaning are largely ignored in writing. From recent work on

the development of writing it seems that, in the earliest stages of the evolution

of most writing systems, the primary aim was not to represent speech directly,

but to find a way of giving permanent visual representation to the meanings

expressed through speech (Harris, 1986). That is to say, the visual symbols that

were used were iconic, and directly represented the categories of objects and


events referred to in speech rather than indirectly through the representation of

the spoken utterances through which they were encoded. So, by the time that

writing systems came to represent the formal organization of speech more

directly, through the incorporation of syntax (e.g., `̀ two horses'' instead of

`̀ horse, horse'') and, in some cases, through the use of a syllabary or alphabet

to represent the phonology of speech, the basic organizing principles had

already been established. Writing represented, first and foremost, the semantic

categories encoded in lexical items and in the syntactic relationships between

them (Harris, 1986). Distinctions of meaning of a pragmatic kind, that are often

signaled non-verbally, such as how the sentence should be takenÐas assertion

or conjecture, description or declarationÐwere left to the reader to infer from

the context. Similarly, attitudinal meanings, such as degree of certainty about

or agreement with the propositionÐto the extent that they were encoded

non-lexically and also non-categoricallyÐwere also left to the reader's imagi-

native reconstruction as he or she read the text aloud.

However, this almost exclusive concern with the lexico-grammatical dimen-

sion is not confined to the representation of meaning in the written medium; it

also extends to the way in which we think about meaning in speech. For,

whatever the earliest relationship between visual representation and speech, as

written scripts came to be linguistically based and thus to represent the sequential

organization of spoken utterance, they also began to function as models of

speech. As a result, as Olson (1994) argues, folk theories of the way in which

meaning is represented in speech are dependent on which features happen to be

captured in the orthographic system used to represent it in writing.

This has a further consequence. Since the `̀ non-verbal'' dimensions of speech

are not captured in standard written texts, they are also more difficult to bring to

the level of conscious awareness. The result is that modes of meaning that are not

directly represented in writing are not readily available for subsequent inspection

and reflectionÐat least not with any specificity.

Evidence for this bias for attention to be given to meaning that is realized

lexico-grammatically can be found in the nature of the glosses and formulations

that participants use to characterize previous contributions when a breakdown in

communication occurs (Jefferson, 1972; Wardhaugh, 1985). Choice of particular

lexical items or of grammatical features, such as tense, may be questioned or

disputed, but rarely is mention made of intonation, gaze or gesture, even though

one of these dimensions may be the cause of the difficulty. The reason, it seems,

is that we simply do not have `̀ meta'' terms, in everyday language, with which to

refer to the distinctions that are perceptible on these dimensions of interactive

behavior or to the differences in meaning that they communicate.

However, that is not to say that, on these non-verbal dimensions, meaning is

always vague and indeterminate. On the contrary, some distinctions are almost

as conventional and categorical in their meaning as lexical items. For example,

312 WELLS

the contrast between rising and falling pitch movement in English is con-

ventionally used to signal the distinction between information that is treated as

shared and that which is offered as new (Brazil, 1975) and this contrast, taken

together with the syntactic arrangement of the subject and first element of the

verb in the main clause, is systematically used to distinguish between the

major classes of speech act function. Other intonational features, such as the

range of pitch movement, variation in pace, and marked stress, are also

regularly used in a fairly conventional manner to communicate the speaker's

orientation both to the content of the message and to the intended recipient

(Chafe, 1985), and for those aspects of meaning that are signaled by these

means the speaker may also be held accountable. Nevertheless, as already

suggested, it is usually more difficult for a listener, if challenged, to explain

the grounds for his or her interpretation.

Similarly, direction of gaze and facial expression can also be conventionally

interpreted, at least to some degree. Looking away when asked a question, for

example, is usually interpreted as unwillingness or inability to give an answer in

the terms proposed by the questioner, and smiling or frowning are normally

interpreted as signaling, respectively, a positive or negative orientation. Some

gestures also have a conventional interpretation, ranging from nods and shakes of

the head to thumbs up and a whole range of scatological gestures employed by

irate drivers.

Nevertheless, with the exception of the latter, it still remains the case that,

when compared with meanings that are encoded lexico-grammatically, distinc-

tions of meaning that are expressed non-verbally are much more difficult to pin

down and identify. And so, lacking means for representing or referring to them,

participants rarely appeal to them in attempting to resolve disagreements about

what was `̀ said'' and what was `̀ meant.'' For the same reason, they are rarely

recognized as making a significant contribution to the meanings that are

co-constructed in face-to-face interaction.4

However, it cannot be assumed that, because these dimensions of meaning

normally operate outside conscious awareness, they do not influence participants'

interpretations of each other's meanings (Schiffrin, 1994). In fact it was because I

suspected that they might have played an important part in the shoebox episodes

that I decided to investigate them more systematically.

ANALYZING NON-VERBAL MEANINGS

As already pointed out, one of the key differences between verbal and non-verbal

dimensions of meaning is that the latter tend towards continuous variation. There

is, thus, an inherent contradiction in proposing to code non-verbal meanings

by assigning tokens to category types. First, there is the problem of deciding,

conceptually, where to draw boundaries between categories and since, unlike


lexico-grammatical categories, such categories do not wear their `̀ names'' on

their sleeves, to find appropriate ways of describing those that are distin-

guished. Second, there are inevitably analytic problems in deciding where a

perceptible change on one or more dimensions is such as to constitute `̀ a

difference that makes a difference'' (Bateson, 1972) as far as the participants

are concerned and so to warrant treating the adjacent stretches of behavior as

tokens of different types. Nevertheless, until at least tentative decisions are

made with respect to these problems, it is extremely difficult to proceed

towards a systematic investigation.

A further problem that has to be solved is that of accurately representing the

video-recorded information that is treated as significant in such a way that it can

be typed into a computer and subsequently displayed in a form that is easily

interpreted by a reader who has access only to the printed text. Since, in practice,

the solution to this latter problem interacts with the former, I shall start by

describing the procedure that, after a considerable amount of experimentation, we

are currently using.5 It goes without saying that we continue to try to improve it.

Transcription

In the past, it was generally our practice to produce the first draft of the transcript

from an audio copy of the sound-track taken from the recorded video-tape. There

are two advantages to this procedure. First, in copying the sound onto an audio

tape, the signal can be passed through a graphic equalizer and adjusted to

maximize interpretability and, second, an audio machine with foot-pedal control

that incorporates a variable rewind function makes the process of transcribing

much easier than when using a standard VCR. Then, following the production

of the first draft of the transcript, the video-tape would be reviewed in order to

check the identification of speakers and to add information about the accom-

panying non-verbal activity. Up until this point, however, we had been sparing

in the amount of information included, simply placing brief glosses or

descriptions (in parentheses) at appropriate points in the verbal text. In other

words, the transcripts we have worked with up until now have been largely

restricted to lexico-grammatically realized meanings.

In the present case, a transcript of this kind already existed, so selected

episodes from this were initially used as the basis for the current work. These

were first reformatted so that each `̀ utterance''6 appeared as a separate text

paragraph and then the whole episode was entered into a table, using the `̀ text to

table'' command in the WordPerfect (Mac 3.0) word processing program. When

the necessary tidying up had been carried out, additional columns were inserted

to the right of the text in which to add codes for the non-verbal dimensions. Next,

tone unit boundaries within utterances were marked (j), where appropriate, and an

indication of pitch movement added immediately preceding the word containing

314 WELLS

the tonic syllable in each tone unit. Five options were recognized: / rise; \ fall;

\ / fall±rise; / \ rise±fall; andÐlevel.7

Finally, on the basis of repeated viewing of each utterance, a plain

language description of the significant non-verbal behavior accompanying

each utterance was added in a last, wider column prior to coding gesture,

gaze and spatial orientation.8

CODING

Although in interaction the many dimensions of meaning making are seamlessly

woven together in a continuous flow, analysis is greatly facilitated if they are

separated and coded independently. Independent coding also makes it initially

easier to explore the relationships between the different dimensions in a

systematic manner. As a first step in this ambitious program, we decided to

focus on a limited number of dimensions, selecting those that we expected to be

relatively easy to code, at least at a gross level of delicacy, and those that we

hypothesized might be particularly significant for the attempt to explain the

unequal manner in which the teacher interacted with the two children. Our

approach to the coding of intonation, or at least one aspect of it (the management

of the `̀ given-new'' contract (Brazil, 1975) was briefly presented above. The

three remaining dimensions are described below.

Spatial Orientation

Participants' orientation to each other and to any artifacts that play a

significant role in their joint activity can be expected to affect the nature of their

participation (Lemke, 1997). For example, when students sit in rows facing the

teacher at the front of the class, the possibilities of peer interaction are

substantially diminished compared with a situation in which a small group is

arranged in a circle with the teacher absent. However, even in the latter case, if

one participant has greater access than the others to a critical artifact such as a

book or computer keyboard, that participant is likely to play a more central role in

determining how the activity proceeds (Wegerif, 1997).

Conventions of varying degrees of delicacy might be devised to represent

information about spatial orientation, although any scheme would quickly

become overly complex if the number of participants were large or if there were

frequent or continual changes of orientation. Fortunately, in the episodes

analyzed here there were only three participants and a single artifact clusterÐ

the experimental apparatusÐas the focus of their shared attention. The solution

adopted, therefore, was to use initials to identify the human participants and X to

indicate the apparatus. A new coding was entered each time there was a major


change in their relative positions. Thus, a typical configuration might be

represented as:

T

J

XA

which, viewed from the position of the camera `̀ below,'' is to be interpreted as

Jasmin and Alex close to the apparatus on adjacent sides of the desk, and Teacher

facing Alex, but further away (than they) from the apparatus. (Significantly, the

end from which the beams were projected through the two slits in the box was the

one facing T; however, this information could only be included in the prose

description, as no concise way of coding it could be devised.)

Gaze

In informal conversation between peers, participants frequently look at each

other and engage in periods of mutual gaze. The speaker typically looks at one or

more of the addressees at the beginning of a turn and again at the end; during

short turns, gaze is often maintained throughout, while in long turns the speaker

may look away for part(s) of the turn. Listeners behave in a similar manner,

although they may spend a larger proportion of the time looking away. Thus, the

moments at which mutual gaze most frequently occurs are found around points of

potential turn change. This pattern is modified when an object, person or event is

the current focus of joint attention. In these circumstances, all participants' gaze

tends to be directed to this `̀ third party,'' although mutual gaze still frequently

occurs at potential turn changes.

Mutual gaze seems to perform a number of functions. First, as its most

frequent temporal location suggests, it plays an important role (together with

other cues available in the auditory channel) in managing the smooth transfer of

the floor from one speaker to another, including the selection of next speaker.

This aspect of gaze has been described in considerable detail in the work of

conversation analysts (Sacks et al., 1974). A second function is that of

monitoring the gestural and postural behavior that accompanies oral commu-

nication. Speakers monitor listeners' uptake of their messages, while listeners

monitor the additional `̀ information'' that speakers provide, particularly through

gesture and facial expression (Schiffrin, 1994). Without mutual gaze, this

important means of negotiating intersubjectivity would be unavailable, as is

the case in telephone conversations. A third function is related to the relative

status of the participants; here, avoidance of gaze may be just as important as

looking one's interlocutor `̀ in the eyes'' in signaling in a culturally appropriate

manner the recognition of one's inferior status.

Determining co-participants' direction of gaze is a skill acquired in very early

infancy (Bower, 1974) and, from early childhood, humans are able to make a

316 WELLS

similar determination for people with whom they are not actually interacting,

provided they can see the full face. This skill even extends to the observation of

naturalistic representations of people in paintings and in the images captured by

still or video cameras. It is therefore relatively easy to determine the gaze of

participants from a videorecording of their interaction, provided one can see their

faces. The difficulty, however, is in deciding how much of the resulting

information to code, and how to do so.

In the present case, we decided to focus on speaker's gaze, and simply to code

whether or not s/he looked at the addressee(s) for at least some part of his or her

speaking turn and, if not, whether his or her gaze was directed to the current focus

of joint attention or elsewhere. We also coded whether mutual gaze between

speaker and addressee(s) occurred during a speaker's turn. This information was

recorded using initials, as with spatial orientation, with > to indicate unidirec-

tional gaze, and <> for mutual gaze. Obviously, a much more fine-grained

analysis could be carried out, including a coding of the periods during which each

participant was looking at one or other of the co-participants.

Gesture

Surprisingly, rather little attention has been given to gesture, and almost none

to its contribution to interaction in educational settings. One of the most

comprehensive general accounts is provided by McNeill (1992), based largely

on videorecorded observations of subjects retelling a Sylvester and Tweety Bird

cartoon. In introducing this work, McNeill makes a clear distinction between

gesture (classified as `̀ gesticulation'' by Kendon (1988), on whom McNeill

draws) and other forms of manual communication (Singleton et al., 1995). Unlike

`̀ emblems'' (and even more, the signs in a sign language such as SLA), which

have a conventional meaning and a standard form, gestures are `̀ idiosyncratic

spontaneous movements of the hands and arms accompanying speech'' (McNeill,

1992, p. 37). He then distinguishes four different kinds of gestures: iconics,

metaphorics, deictics and beats.

We found the continuum from gesticulation, through emblems, to full sign

languages a helpful starting point. Like McNeill, we were mainly interested in the

gesticulation end of this continuum. However, on viewing our data in the light of

the criteria discussed, we decided that the continuum needed to be extended

further to the `̀ left'' to include bodily movements that, while not intended to

communicate, are normally interpreted as meaningful by other members of the

culture and oriented to, where necessary, in order to coordinate joint activity.

What we are referring to here are the movements that make up a recognizable

action or succession of actions, such as making a bed or sawing and nailing

pieces of wood to construct a box. In our data, the second episode contained a

considerable amount of such interpretable action, as the participants together


invented and applied the test for determining whether the beams of light bent as

they passed through the jar of water.

However, what finally persuaded us that such action should be included at one

end of the continuum of semiotic behavior was a category intermediate between

action and gesture as defined above. We noticed that when the two children had

completed their experiment, they were anxious to demonstrate it to others. Now,

however, although they repeated many of the same actions, they did so not to

achieve the experimental result but rather to demonstrate to others how the action

was performed.9 On these grounds, we are proposing an extension of `̀ Kendon's

continuum'' as follows:

actionÐdemonstrative actionÐgestureÐemblemÐsign language

Equally fruitful as an initial framework was McNeill's categorization of the

different kinds of gestures. However, here again, we found that our data

suggested the need for some further distinctions. Currently, our scheme distin-

guishes the following categories and sub-categories:

Category CodeAction ADemonstrative Action Da

Deictic:

(a) Point to a co-participant Pv

(b) Point to concrete object or location Po

(c) Trace the path of an object or phenomenon Pt

(d) Abstract Point to a `̀ concept'' referred to in either concur-

rent or preceding speech

Pe

Iconic: Gesture that iconically represents an object Go

or action referred to in the concurrent utterance Ga/t

Metaphoric:

(a) Gesture that iconically represents a `̀ concept'' in the

concurrent utterance (e.g., closing the fist to represent

`̀ contraction,'' or a horizontal movement of one hand away

from the body, palm down, to represent ``it was

straightforward.'')

Gc

(b) Gesture that represents an abstract relationship, such as

`̀ equivalent to'' or `̀ contrasted with''

Gr

(c) Gesture representing a mental state (e.g., touching one's

temple to represent `̀ thinking'')

Gm

(d) Gesture representing communicative action, such as

`̀ telling'' or `̀ writing''

Gc

Attitudinal:

(a) Gesture that represents an affective state, such as pleasure Gf

318 WELLS

In addition, gestures were coded as either underlining (u) the meaning of the

accompanying speech or substituting (s) for speech.

As with the development of most coding schemes, the categories above were

arrived at after a lengthy period of to-and-fro between different versions and their

application to increasingly large segments of the data. Finally, all three episodes

were coded using the above categories and an * placed before the word in the

utterance on which the `̀ stroke,'' or point at which the gesture comes to fruition,

occurred.10 For the most part, little difficulty was experienced in assigning

gesture tokens to types.11

Table 1 presents a short coded extract from episode one in order to illustrate

the way in which the coding was carried out. (In transcribing the speech, the

following conventions are used: an inaudible segment of word length is shown as

x; words spoken with particular emphasis are in CAPS; a period indicates a

perceptible pause, with an additional period for each second; segments spoken

simultaneously are underlined).

For this investigation, only three of the six episodes identified earlier were

analyzed for the dimensions just described. These are the ones in which I interacted

with the two children in the role of `̀ visiting teacher'' (episodes 1, 2, and 4).

RESULTS OF THE ANALYSIS

At this early stage in our investigation of the contribution of non-verbal

dimensions to the meanings negotiated and co-constructed in face-to-face

interaction, it would be inappropriate to attempt to draw conclusions of any

generality. Instead, the results of the analyses will be used to illuminate the

particular task-related interaction from which they are derived. In particular, I

shall use them to compare the interactional behavior of the two children and of

the teacher as he interacted with each of them. For this purpose, I shall focus

chiefly on the results of the analysis of gesture. First, however, I will comment

briefly on the other dimensions examined.

SPATIAL ORIENTATION

In all three episodes, the three participants tended to situate themselves on three

sides of the desk with the shoebox in the center. In the first two, while the

(b) Gesture that represent an epistemic attitude with respect to

the concurrent utterance, such as uncertainty

Gz

Emphatic: Referred to as `̀ beat'' by McNeill (1992), this

typically up±down gesture marks emphasis or `̀ indexes the

E

word or phrase it accompanies as being significant, not for its

own semantic content, but for its discourse±pragmatic

content'' ( p. 15)


Table 1. Sample of Coded Transcript

44:25 T Now . can you \tell me . . . T < > J [A is off picture] T sits

on chair

what question have you

\answered?

TJA < X J A comes to stand opposite T

All look at shoeboxT X A

or what question . have

you found \out about by

doing this?

A continues to look down

at box

J Well . we x- T < J

Well we had- . well first

our question \was .

J < X J holds box, lifting it up

and down from table, nodding

head from side to side

can light- does light

always shine */straight .

J <> T Gcu turns to face T on `̀ can'' and

makes strong rh horizontal

gest. from r to lacross chest

on `̀ straight''

/straight*forward? J <> T Gcu repeats gest. into lh

T And /does it? T <> J

J / \No . J <> T Gzu J wobbles head = qualified

cos it sort of *\bends Pe points with rh to inside of box

cos before when we put

this *RIGHT\ / there .

J < X Dau picks up jar on `̀ cos'' and

places it in box on `̀ right''

T Yeh

J and this part right through

there it was very

said as an addition

*\ /dark . Pe looks to right, rh point

but we had trouble getting

it /dark

J <> T laughs after `̀ dark''

and it was just a/ \ torch

and we put it *right /there . Dau J demonstrates where they

held the f-light, as if holding

it, close to the slits

A xxx Dau JA A moves close to J's side and

also shows where they held

the light

G X

J but it couldn't be *right /

close to it

J < T Dau (marked rising inton.) J

demonstrates light close

to box

T uhhuh

J because it would be *too

bright and straight

Gzu J open hand gesture, both

hands, indicating excess

A It was just too straight

xxxx

A < T Gtu A rh horizontal gesture along

path of light through box

(repeated)

320 WELLS

experiment was being described and then extended, the shoebox was oriented

with the slits on the left (from the camera's perspective). Initially, Jasmin was

positioned on one side of the box, facing the camera, with the teacher at the end

with the slits and Alex facing him at the other end. However, as they demon-

strated what they had done, Alex moved to stand by the side of Jasmin in order to

share in the demonstration.

Episode two started with the same arrangement as did episode one but, early

on, the teacher changed places with Jasmin, thus putting her in the privileged

position of controlling the flashlight and directing its beam through the two slits.

On one occasion, Alex leant across the box in an effort to direct the light, but

Jasmin refused to cede control. Alex did, however, take the initiative in finding an

exercise book to cover his end of the box so that the beams emerging from the jar

could be more easily seen. Then, more or less excluded from the action, Alex

wandered away for a few seconds but, on returning, he again leant over the box to

test the effect of covering one of the slits with his finger.

In episode three, as they began to plan the report, Alex and the teacher were

sitting on two adjacent sides of the table, Alex facing the camera. After a short

delay, Jasmin returned to the table and took up a position standing opposite the

teacher; initially, the shoebox was oriented as previously. Part way through this

episode, two boys came to show the periscopes they had made and both Alex and

Jasmin engaged in conversation with them. Jasmin turned the box so that the slits

were facing her and she demonstrated what they had been doing to one of the

boys. A little later, Alex stirred the water in the jar with his finger and briefly

continued the conversation with one of the boys, who looked at Alex through his

periscope. Then, on Jasmin's insistence, he turned his attention back to the

planning of the report.

In reviewing the results of this analysis, three things are apparent. First,

whether by design or by chance, Jasmin always takes up a more dominant

position than Alex; throughout, it is she who controls the flashlight and, in the

fourth episode, she stands facing the teacher while Alex sits to one side of him.

Second, although Alex makes several attempts to play a more central role in the

action, Jasmin does not allow him to do so. Third, despite this, Alex remains

involved and, apart from a few brief interludes, appears to give his full attention

to what is going on.

GAZE

There is a strong contrast between the gaze behavior of the two children across

all three episodes. Except when all three participants are looking at the

apparatus, Jasmin looks at the teacher when she speaks, often with a smile or

some other facial expression showing interest. On almost all occasions, he also

looks at her while she speaks and there is a period of mutual gaze. By contrast,


Alex rarely looks up when he speaks and, when he does, it is more often at

Jasmin than at the teacher. He rarely makes eye contact with the teacher. The

difference between them in this respect can be captured in the finding that

whereas Jasmin looks at her addressee on more than 90% of her speaking turns,

the proportion for Alex is only about 25%. When to this is added that Jasmin

speaks twice as often as Alex, and typically at much greater length, it is apparent

that there are many occasions of mutual gaze between the teacher and Jasmin

and hardly any with Alex.

This finding goes a considerable way towards explaining the differential

distribution of the teacher's attention: Jasmin, to a much greater degree that Alex,

kept up her end in the interaction. Through frequent exchanges of mutual gaze,

she signaled that she was on the same wavelength as the teacher and so elicited

further interaction. By contrast, Alex rarely gave such signals and so was

interpreted as not wishing to engage in the interaction. Whether this was an

appropriate interpretation will be further considered below.

GESTURE

As previously described, all gestures were assigned to one of the five categories

and, where appropriate, to one of the subcategories within them. In addition, where

a non-verbal action that accompanied or took the place of a speaking turn could be

considered meaningful within the framework provided by the activity, it was coded

as either `̀ action'' or `̀ demonstrative action'' according to the criteria described

above. Then, as frequencies in some subcategories were very low, a regrouping

was carried out to create the following seven clusters: Action; Pointing to material

objects and locations; Demonstrative Action; Iconic gestures; Metaphorical

gestures (including metaphorical points); Attitudinal gestures; and Emphatic

gestures. Table 2 shows the distribution of types of gesture by participant.

Table 2. Distribution of Gestures by Category and Participant

Category Jasmin Alex Teacher Total

Action 16 14 2 32 (19.8%)

Point (to object or

aspect of situation)

11 6 23 40 (24.7%)

Demonstration 10 2 0 12 (7.4%)

Gesture: Iconic 5 9 6 20 (12.3%)

Gesture: Metaphoric

(incl. abstract point)

6 3 5 14 (8.6%)

Gesture: Attitudinal 12 1 2 15 (9.3%)

Emphasis 8 2 19 29 (17.9%)

Total 68 (42.0%) 37 (22.8%) 57 (35.2%) 162 (100%)

322 WELLS

As can be seen, overall, Jasmin produced the greatest proportion of gestures,

and Alex the least. However, what is more interesting is the different rankings of

gesture types within participants. Although for both Jasmin and Alex, action was

the most frequent form of non-verbal communicative behavior (most tokens of

which occurred in episode 2) the two children differed in their next most highly

ranking categories. For Alex, it was iconic gestures, and, in particular, gestures

tracing the path of the beams of light; for Jasmin, on the other hand, it was

attitudinal gestures (mainly her affective attitude to the things she was talking

about) that ranked second, closely followed by pointing and demonstrative

action. Meanwhile the teacher made most use of pointing gestures and gestures

that emphasized the ideas he wanted the children to grasp. It would not be

unreasonable, I believe, to take these relative distributions of gesture types as

indicative of the overall stance of the three participants to the activity in which

they were jointly engaged.

Equally illuminating is to look at the distribution of gesture types across the

three episodes.12 From Table 3, it can be seen that the explanation-oriented

activity in episode two evoked the largest overall number of gestures and that, of

these, more than 75% took the form of goal-oriented action and deictic points

intended to make explicit the specific material phenomena between which a

causal relationship was being proposed, e.g., `̀ T: and the light that goes in *\ /

here (pointing to left slit) . is showing up on */ \ that side (pointing to bar of light

on right).'' Not surprisingly, most of the action was performed by the two

children, whereas most of the pointing gestures were made by the teacher in his

attempt to focus the children's attention on the phenomena that he considered

particularly relevant for the construction of an explanation.

In the much shorter first episode, by contrast, there is a more even distribution.

Here, a variety of gesture types support the children's attempts to describe and

demonstrate to the teacher the actual procedures they had used in producing the

Table 3. Distribution of Gestures by Category and Episode

Category Episode 1 Episode 2 Episode 3

Action 1 (3%) 22 (34%) 9 (14%)

Point (to object or

aspect of situation)

6 (18%) 28 (42%) 6 (10%)

Demonstration 8 (24%) 1 (1%) 3 (5%)

Gesture: Iconic 8 (24%) 6 (9%) 6 (10%)

Gesture: Metaphoric

(incl. abstract point)

4 (12%) 1 (1%) 9 (14%)

Gesture: Attitudinal 6 (18%) 4 (6%) 5 (8%)

Emphasis 0 4 (6%) 25 (40%)

N = 33 (100%) 66 (100%) 63 (100%)


`̀ criss-cross'' effect that the book had instructed them to look for. It is in this

episode too that, through her attitudinal gestures, Jasmin expresses her strong

involvement in solving the problems posed by the `̀ experiment'' that they have

chosen to carry out.

Episode three shows yet another distributional pattern, with a high pre-

ponderance of emphatic gesturesÐlargely contributed by the teacherÐas the

three participants consider the content and sequence of the different stages that

should be included in the report that the children are required to write about

the activity in which they have engaged. It is significant that it is in this

episode, too, that the majority of metaphoric gestures occur, as Jasmin

considers the (absent) audience and the sort of information that they would

want to know. Action and demonstrative action, by contrast, occur chiefly

when Alex and Jasmin are showing what they have been doing to the two boys

who pay a visit.

Taken together, the results of the comparison across episodes suggest that

participants produce different types of gesture according to the nature of the task

in which they perceive themselves to be involved. That is to say, as with speech,

participants make choices in terms of the genre appropriate to the task in hand

(Kamberelis, 1995; Miller, 1984). Episodes one and two are clearly distinguish-

able from each other in this respect, with the first being concerned with

demonstration and the second with making an explanation. In these two episodes,

all three participants seem to share a similar orientation to the task and use a

similar range of gestures (including action and demonstrative action as categories

of gesture for this purpose). In episode three, by contrast, the participants appear

to have adopted somewhat different orientations; while the two children were

exploring the alternatives of content and organization for their written reports, the

teacher was providing guidance and support from the perspective of one who had

already considered the options.

In work with fifth and sixth grade science students, Crowder (1996) found

evidence of a somewhat similar matching of gesture type to task undertaken.

However, for the pairs of students she compared, the task was relatively constant:

to explain a phenomenon. What differed was the orientation to the task adopted

by the different students. Where students were describing a model or explanation

with which they were already familiar, their gestures tended to be iconic or

points which added no information that was not already present in the

accompanying utterances (`̀ gesture-redundant talk''). In other words, their

gestures tended to add emphasis for the benefit of the audience. The gestures

of students who constructed explanations in the moment, on the other hand, often

added meaning to what was verbally expressed or even substituted for a word

that was unavailable (`̀ interwoven gesture talk''). These gestures often fore-

shadowed the meaning that was verbally expressed, and the `̀ stroke'' was much

less likely to coincide with the intonationally marked tonic syllable in the

324 WELLS

accompanying utterance than was the case where students described an already

familiar model.

In the data from the shoebox experiment, the children provide some

evidence of both types of orientation. In the first episode, both are describing

the experiment they have just carried out and their gestures serve to reinforce

the meanings expressed in speech, while at the same time giving directional or

locational precision to words or phrases such as `̀ very far,'' `̀ too straight'' or

`̀ like that.'' In the second episode where, in response to the teacher's request,

they are attempting to prove that the beams of light bend, it is their actions

that are the primary mode of meaning making and their speech provides an

interpretive commentary on what they are doing and observing. In episode

three, where the task involves both envisaging the likely informational needs

of their audience and reviewing the procedures that might need to be

described, both children produce metaphorical gestures that seem to help

them to clarify what they will need to report. However, the ratio of gestures to

utterances is lower than in the previous episodes and they rarely substitute for

meanings unspoken.

The analysis reported by Crowder (1996) is considerably more detailed than

the one attempted here and the students to be compared were chosen specifically

because they were judged to demonstrate contrasting orientations to the same

task. In the present case, with a lesser degree of analytic delicacy, it is genre

differences between tasks that are most salient, rather than individual differences

in task orientation. But for this reason, the two studies can be seen to complement

each other in demonstrating genre-related differences in choice of gesture, both

within and between tasks.

However, Crowder's distinction between `̀ gesture-redundant'' and `̀ gesture

interwoven'' talk prompted me to look more closely at possible differences

between Jasmin and Alex in the way their gestures and talk were related. Most

apparent is that Jasmin both speaks and gestures more than Alex (see Table 4).

She also produces many more attitudinal gestures than he does and, leaving aside

metaphorical points, more metaphorical gestures as well. However, what Table 4

reveals is that Alex's ratio of gesture to utterance is slightly higher than Jasmin's,

and that this is most apparent in episode two, where making an explanation is the

Table 4. Percentage Distribution of Utterances by Participant within Episode (Percentage

Distribution of Gestures in Parentheses for Comparison)

Jasmin Alex Teacher

Episode 1 56.6 (63.6) 18.9 (21.2) 24.5 (15.2)

Episode 2 41.7 (40.9) 12.6 (21.2) 45.6 (36.3)

Episode 3 36.4 (31.7) 28.8 (25.4) 34.7 (42.9)

Total 42.3 (42.0) 20.8 (22.8) 36.9 (35.2)


task in hand. On more than one occasion in this episode, Alex silently performs

task-related actions and makes gestures that trace the path of the beams of light

while Jasmin does the talking. The same occurs in episode three. It seems,

therefore, that Alex is constructing an explanation in the moment through action

and gesture, even though, from the record of his speech alone, one might judge

that he was less fully engaged than Jasmin.

DISCUSSION

I started this article with a number of possible explanations for my unequal

treatment of the two children in the interaction I had with them around the

shoebox experiment. And because the original transcript of the verbal

dimension of our interaction did not provide an adequate basis for testing

all my hypotheses, I decided to investigate certain non-verbal dimensions as

well. By examining participants' spatial orientation, gaze and use of

gestures, in addition to their verbal contributions, I gained a much more

complete understanding of the factors that influenced the way in which the

interaction unfolded.

In general, my hypothesized explanations were confirmed. Jasmin was indeed

more socially adept at interacting socially with an unknown adult than AlexÐat

least on the terms that the adult took for granted. Alex's habit of looking down

when speaking instead of at his addressee(s) led me to undervalue his contribu-

tions and to give more of my attention to Jasmin, who behaved in a way that I

found more responsive and engaging. However, having now noticed Alex's

preference for avoiding mutual gazeÐa preference of which I was not con-

sciously aware at the timeÐI am left wondering whether this was a cultural

difference to which I was insensitive or simply a form of shyness. In either case,

his gaze behavior should not have triggered such an inappropriate response on

my part and I shall endeavor to be more sensitive to patterns of behavior that

differ from my culturally-based expectations in my future encounters with

students, whatever their cultural background.

What I was not aware of until I viewed the whole videotape, including the

episodes in which Jasmin and Alex chose, set up, and carried out their

experiment, was the extent to which they had established a relationship of

unequal status, in which Alex allowed Jasmin to make most of the decisions and

to direct their joint activity. When I joined them, therefore, I was unwittingly

drawn into their role relationship and, for the reason given above, I was also too

ready to collude in it (cf. Buzzelli & Johnston, 1997). Close examination of the

two children's spatial positioning in relation to the apparatus and to me provides

ample evidence of Jasmin's assumption of a controlling role, both in the

manipulation of the flashlight in relation to the slits cut in the boxÐto which

she had greater access than AlexÐand in her tendency to speak for both of them,

326 WELLS

frequently speaking over Alex so that his contributions went unnoticed. Again,

had I been more sensitive to Alex's relative disenfranchisement, I would have

made a greater effort to ensure more equal participation.

But just as important as the discovery of my own inappropriate way of

privileging Jasmin at Alex's expense was the enhanced understanding I gained,

from attending to the fine detail of extra-linguistic behavior, of the complex,

multidimensional, and mutually constructed nature of face-to-face interaction in

any situation. From this more dispassionate perspective, I came to see that Alex's

apparently less intellectually engaged involvement in the `̀ scientific'' activity of

explaining the phenomenon of bending light was not simply the result of his

individual attributes or of the teacher's ineptitude, but was jointly constructed by

all three participants through the totality of their communicative actions and their

responses to each other's actions in ways of which they were probably to a large

extent unaware.

However, the real surprise came from what I discovered by closely attending

to the participants' gestures. My own preference for deictic points and

emphatics made me realize that I had more of an `̀ instructional'' agenda than

I had thought. Jasmin's fluent metaphorical and attitudinal gestures confirmed

my impression that she really enjoyed being challenged to move beyond simply

carrying out the instructions found in the book to developing and testing an

experimental proof and then sharing her discovery with others. (This impression

was amply confirmed by her reader-oriented and copiously illustrated report;

see Haneda, (forthcoming).)

But the greatest surprise was to discover how fully Alex remained

involved in the whole process. As our interaction proceeded in real time, I

did not notice how he was shadowing the talk between Jasmin and myself

with gestures that showed he fully grasped the significance of the `̀ finger

test'' and how later, in episode three, he used a tracing gesture to explain

what he could not put into words. Clearly, he was not as uninvolved as his

lack of speech, his willingness to be distracted by the interventions of his

peers, and his apparently aimless stirring of the water in episode three, had

led me to suppose.

THE MUTUAL CONSTRUCTION OF DIFFERENTIAL

PARTICIPATION

All these issues are brought into focus through close examination of a short

sequence in episode three where, having decided upon the intended

audience for their report, the two children begin to consider the form that

the report should take (see Table 5). The teacher is clearly involved here in

scaffolding the construction of a genre appropriate to the chosen audience.

However, his initial prompt, although intended for both children, is more


Table 5. The Mutual Construction of Alex's Peripheral Role

4.10 T so . /what is it you want to

\tell them?

T < J A

T X J

T removes glasses on

`̀ what'' and holds them

across chest in r.h. J

looks at cap of jar, picks

it up and puts it down

again

4.11 J Tell them- er first we've

got to write the \ /question

J <> T J pushes hair behind l.

ear

4.12 T \ /Yep T <> J

4.13 A Then we've got to put- A < X

T < A

A A continues to manip-

ulate f-light; he does

not look up when he

speaks

4.14 J -and then- . J >T < A

4.15 A -we've got to write

about . . .\ /what we did

A < X

T < A

A looks at f-light and

then briefly up at J as

she takes over

4.16 J -write about-

xhdidi4.17 T /Yes T > J T turns head to look at J

at end of A's utterance

4.18 J and um . the

*/materials

we used

J < >T ? J brings both hands,

palms in, against

lower body on

`̀ materials''

4.19 A then . . .

how-whyhit doesiA < X A doesn't look at others

and his speech is ig-

nored. He looks briefly

at T at end of utterance

4.20 T OK so what */ \first d'you

think, the /materials you

used or what you \did?

A < T

T <> J

E T makes vague gesture

with l hand open and

vertical

4.21 A What we \did A < X

4.22 J The materials J < A

4.23 A

J

(laugh) A <> J

T < A

4.24 T *\ /OK, so why-why d'you

think . why d'you think .

\`̀ what we did?'' (to A)

T < A Pv T points towards J;

leans forward, turns

head to look at A

4.25 A I think . \hreading materi-

alsi[ = ?reading the book to

identify their question and

the materials to be used]

A- A J leans over and turns

off f-light that A is

holding

4.26 T \/OK. . . T <A A continues to fiddle

with f-light

4.27 Why d'you think materials

\first? (to J)

T <>J T sits up and looks up at

J

328 WELLS

directed to Jasmin than to Eric; because she is standing opposite him while

Eric is sitting on his left, it is to Jasmin that the teacher looks while

posing the question and there is a brief period of mutual gaze between

them. Not surprisingly, therefore, it is Jasmin who offers the first contribu-

tion. Rather than directly responding to the teacher's question, though, she

proposes that the report should start with the `̀ question'' that they were

trying to answer (4.11). In answering in this way, Jasmin shows that she

sees herself as a co-constructor of the plan for the report, thereby

confirming the teacher's evaluation of her as an interested and engaged

student. Her no doubt unconscious movement to push her hair behind her

ear as she answers suggests a certain degree of nervous tension in this

challenging situation.

Alex is obviously keen to contribute too. His proposal (4.13, 4.15) connects

with Jasmin's by presupposing that they are together constructing an ordered

sequence of elements: first the question, then what they did. Unlike Jasmin,

however, Alex does not make eye contact with either member of his audience

while he speaks; instead he fiddles with the flashlight. As a result, when he

pauses part way through his utterance, Jasmin is able to begin a comparable

proposal, actually completing hers at the same time as he does; significantly, she

also looks at the teacher as she speaks. When the teacher accepts the proposal,

therefore, it is to Jasmin that he responds; although he had been looking at Alex

while he spoke, the combination of Jasmin's directed gaze and her more forceful

manner, together with Alex's lack of eye contact, apparently led him to treat

Jasmin as the principal interlocutor. Alex gives the impression of accepting this

hijacking of his contribution, since he looks up briefly to Jasmin at the end of her

utterance and then down again at the flashlight, giving no indication of surprise

or dissatisfaction.

But he has not given up. As Jasmin confidently continues with a further

proposal (4.18), he takes the opportunity provided by her voiced hesitation to

attempt to get back into the discussion (4.19). However, because of his own

hesitation and his lack of eye contact, he is ignored by the teacher, who speaks

over him as he asks Jasmin which of the two proposed stages should come first

(4.20). Despite this, Alex still persists. Together with Jasmin, he responds to the

teacher's question, giving a different answer from hers. They both laugh at this

disagreement and at this point Alex does at last make eye contact; but it is with

Jasmin not with the teacher.

Nevertheless, the teacher has noted Alex's answer and looks at him, asking

him to justify his proposal. Once again, however, Alex does not make eye

contact as he speaks and his answer is hard to hear (4.25); he does not offer

the requested justification but seems to have changed his mind and to be

supporting Jasmin's preference for starting by describing the materials they

used. The teacher makes an accepting move that invites him to say more


(`̀ OK'' said with fall±rise intonation), but Alex does not continue. Perhaps he

is disconcerted by Jasmin's assertion of dominance when she stretches over to

switch off the flashlight that he is holding while he speaks. In any case, after

a couple of seconds of silence during which Alex, with downcast eyes,

continues to fiddle with the flashlight, the teacher turns to Jasmin and asks

her to justify her proposal. She responds, fluently and at length, with dramatic

gestures that underline the mental and affective states imputed to the

imagined readers of the report. Quite evidently, she is more than willing to

speak for both of them, taking on the role as if by right and with the

teacher's apparent encouragement.

In what follows, this pattern is repeated. Alex does not opt out of the

discussion, but his contributions are spoken softly, often simultaneously with

those of Jasmin, and he consistently looks down at the apparatus rather than

at his interlocutors as he speaks. Jasmin, on the other hand, speaks

confidently, and with strong gestures; she engages in frequent mutual gaze

with the teacher and on several occasions overrides Alex's attempts to enter

the discussion. Because of the two children's different modes of participation,

and because Jasmin's contributions better fit his scaffolding agenda, the

teacher colludes with her in her assumption of the role as spokesperson.

The result is that, although Alex is equally engaged in the activity, his less

assertive manner allows his contributions to be ignored by the other two

participants; what is more, he appears to be quite content to accept this less

central role.

CONCLUSION

Although I had no intention of investigating my own `̀ teaching'' when I

selected this videotaped interaction as a `̀ text'' for practical work in my class

on discourse analysis, the post-hoc action research in which I have engaged as

a result of my colleagues' interpretations of it has been doubly beneficial. First,

I have learned just how `̀ logocentric'' I have tended to be, both in my

interactions with others and in my studies of classroom discourse. From now

on, I shall try to be more attentive to all dimensions of meaning making in

face-to-face interaction.

Second, I have become fascinated by gesture itself and by the role that it plays

in interaction. Since I was apparently not aware of the significance of Alex's

gesturesÐnor indeed of my ownÐuntil I studied them on the recorded

videotape, it appears that gestures cannot function on the same conscious level

as the verbal meanings that are co-constructed in interactionÐat least not in the

normal course of events. On the other hand, where informal speech is not

accompanied by spontaneous gestures, it is typically experienced by listeners as

330 WELLS

lacking in conviction; similarly, for many speakers, to be debarred from

gesturing is to be deprived of an important means of communicating their

meanings. However, if neither speakers nor listeners are regularly conscious of

the gestures that accompany the utterances that they produce or hear, what

exactly is the nature of their contribution? This is a topic that clearly merits

further investigation.

From a broader educational perspective, however, the most important lesson

I have learned from this investigation is not to base analyses of interaction

solely on those aspects that are captured in a linguistic transcription. As I

hope I have shown, much that is significant for the manner in which joint

activities are so-constructed over time is to be found in semiotic behavior that

is non-verbal. Although participants may not normally be conscious of the

meanings that are realized through these modes, they can still be an important

mediator of coming-to-understand; furthermore, they can also provide valid

evidence of students' learning, if one is attuned to attending to them.

There still remains the question of how appropriate it is to analyze these

continuous dimensions of meaning in categorical terms. Certainly, with in-

genuity, means can be found to represent the behaviors themselves in an

analogical manner, using some of the multimedia resources that are now

becoming available. However, the greater challenge it seems to me, is the

absence of an appropriate metasemiotic in which to interpret the meanings that

are realized through these non-verbal modes and to theorize the relationships

among them.

Perhaps this state of affairs is itself what needs to be problematized. What does

it suggest about the relationship between practical and theoretical modes of

knowing and about the semiotic means that mediate them? To answer this

question it may be necessary, as Vygotsky (1981) suggested, to adopt a `̀ genetic''

approach and to look more closely at the trajectory from pre-speech modes of

representing in face-to-face joint activity to the visuographic modes that enable

asynchronous participation in a shared virtual reality. Phylogenetically, the

emergence of speech and writing can be situated on this trajectory; in each case,

the new mode added to the toolkit of available semiotic resources and, at the

same time, modified the roles of those that were previously available. Viewed in

this light, the new multimedia modes of representing can be expected eventually

to have equally far reaching effects, including perhaps the provision of an

alternative mode of interpretation to the categoriality that is so dominant in

written language. For the present, however, there seems to be no viable

alternative to written text as the chief mediator of theoretical discourse. We

should certainly exploit the new possibilities for representing those dimensions of

interaction that are continuous in their manifestation; but it seems that we shall

still have to use the categories provided by natural language to theorize together

about their significance.


NOTES

1. For both Jasmin and Alex, English was a second language. Both are Chinese± Canadian

students, born in Canada, whose first language is Cantonese. Both have been speaking English in

school since kindergarten, but Jasmin uses English more often at home; she has a native-like fluency,

whereas Alex seems to have a more limited command, particularly with respect to vocabulary.

However, this may be as much a lack of confidence as a lack of command of English.

2. By `̀ non-verbal,'' I mean those distinctions of meaning that are not encoded through

choices from the lexical or morphosyntactic systems of a language. Some aspects of intonation

in English, I concede, are grammatical (cf. Brazil, 1975; Halliday, 1970), but there are also

dimensions that are not, and it was these in which I was particularly interested. I should also

make it clear that my observations are based on my own dialect of (standard British) English,

although I believe they apply to a large extent to other dialects of English and to other

European languages.

3. See Deacon (1997) for a discussion of the possible evolutionary antecedents of this distinction.

4. This is not to say that non-verbal dimensions of interaction have gone unstudied; there is, in

fact, a considerable body of research on each of the channels I have mentioned. My point is, rather,

that they do not play a significant part in the lay-person's `̀ theory of speech''; nor has their

significance been widely recognized in work on discourse in the classroom.

5. I should like to acknowledge the significant part played by Mari Haneda in the development

of the scheme presented in this paper.

6. The term `̀ utterance,'' as used here, refers to the smallest unit of spoken text that, in

context, makes a potentially complete contribution to the ongoing discourse; it corresponds, we

believe, to what Chafe (1985) calls an `̀ idea unit.'' The segmentation of spoken text into

analytic units is notoriously problematic, and usually makes appeal to several criteria. In earlier

work on language development (Wells, 1981) we defined an utterance syntactically as `̀ one

independent clause and any elements dependent upon it''; however, intonation and pausing

were also taken into consideration. In our current work, we would give greater weight to

intonation, defining an utterance as one potentially free-standing tone unit and any tone units

dependent on it.

7. As we are well aware, a system with only five options falls far short of representing the full

range of distinctions in pitch movement found in most dialects of English. However, it was

considered adequate for the present purpose. More delicate descriptive schemes can be found in

Halliday (1970) and Brazil (1975). (For an alternative and more comprehensive notation, see Wells,

(1981), p. 278.)

8. Many aspects of non-verbal behavior are potentially significant in accounting for the way in

which an interaction develops. In the present case, only those aspects relating to the dimensions

selected for coding were systematically included in the accompanying description.

9. It has been pointed out that this distinction beteen action and demonstrative action is

somewhat similar to that between use and citation in linguistic description (Johnston, personal

communication, Nov. 1997).

10. McNeill's scheme of analysis is considerably more detailed than the one employed here.

Not only does it describe the shape of the gesture and its location in relation to the gesturing space,

but in mapping it onto the accompanying utterance, three phases are marked: preparation, stroke,

and retraction.

11. Because the results are based on the same data-set that was used to develop the coding

scheme, no attempt was made to carry out a formal evaluation of inter-coder reliability.

12. In interpreting these frequencies, it must be borne in mind that the three episodes were not of

equal duration. Episode one was substantially shorter than the other two, and episode three was the

longest (in part due to the interruption caused by the two boys with periscopes).

332 WELLS

REFERENCES

Bakhtin, M. M. (1986). Speech genres and other late essays. (Y. McGee, Trans.). Austin: University

of Texas Press.

Bateson, G. (1972). Steps to an ecology of mind. New York: Ballantine.

Berry, M. (1981). Systemic linguistics and discourse analysis: A multi-layered approach to exchange

structure. In M. Coulthard & M. Montgomery (Eds.), Studies in discourse analysis. London:

Routledge and Kegan Paul.

Bower, T. (1974). Development in infancy. San Francisco: W. H. Freeman.

Brazil, D. C. (1975). Discourse intonation. Discourse Analysis Monographs 1, University of

Birmingham, English Language Research.

Buzzelli, C. A., & Johnston, B. (1997). Expressive morality in a collaborative learning activity: The

creation of moral meaning. Paper presented at the Annual Conference of the American

Educational Research Association, Chicago, 24± 28 March, 1997.

Chafe, W. L. (1985). Linguistic differences produced by differences between speaking and writing. In

D. R. Olson, N. Torrance, & A. Hildyard (Eds.), Literacy, language, and learning. Cambridge:

Cambridge University Press.

Crowder, E. M. (1996). Gestures at work in sense-making science talk. Journal of the Learning

Sciences, 5(3), 173±208.

Deacon, T. W. (1997). The symbolic species: The co-evolution of language and the brain. New

York: Norton.

Halliday, M. A. K. (1970). A course in spoken English: Intonation. London: Oxford University Press.

Halliday, M. A. K. (1993). Towards a language-based theory of learning. Linguistics and Education,

5, 93±116.

Halliday, M. A. K., & Hasan, R. (1976). Cohesion in English. London: Longman.

Haneda, M. (forthcoming)

Harris, R. (1986). The origin of writing. London: Duckworth.

Jefferson, G. (1972). Side sequences. In D. Sudnow (Ed.), Studies in social interaction. New York:

The Free Press.

Kamberelis, G. (1995). Genre as institutionally informed social practice. Journal of Contemporary

Legal Issues, 6, 115± 171.

Kendon, A. (1988). How gestures can become like words. In F. Poyatos (Ed.), Cross-cultural per-

spectives in nonverbal communication. Toronto: Hogrefe.

Lemke, J. L. (1990). Talking science: Language, learning, and values. Norwood, NJ: Ablex.

Lemke, J. L. (1997). Making towers, making withs. Electronic Journal of Science Education, 2(1),

(http://unr.edu/homepage/jcannon/ejse/kamen_etal.html).

Lemke, J. L. (in press). Typological and topological meaning in diagnostic discourse.

Discourse Processes.

McNeill, D. (1992). Hand and mind: What gestures reveal about thought. Chicago: University of

Chicago Press.

Miller, C. R. (1984). Genre as social action. Quarterly Journal of Speech, 70, 151±167.

Olson, D. R. (1994). The world on paper. Cambridge: Cambridge University Press.

Sacks, H., Schegloff, E. A., & Jefferson, G. (1974). A simplest systematics for the organization for

turn-taking for conversation. Language, 50(4), 696±735.

Sinclair, J. M., & Coulthard, M. (1975). Towards an analysis of discourse: The English used by

teachers and pupils. London: Oxford University Press.

Singleton, J. L., Goldin-Meadow, S., & McNeill, D. (1995). The cataclysmic break between

gesticulation and sign: Evidence against a unified continuum of gestural communication.

In K. Emmerey & J. S. Reilly (Eds.), Language, gesture and space. Hillsdale, NJ: Erlbaum.

Schiffrin, D. (1994). Approaches to discourse. Cambridge, MA: Blackwell.


Vygotsky, L. S. (1981). The genesis of higher mental functions. In V. Wertsch (Ed.), The concept of

activity in soviet psychology ( pp. 144± 188). Armonk, NY: Sharpe.

Wardhaugh, R. (1985). How conversation works. Oxford: Blackwell.

Wegerif, R. (1997). Factors affecting children's talk at the computer. In R. Wegerif & P. Scrim-

shaw (Eds.), Computers and talk in the primary classroom. Milton Keynes, UK: Open

University Press.

Wells, G. (1981). Learning through interaction: The study of language development. Cambridge:

Cambridge University Press.

334 WELLS

Documents

Modes of Meaning in a Science Activity