The signature form of abstract deixis is the classic pointing G-hand aimed at absolutely nothing. This gesture is extremely widespread and frequent in conversational discourse. A speaker says "and there was [this guy]", and flexes an index finger at a certain locus, although there is no person there. In effect, the gesture `creates' a reference, in this case an imaginary person identified as "this guy". What I am terming `abstract deixis' is, according to Bühler (1982), concrete deixis that has been transposed to the realm of imaginary referents. His term for this process was deixis at phantasma. As in all cases of pointing, abstract deixis entails a coordinate center, or origo, and an indicated line connecting this center to a locus (the G-hand is an iconic sign of this line). A `target' then is created at this locus that may be used for later anaphoric references. The creative use of pointing in abstract deixis fleshes out the deictic paradigm with new meanings for the target and origo cross-hairs. Another component of pointing is the deictic field or ground over which the pointing act is understood to extend. In the concrete case, the deictic field is spatial and ranges from the immediate perceptual present to vast regions such as the entire Yucatan peninsula that Haviland (in press) describes in acts of pointing to distant locales. In the abstract case, the ground takes on non-spatial meanings. The components of deixis are all part of the referential creation of abstract pointing, and can be analyzed as parameters set by the speaker for the context. Creative deictic use is prominent in the example that I will now present.
Example of the creative use of deixis
The snippet in Table 1 is from a longer conversation recorded some 20 years ago by Starkey Duncan in the then Department of Behavioral Sciences at the University of Chicago. It features two previously unacquainted male graduate students. One student will be called A, a law student. The other student is B and he was a social work student.
The excerpt picks up at the end of what has been an extended effort by A to play the interactional game that Michael Silverstein (1993) has dubbed Getting to Know You. This takes a form typical among students, which is to exchange academic biographies. A has pursued this line for a number of turns. He had earlier asked, "Where did you come from before?", and B offered, "Mm, Iowa. I lived in Iowa." This led A down the garden path, however, since B proved unable or unwilling to contribute to Iowa as a topic, but the Iowa theme is relevant since it led to the exchanges in Table 1. After Iowa petered out, A resumed his quest for B's biography (Q means a question, R means a reply, A or B means the speaker, and the number of the question or reply is the ordinal position of the item in the snippet).
|Speaker A||Speaker B|
|QA6 how do you like Chicago compared to|
|QA7 did you [go to school thére] or uh
points to shared space
|RB7.1 I did go to school [there]
points to shared space
|RB7.2 [I went to school hére]
points to left
circles to left
|RB7.5 [ / um]
points to left
|RB7.6 so I [came back]
points to shared space
|RB7.7 [kind of /]
points to right
|QA8 an' [you wént to undergraduate hére or...............
points to shared space
|(A's gesture held)..............]|
|RB8 [in Chicágo] át, uh, Loyola
points to shared space
Silverstein's analysis of the text
Silverstein identifies "stretches of interactionally-effected denotational text". These are runs of local cohesion indexed via deictic references to past or present locations. In QA7, for example, A formulates his probe about B's past temporal location, `thenB', with "go to school". His goal was to elicit information about B's relationship to the past spatial location, `thereB'. This indexical probe, "go to school", carries onto the next step of the conversation a framework for coherence, for things holding together. The most recent denotational frame before QA7 was that of Iowa (either Iowa City and/or State). This would have been the default frame for the emphasized "thére" of QA7. This frame is picked up by B in his reply at RB7.1, when he says "I did go to school there". Yet ambiguity remains since A's "thére" can be a substitute for either "in Iowa" or "at Iowa" (and which, "in" or "at", is left unsaid). In other words, it could equally designate `C/SthereB' ("in") or `UthereB' ("at"). (C means City, S state, and U university.)
B does nothing to disambiguate the frame in RB7.1, where he repeats the precise formulation of A's QA7, using the same predicating phrase "go to school there". B however continues by clarifying the temporal order of the paradigm that he has set up, but not the institutional affiliation: he has gone to school `hereB', he says in RB7.2, as well as `thereB' in RB7.1. The result as Silverstein points out is a deictically organized progression of references that sketches B's academic biography ([emptyset]t means temporal succession):
in or at U/CChicago [emptyset]t in or at U/SIowa [emptyset]t in or at U/CChicago
This contains multiple ambiguities of deictic reference between "in" and "at", but the most important of these for the remainder of the snippet is, what `Chicago' is B speaking of: `The University of' ("at") or `the City of' ("in")? A pursues the topic once again and asks in QA8 if B had been an undergraduate at the University of Chicago ("an' you went to undergraduate here or"?), using a noninverted, confirmatory question that preserves the exact surface form of B's RB7, the predicate "go to school". Even this formulation by A is not without "denotational-textual wiggle-room". It would have been possible for B to have replied as though A had been asking if he had been an undergraduate `Chére', that is, in the City of Chicago, simply by saying "yes", for example.
But "for reasons unknown", B chooses to reveal that "most important of emblems of identity in professional- and upper-class America, the `old school tie'", and supplies the long-sought information in RB8 ("in Chicágo át, uh, Loyola"). B at last differentiates City and University -- though with some unwillingness. The result is the following now much clarified deictic structure in which the only ambiguity still present is whether `Iowa' meant the University as well as the State:
in CChicago, at ULoyola [emptyset]t in or at U/SIowa [emptyset]t in CChicago, at UChicago.
As it turned out, A also had "went to undergraduate" at an (albeit different) Jesuit institution. The conversation thereupon took flight and B's hard-won revelation led to many non-problematic exchanges on the theme of Jesuit education.
Analysis of the pointing
Not considered in Silverstein's textual analysis is the creative use by A and B of the gesture space via pointing. Analysis of the abstract deictic structures in the conversation will lead to an explanation of B's unexpected capitulation, in RB8. In general, the patterns of pointing were:
A points only into the shared or landmark space.
B points into this space and also points to the left and once to the right.
In the rest of this paper, `left' and `right' refer to directions from the viewpoint of the speakers (thus `left' is on the right of the screen, etc.). A is seated to B's right and the shared space or deictic field is between them. This space is roughly the overlapping part of their personal spaces (cf. Özyürek, in press). A and B were angled toward each other and were seated in such a way that they were equally visible to the camera. This has the advantage for our purpose that the left, center, and right spaces were the same for both speakers, thus obviating the ambiguity that could have arisen if one or both speakers had transposed the gesture space to the other's coordinate system. Even if this had happened, it would not have altered the left-center-right space and would not affect the following analysis.
The overlapping shared space acquires meaning as the discourse topic, and this meaning and its shifting values and the contrasts of other gestures to it are the subject of the analysis. The shared space initially had the meaning of B's academic past in Iowa, `Iowa-then'. As noted previously, this reference is ambiguous between the State of Iowa and the University of Iowa, and which was meant was not spelled out. The meaning at RB7.1 when B pointed to the shared space and said, "I did go to school [there]", thus could have been either the State or the University of Iowa.
A corresponding ambiguity exists during RB7.2-3 when B continued, "[I went to school hére] [álso], and pointed two times to the left, that is, away from the shared space. As with the verbal deixis, "hére", this left space could have meant either the City of Chicago or the University of Chicago, and following Silverstein, this will be written `C/UChicago-then'.
The meaning of the deictic field clearly changed for B at RB7.6, when he said "so I [came back]" and pointed to the shared space that previously had meant `Iowa-then' (the status of the shared space at RB7.4-5 is unclear). This meaning shift could have hinged on temporal updating. B wanted to move the topic into the present and thus contrasted `now' to the `then' that had been the left space at RB7.2-3. This contrast put `now' into the deictic field and `Chicago' came along with it. However, once imported, `Chicago' too became part of the deictic field for B. Thus, at RB7.6, the deictic field meant `Chicago-now', and this became B's new thematic reference point. But which `Chicago'? -- the City or the University?
I will argue that, at this moment, if not sooner (for we can't be sure about RB7.4-5), the deictic field for B meant the City. The crucial indication is that B pointed to the right at RB7.7 and hedged the reference to coming back with "[kind of /]". He was evidently saying that he had come back to Chicago, but hadn't come back to Chicago, and placed this Chicago1 vs. Chicago2 opposition on a new shared vs. right space axis.
I claim that the shared and the right space cannot have the same meaning; that one is the City and the other is the University (or is at least not-the-City), although we cannot say from the spatial contrast itself which space has which meaning. Subsequent pointing however soon makes this clear.
A now asks his fatal question (QA8): "an' [you wént to undergraduate hére or]" and points again to the shared space with an extended hold that is maintained during B's response. A's use is unambiguous: the space means for him the University (see note 3). B's response at RB8 also points to this space while he says, crucially, "[in Chicágo] at, uh, Loyola" -- the unexpected capitulation after a career of so much evasion.
The preposition "in" shows that B was indicating the City as opposed to the University. Thus the shared topic space for B at this point meant the City, not the University. This in turn suggests that the right space at RB7.7 meant the University as opposed to the City.
This meaning moreover would explain the hedge, "kind of". What B meant when he said "so I came back kind of", was that he had returned to one kind of Chicago (the City), but it was not the Chicago that might have been presupposed in this conversation (the University), where he and A were students and were talking (or alternatively, the "kind of" flagged not-the-City).
That B hedged and introduced a new spatial contrast also suggests that he was aware of the `C/UChicago' ambiguity. Had he been thinking only of his own meaning of `CChicago' for the deictic field, there would have been no motive for introducing a new space for `UChicago' (or `not CChicago') and the hedge. In other words B, without realizing it, revealed that his persistent `C/UChicago' ambiguity had been intentional.
That A and B had conflicting meanings for the shared topic space and that B was aware of this also explains why B suddenly gave up his resistance at this very moment. This is the puzzle that remains after Silverstein's analysis. B easily could have continued avoiding A and perpetuate the ongoing ambiguity, had he wished, by merely answering QA8 with "yes".
However, the deictic field and "here" had the City meaning for B while it meant the University to A. This contradiction confronted B with an interactional problem on a new level: the need to cease being merely evasive and to start lying; apparently B did not make this choice. (B could not avoid his dilemma by not pointing: A had already pointed into the shared space with the unambiguous meaning of `UChicago' and B had previously pointed to it with the opposite meaning of `CChicago'; moreover, A was continuing to point at the shared space with the contradictory meaning; B's confrontation with morality was inescapable.) That A maintained his pointing gesture during the entirety of B's response suggests that for A, also, there was a sense that the central gesture space had become a field of contention.
Thus the role of pointing into the deictic gesture space was an active one in this stretch of conversation. Pointing contributed to the dynamics of the conversation and included such interpersonal factors as evasion, probing, and confession. Table 2 summarizes the meanings given to the right, shared, and left spaces in the snippet.
|QA7 did you go to ...||S/UIowa-then|
|RB7.1 I did go to ...||S/UIowa-then|
|RB7.2 I went here||C/UChicago-then|
|RB7.6 so I came back||CChicago-now|
|RB7.7 kind of||UChicago-now|
|QA8 you went to undergraduate here||UChicago-now
(held through the following)
|RB8 in Chicago at Loyola||CChicago-now|
There may be other cultures in which handshapes or pointing angles differentiate dimensions of deixis. If A and B had been members of such a culture, there might have been less room for ambiguity and less potential for evasiveness. Consider, for example, the effects of having a system in which near and far are systematically differentiated in the form of pointing. B might have used the `far' version when he said "came back" and had in mind the City and not the University (where he physically was). The form of pointing then could have been a signal to A that B must have returned to the City, not the University, and B's evasion would have ended there. In other words, the smalltime drama of this conversation and B's moral dilemma were enabled, in part, by the parameters of handshape and space in pointing for a North American English speaker.
Distribution of pointing in the total conversation
Pointing was not uniformly distributed across the total A - B conversation. The conversation breaks naturally into three unequal parts. First there was a preliminary phase, prior to the initiation of A's attempt to discover B's academic history. A and B performed 14 gestures in this phase. Of these, 57% were metaphoric gestures of the `conduit' type (McNeill, 1992), 28% were abstract deictics, and 14% were other types or were difficult to classify.
The second phase was A's probing of B, which climaxed with the snippet that is the focus of this paper. A and B together produced thirteen gestures in this phase, 23% metaphoric and 77% abstract deictic (and no others). Thus, there was a dramatic upsurge of abstract deixis during this phase.
The third phase began immediately after RB8 with the following response from A:
A: óh óh óh óh óh I'm an óld Jésuit Boy mysélf / / unfórtunately
This phase was the actual start of the conversation in the sense that, from now on, A and B talked about a mutually agreed upon theme, the character of Jesuit education, the ways in which it is special and how it compares to experiences at the University of Chicago (too intensely intellectual), with A's "unfortunately" announcing the end of pursuit and the beginning of fellow-student camaraderie.
And the gesture situation again changed dramatically. In the third phase (by far the largest part of the conversation), abstract deixis virtually disappeared. Of 110 gestures from A and B, fully 93% were metaphoric, and only 6% were abstract deictic.
This near total disappearance of pointing can be explained with the aid of the concepts of the origo and the deictic field. What abstract deixis embodies is an orientation of the speaker toward a possible significant meaning (McNeill, Cassell & Levy, 1993). The gesture resolves the speaker into an origo and the proffered meaning into a deictic field. The gesture itself becomes the orientation of the first toward the second. In the second phase, when the topic of conversation was being negotiated, orientation to proffered meanings was a dominant cognitive mode of interaction, and abstract pointing thus prevailed and there was the above noted upsurge of abstract deixis in this phase. But once A and B had found their topic this motivation disappeared and with it the urge to point at abstract, objectively empty space. Other forms of gestural metaphor took over in its place.
Analysis of B's growth point in RB8
Although B's utterance "[in Chicágo] át, uh, Loyola" possesses minimal linguistic structure, it is interesting for the conditions under which it was formed. These conditions can be analyzed with the concept of a growth point (McNeill, 1992). This analysis will help generalize the growth point concept by exhibiting a case in which the core idea is abstract and moral rather than (as in most earlier examples of growth points) visual and spatial.
Under the prevailing imperative to orient himself to the proffered topic of his personal biography, B's thinking was dominated by the distinction between CChicago and UChicago on which his biography turned, and his apparent wish to blanket this distinction under the cloud of ambiguity provided by the word `Chicago'. The interpersonal context had developed through the contributions of both participants and included, critically, the shared use of the central gesture space. Thus, a dialectic between shared and individual cognition is a phenomenon of this snippet.
The conditions leading to RB8 include: (1) A and B's joint orientation to the shared gesture space, (2) B's awareness of his contradiction with A over the meaning of the shared space, and (3) the role of this contradiction in creating the moral dilemma that B ultimately confronted. The contradiction with A was one pole of the utterance and the resulting moral dilemma for B was the other. The contradiction was highlighted by A's protacted pointing to the shared space while B invoked the `CChicago' meaning. Together, these poles were the direct determinants of the form of utterance that we observe (in other words, this utterance could not have had any other form).
Idea units with multiple components are readily modeled by `growth points'. A growth point or GP is a psycholinguistic unit based on contrasts within `fields of oppositions' (it is called a growth point because it is the theoretical starting point of a thinking-for-speaking unit). The contrast and the field of oppositions within which the contrast appears are linked meaning structures that are partly under the creative control of the speaker at the moment of speaking (McNeill, 1992; McNeill & Duncan, in press; McNeill, in press).
In the case of RB8 the field of opposition, as B construed it, was something like To Lie About Loyola vs. To Tell The Truth About Loyola. B's contrast was the To-Tell-The-Truth pole. That is, B's meaning at this point was not just the denotational content of "in Chicago, at Loyola", but the moral content of coming out with the truth when the alternative was lying. This hidden content was, I believe, the core of his meaning at this moment, and the various parts of this meaning were materialized in one or both of the modalities, speech and gesture. B's contradiction with A materialized via pointing at the space that A had designated as `UChicago' but meaning by this space, `CChicago'. The "in" lexical choice brought out CChicago, which is the `Truth' alternative. The "in" - "at" succession arose from the `C/UChicago' ambiguity that B had been perpetuating. Having separated the City meaning with "in", B went on to lay out the University component with "at". The stress pattern, "in Chicágo - át", displays this contrast within a consistent rhythmic pattern (that is, "ín Chicago - át", or "in Chicágo - at Loyóla" -- the two other possible combinations -- twist the rhythm and lose the contrast that splits out the University concept as something distinct from `CChicago'). The "át", in turn, led to "Loyola" but with hesitancy as if completion of the City-University paradigm had taken on a life of its own and was unfolding somewhat against the will of the speaker. The GP thus incorporated information about the contradiction with A and B's awareness of it, plus B's sense that he was confronting a moral dilemma and his decision to resolve it. B's unpacking of the GP into "[in Chicágo] át, uh, Loyola" grew out of the contrasts built into it, despite B's squeamishness over the final revelation. Thus, according to this model, the utterance was a product of B's individual thinking at a particular moment in a specific pragmatic-discourse context.
The intrapsychic/interpsychic interface
The process just described illustrates the intersection of Vygotsky's (1987) two planes, the interpsychic and the intrapsychic (only the interpsychic tends to be discussed in the conversation analytic literature, where Vygotsky's theory often undergoes an `intraectomy'; cf. Duranti and Goodwin, 1992). The GP is intrapsychic in the Vygotskian dichotomy (awareness of the contradiction, the moral dilemma) and yet it intersects interpsychic material (the contradiction with A, the decision to confess). It is important to maintain this distinction, lest the mind be seen as no more than a passive imprint of the social interaction. The challenge, seen by Vygotsky, is to figure out how the mind remains intact while it merges with the social context. The GP offers a picture of this. The same gestures and linguistic categories appear on both planes and thus bridge them. The GP describes how individual thinking internalizes and brings content from the "interactionally effected" framework into idea units that support, indeed cannot help but generate, textual coherence.
Though the same interactional content appears on the two planes, the content has different functions on each plane. The most visible manifestation of this functional differentiation in our snippet occurred at RB8 when both B and A were simultaneously pointing at the shared space but with opposite intended meanings. Interpsychically, this was a tussle over the meaning of the space. Intrapsychically, it embodied B's dilemma, whether to lie or tell the truth. On the intra- plane, the content was part of B's personal mental life and subject to autochthonous forces of his own, on the inter- plane it was subject to the social forces of the interaction between A and B. The point is, both planes are sources of the representations running through B's mind at this moment. The GP is the unit, the point where various forces come together. While the GP is on the intra- plane, it ties together the influences on thought and action that scatter over the interpsychic and intrapsychic planes. Vygotsky said that everything appears in development twice, first on the social plane, then on the individual. The same logic applies to the GP. He saw the necessity of unit formation to encompass this transformation, invoking the concepts of psychological predicates and inner speech to express the unity of thought and action in the minds of agents. The growth point concept is heir to this insight.
Bühler, C. (1982). The deictic field of language and deictic words. In R.J. Jarvella & W. Klein (eds. and trans.), Speech, place, and action (pp. 9-30). London: John Wiley & Sons.
Duranti, A, & Goodwin, C. (1992). Introduction. In Duranti, A. & Goodwin, C., Rethinking context: Language as an interactive phenomenon. Cambridge: Cambridge University Press.McNeill, D. (1992). Hand and mind: What gestures reveal about thought. Chicago: University of Chicago Press.
McNeill, D., Cassell, J., & Levy, E.T. (1993). Abstract deixis. Semiotica. 95-1/2, 5-19.
McNeill, D. & Duncan, S.D. (in press). Growth points in thinking-for-speaking. In D. McNeill (ed). Language and gesture: Window into thought and action..
McNeill, D. (in press). Models of speaking (to their amazement) meet speech-synchronized gestures. In D. McNeill (ed). Language and gesture: Window into thought and action..
Özyürek, A. (in press). The influence of addressee location on spatial language and representational gestures of direction. In D. McNeill (ed). Language and gesture: Window into thought and action..
Silverstein, M. (1993). A minimax approach to verbal interaction: Invoking `culture' in realtime discursive practice. In Workshop on language, cognition and computation. Barcelona: Fundació Catalana per a la Recerca (Nov. 24-26).
Slobin, D. 1987. Thinking for speaking. In J. Aske, N. Beery, L. Michaelis, & H. Filip (eds.), Papers from the 13th annual meeting of the Berkeley Linguistics Society (pp. 435-445). Berkeley, CA: BLS.
Vygotsky, L. S. 1987. Thought and language, A. Kozulin (trans.).
Cambridge, MA: MIT Press.
Preparation of this paper was supported by a grant from the Spencer Foundation. I wish to thank Michael Silverstein for commenting on an earlier draft. The term `abstract deixis' was used by McNeill, Cassell & Levy (1993).
 Pointing is not limited to the G-hand and requires only an extensible body part -- hand, arm, and head are the most common. The analysis in this paper does not depend on the specific form of gesture.
 "[A] seems to blend two simultaneous informational quests in his utterance, which makes for a rather strange discontinuous colloquial phrase with focal stress, `wént ... hére' superimposed upon the ... repetition of [an] earlier construction... . The different focalization of these two blended constructions leaves no doubt to us analysts which is the more important piece of information being asked for; it is the undergraduate institution with which Mr B's identity can be affiliated" (Silverstein, 19XX, p. 48).
 In describing gestures, square brackets show the speech with which the gesture co-occurred. "/" signifies an unfilled speech pause. All gestures are deictic, mostly with the classic G-hand shape.
 The A - B conversation was transcribed by Starkey Duncan and was first used for targeted analysis in the early 1980s by a team known as the Anaphora Workshop at the University of Chicago. The Workshop included, if memory serves, Elena Levy, Maya Hickmann and Rebecca Passaneau.
 Abstract pointing is also a form of gestural metaphor, in which non-spatial contexts are structured in terms of origo, target, deictic field, and orientation.