Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
■ Michael Silverstein
THE UNIVERSITY OF CHICAGO
M
y largest point is that interdiscursivity plays on “-evals,” that is, on the logic
of the chronotope in Mikhail Bakhtin’s (1981:84) sense. This dvandva com-
pound of course should now be familiar to any linguistic anthropologist
studying discourse, denoting the temporally (hence, chrono-) and spatially (hence,
-tope) particular envelope in the narrated universe of social space-time in which and
through which, in emplotment, narrative characters move.1 Bakhtin’s nice coinage it-
self was, of course, pointedly interdiscursive with general cultural appropriation of
theories of the relativity of “space-time” in physics and other sciences (cf. Clark and
Holquist 1984:278ff.), though he achieved better poetic effect in the ordering of syl-
lables by taking “time” before “space.”
Journal of Linguistic Anthropology, Vol. 15, Issue 1, pp. 6–22, ISSN 1055-1360, electronic ISSN 1548-1395.
© 2005 by the American Anthropological Association. All rights reserved. Please direct all requests for per-
mission to photocopy or reproduce article content through the University of California Press’s Rights and
Permissions website, at http://www.ucpress.edu/journals/rights.htm.
6
03.JLIN.15.1_06-22.qxd 5/12/05 12:09 PM Page 7
I use the more conceptually primordial term interdiscursivity here, rather than the
term intertextuality (which we will get to in its due place), and I index by this choice
of stem the primordiality of discourse, the processual, real-time, event-bound social
action that it is our task to analyze, discovering therein textual structures, according
to principles of arrangement of the very signs of discourse. Communicational inter-
discursivity is a relationship of event to event and is projected from the position of
the personnel—authorial and/or animating senders, responsible receivers, nonre-
sponsible monitors, et cetera—of some particular event in respect of one or more oth-
ers. Intertextuality—from which we can caption the “intertext,” a structural principle
of textuality that remains constant or recognizable across discursive events—is a di-
rectionally neutral state of comparability of texts in one or another respect, to be sure.
But as linguistic anthropologists we are, of course, interested in how intertexts are
created, that is, how they are generated in events of communication through tech-
niques of interdiscursivity deployable as role strategies of the participants.
At the outset, moreover, we must keep in focus the kind of semiotic phenomenon
we are dealing with. Hence, to approach interdiscursivity at the level of commu-
nicative events I invoke the appropriate, if more semiotically elementary, sign rela-
tionships of iconicity and indexicality in the Peircean idiom. These are the
characteristic modalities of language in use, rather than the specifically linguistic
property of symbolic form, of full Saussurean abstract signs that are relevant to the
analysis of sentence structures as instances of denotational codes. After all, films,
paintings, theatrical productions, spectacles, events in the depicted universe of a set
of 2002 marketing-year Volkswagen ads, et cetera all have interdiscursivity and even
intertextuality about them—hardly linguistic objects in the strict sense, even if some-
times language is a component medium of their realization as events.
Events of using language (or any other cultural form, for that matter) can be in-
terdiscursive one with (an)other(s) if they seem to form a set of some kind. What
kind of set? A set defined by perceived “likeness” of some sort—hence, “iconicity”;
likeness of one with the other or others in the set, to whatever highly specific or
vaguely schematic degree. And co-membership in a “likeness” set, a class of objects
“like” one another in some respect, is, of course, an “indexical” relationship, a rela-
tionship of co-occurrence within a frame. Now, from what we know about language
and its discursive deployment, in what respect can such discursive events be “like”
one another? They can be alike as to what Bakhtin (1986) terms speech genre, that is,
in respect of entextualization (bounded textual qualities) and/or contextualization
(indexical relations to event frame), in great detail or only in more abstract[able]
characteristics.
structures of likeness come into being having no chronicity as such—for time as such
is logically obliterated in the symmetric STRUCTURAL RELATIONSHIP of likeness—such a
structure of likeness is termed here intertextuality.7 Under intertextuality, two or more
texts-in-context (individuable objects) become TOKENS OF A TYPE, thus “the same” in
some respect or respects.8 Note how grammar, too, the presupposed denotational
code underlying denotational textuality, is such an achronic condition on all possible
denotational intertextuality, with infinitely many realizable tokens past, present, and
future in a plane of token-level coevalness termed synchrony. Of course, grammar in
its distinctively human form and infinite generative capacity instantiates as well a
different order of semiosis, the symbolic, in addition to the iconic-indexical with
which we have been concerned.
reference at that moment, serves to extend the referent uniquely under those discur-
sive conditions by indexically invoking its associated description or characterization.
Now where do the descriptive backings, the characterizability conditions on ex-
tendable denotata, as I prefer to term them, come from?10 Where, in particular, does
the characterizability condition come from that is associated with your standard av-
erage definite referring expression headed by a common noun, for example, English
(the) man over there with the martini (headed by man)? For some referring expres-
sions, such as personal deictics (English I/we, you) or demonstrative deictics
(this/these, that/those), the characterizability conditions are clearly just descriptions
of the relationship between the extended entity, the “Referent” of the communicative
event, and the persons inhabiting the indexically presupposed roles of “Sender” and
“Receiver” in their spatiotemporal chronotope.11 Hence, I is a referring expression
because when a token of it is used, it invokes-as-true of the Referent that it is identi-
cal with the entity in the role of Sender, et cetera. As I have noted (Silverstein 1976),
these characterizability conditions invoked by various deictic paradigms are inher-
ently metapragmatic in character; they are descriptions of some aspect of the very
context in which tokens of the forms are used.
Proper names are interesting in this respect, and they play a central role in the
Kripke-Putnam approach to the relations of intensions as characterizability condi-
tions on referring expressions to extensions, actual extended denotata in acts of ref-
erence. Though their ideas need much more in the way of sociocultural sensitivity to
constitute an empirical theory adequate to the task, the general observation is that
any use of a proper name as a referring expression is, in effect, either its “baptismal”
use, an event of performative nomination, or a use that harkens back to some such bap-
tism by interdiscursive renvoi. A proper name, recall, is a linguistic form that refers
to some unique individual entity across all possible worlds every time it is used; it
is, as Kripke terms it, a “rigid designator” (1972:269–270). Hence, once a token of the
linguistic form is, on some occasion or other, attached to an entity-as-referent by duly
authorized—that’s key!—baptism, it creates an event-origo in the universe of all pos-
sible worlds calibratable to that event in relation to which all further tokens of the
name will still apply to that unique entity so baptized. Of course, if it is an event of
baptism we are speaking of, it has roles—Sender, Receiver(s), baptized Referent cer-
tainly, Audience perhaps, et cetera—which are inhabited by individuals through
rules of role-recruitment (what kind of person can use what form to what kind of
person referring to what kind of thing, etc.). Every proper name, then, rests on nor-
mativities of a system of role-recruitment in which social differentiations of “natural”
social kinds of entities are stipulated, to varying degrees, both positively and nega-
tively as the inhabitants of the various performative event roles (as we would of
course expect from so-called speech act theory). Parents of a newborn in bourgeois
mainstream America in effect whisper their choice of name to a physician or equiv-
alent named licensee of the state who, filling out a birth certificate, baptizes the
neonate for secular and civil purposes, performatively creating a new individual; a
religious officiant does so, on the recommendation of the parents, creating an indi-
vidual within the religious corporation. For first, given, or “Christian” names, there
are two great, but slightly overlapping, grammatical sets of name-forms such that to-
kens of one set are used when Referent is recruited from the “natural” social category
of males, and tokens of the other set are used when recruited Referent is female. For
some, ethnic categoriality is a basis of role-recruitment of the Referent of a baptism,
whether positively or negatively. And so on. As we see, for baptism as an event, it is
the metapragmatic description of the categories from which Referent and sometimes
Sender, Receiver, Audience, et cetera are recruited to these roles that constitutes
whatever we might call the descriptive backing or characterizability condition on the
proper name form used in the event.
But not only in the event of baptism: ever after. Any subsequent use of a token of
the form that is the proper name of an entity in an event of reference may have its
03.JLIN.15.1_06-22.qxd 5/12/05 12:09 PM Page 12
So, reading back purely at the type level of this register-form’s effect in our own
experience, it was difficult at first for the students to appreciate Shakespeare’s art of
ironical voicing as he animates social identities for the delight—and disdain—of his
Elizabethan audience.
lexical item, and it is that lexical item that is now used even in the performative one-
word utterance.
Note how delocutionaries are inflectable with tense, aspect, mode, et cetera as
verbs, nouns, or any other conforming lexical class, as shown as well by Benveniste’s
examples of French saluer (to greet) and such. Etymologically, they are a growing
edge of the metapragmatic vocabulary essential to having grammaticalized explicit
primary performative formulae, Searle’s (1969:30ff.) “illocutionary force indicating
devices.” The fact is, such explicit primary performative verbs come and go in the
history of languages. The set of so-called primary IFIDs grows especially from delo-
cutionary processes that constitute lexical forms themselves ultimately undergoing
all of the normal lexical possibilities for derivation.
poraneity or “eval”ness of whose evil is central to our engagement with the allegory.
Within the filmic diathesis, both characters/periods constitute origines from which
the other is projected, with Austin Powers bringing to the 1990s global pop-cultural
and pop-political horizon an invariantly 1960s/70s sensibility and style that some-
times, like the proverbial stopped clock, is, by chance (Chance the Gardener?), effec-
tive, if generally inappropriate in the new “-eval” context.
In a way, too, for the audience the fun is designed to be the inverse of my students
vis-à-vis Shakespeare; it requires the recognition by the viewers of these films that
the 1970s are gone—good riddance!—in both their Bondesque and Powersian ver-
sions, because the 1990s are so cool, and yet plus ça change . . . . Austin Powers, Dr.
Evil, and the rest of the characters both in the 1970s and in the late 1990s/2000s con-
stantly travel back and forth and encounter each other interdiscursively in the coeval
and noncoeval modes (notwithstanding the Evil Family history!).
Rip Van Winkle, Woody Allen’s “Sleeper,” Flash Gordon and Ming the Merciless
of the WWII era, various more recent space/time-travel movies, as well, of course, as
“007” constitute the intertexts here of the first-order interdiscursive spoof, which is
merely the starting point for what we might term the “higher” interdiscursivity of
stylistic types within the narrated filmic universe.
Notes
Acknowledgments. This article was originally prepared, in telegraphic style, for the sympo-
sium “Intertextuality,” organized by Asif Agha and Stanton Wortham, at the annual meeting
of the American Anthropological Association, New Orleans, 21 November 2002. On the basis
of revealing reactions to the telegraphy from three referees and from the editors, it has been
considerably expanded while being revised for publication, a process of—one hopes!—clarifi-
cation that thus owes much to my interlocutors, known and unknown.
1. Of course, as many writers have shown in one or another aspect, following Bühler (1990),
the chronotope of the narrated world, where characters (denotata) “live,” is indexically pro-
jected from the here and now of the world of narration, the world of discursive practice.
Grammatical deixis (indexical denotation) of the usual spatiotemporal sort is a central formal
mechanism for calibrating the two universes, as are such paradigmatic deictic markings as
person, mode, et cetera. The deixis necessary to align the two realms and yet keep them inter-
actionally distinct is obviously a relative (and relativistic) one, experienced as “transposition”
of deixis within the world of narration. At the same time, actual tropic deictic transposition—
as in the use of the narrative present tense for non-ongoing narrated events—does, in fact,
occur. Haviland (1993, 1996) and Ochs (1994, 1996, 2004; Ochs et al. 1996), among others, have
explored such transpositions and their effects, Haviland in particular in respect of the roles of
verbal and gestural deixis in achieving “transposition.” As well, Bakhtin’s (1981:301ff.) central
point about the (to him) high-water mark of literary modernity, the realist novel, can be simi-
larly understood. He notes how authors use the natural “heteroglossia” (inherent sociolin-
03.JLIN.15.1_06-22.qxd 5/12/05 12:09 PM Page 18
guistic variability) of any language community indexically to create narrative “voices” of char-
acters. Such “voices” exist, of course, only in relation to presupposable social positionali-
ties that variant uses index in the universe of narration shared by a novel’s author and
readership—as well, in realism, by a novel’s characters in its narrated world. See Hill 1995;
Silverstein 1999; Wortham 1996, 2001, 2003; and Koven 2001, 2002 for some applications of
Bakhtinian technique to situations of narration within interactional frames.
Bakhtin further points out that heteroglossia manifests as an inherent macro-sociological in-
terdiscursivity in all discourse, making narration about characters in effect doubly “dialogic”
in a special and idiosyncratic sense of this last term, once explicitly and once implicitly. In het-
eroglossia-based “voicing,” the narrative artist uses phenomena of interdiscursivity to create
component partials, “voices,” of an intradiscursive textual structure. Nevertheless, it is im-
portant to keep the concept of interdiscursivity as developed here for the universe of discur-
sive interaction conceptually distinct from both narrated world (denotational) chronotopes
and from heteroglossia-based “voicing.” This footnote responds, with thanks, to several use-
ful queries of anonymous referees about differentiating these various concepts in the linguis-
tic anthropological literature.
2. Greg Urban has recently (2001) written about Metaculture in relation to regimes of virtual
circulation from originary to secondary to . . . n-ary occasions, generalizing from the specifi-
cally discursive situation, though exemplifying his points principally from the realm of dis-
course, in which the key entity or form is the text, a causally consequential structure of relative
coherence that emerges in and is precipitated by discourse. Some discourse is explicitly
metadiscursive; some only implicitly so. We must be careful, of course, to differentiate the text
in this sense, a completely socio-spatiotemporal entity, from the text artifact, with its “thingy”
quality of potential physical movement through time and space in its own regime of circula-
tion, for example, in the commodity form.
3. I have spelled this out in elaborate detail in a paper at the 1996 San Francisco AAA meet-
ing, “The ‘simultaneity’ of grammar and the ‘poetics’ of simultaneity,” but summarize here.
4. Note that self-styled “Conversation Analysis,” with its unique focus on the adjacency pair
in turn-taking, is hard-pressed to account for these hierarchical layerings—long in the litera-
ture—and the poetic structures formed in interaction with relevantly “equivalent” (cf.
Jakobson) sequential units at some durational remove one from another. (See Goffman 1976;
Merritt 1976; and Silverstein 2003a:197–201, reanalyzing a transcript in Levinson 1983:305.)
The best one can do within the strictures of CA is to model interaction as a kind of interac-
tional-text-level X-bar structure of constituent pair-part fractions, “headed” by a highest-level
interactional “move,” for example first pair-part “request” by someone as the culminative se-
quel to a “pre-request” sequence, perhaps. This is, of course, a theoretical aporia in an ap-
proach that purports to model only earlier-to-later local adjacency structures; worse, it is not
actually locally determinable only “in” transcript.
5. In Silverstein 2003a:197–200 I elaborate the point that text partials seem functionally to
nest one inside another, yielding a kind of hierarchical structure. This results from the fact that
text-forming metricalizations, though subtending different-sized scopes of signal duration, are
laminated one onto another in the total (co(n))textual effect. So, once indexically established,
certain aspects of role inhabitance perdure over interactional event-time, in effect “framing”
other aspects of role inhabitance. In the example referred to here, once each of two interacting
people have assumed the roles of “customer” and “salesclerk,” these identities endure and
frame what emerges in the more fluid and quicker-cycle inhabitance of the roles in conversa-
tional dialogue. Customer and salesclerk alternate as “initiator” and “respondent” of various
adjacency-pair Q-and-A metricalizations of turn-taking, particularly as the salesperson tries
coaxing a necessary piece of information out of his presumed customer—information that
never comes.
6. Recall here Jakobson’s widened usage of the term to mean “having determinate semantic
value one with respect to another,” as denotationally distinct elements within a syntagmati-
cally delimited paradigm or a positionally delimited “poetic function.”
7. Pointing up the importance of the distinction between interdiscursivity, with its event-
based vectorial dynamism, and intertextuality in the sense just introduced, one referee use-
fully noted that “[interdiscursivity] juxtapose[s] texts in such a way as to challenge their
apparent similarity, to construe them as fundamentally not tokens of a type, as in ironic, satir-
ical, and parodic intertextuality” (emphasis added—MS)—to which one can only say, “Yes!
That’s the point!” These are tropic transformations of otherwise cotextual relationships, for ex-
ample, negating a text and its taken-for-granted interpretation by metacommunication, negat-
ing it by subtle substitution, and negating it by exaggeration.
03.JLIN.15.1_06-22.qxd 5/12/05 12:09 PM Page 19
8. Here is the birth of the concept of textual genre, though the concept more usefully lies
within a metapragmatics of interdiscursivity as in Bakhtin’s elaboration of the “speech genre”
as the unit of analysis.
9. Indeed, one of the lines of division between self-styled “formalist” and “functionalist” ap-
proaches to language (grammar) is the issue of the reducibility of principles of grammar,
specifically of syntax, to principles of what we are terming here text-in-context. See Nichols
1984 for an early summary of the issues along the formalist/functionalist divide. Hopper and
Thompson (1984, 1993), Hopper (1996, 1997), and Bybee and Hopper (2001) all develop views
that largely eschew a distinct formal organization identifiable as grammar. By contrast, note
the formalist views about the well-defined autonomy of at least syntax put forward by Baker
(2003) and Newmeyer (1998, 2001, 2003).
10. I have treated this matter from a grammaticosemantic and -pragmatic point of view in
two earlier articles (Silverstein 1981, 1987) about the organization of categories of denotation
in relation to types of descriptive backing associated with each of the grammatical types: per-
sonal deictics, demonstratives, anaphors, proper names, relational and absolute status terms,
. . . , abstract nominalizations. Such grammatically (formally) differentiated denotational ex-
pression-types differ systematically in the ingredients of the intensional information commu-
nicated-as-true about denotata by their use. The systematic differences can be captured in the
figure of a multidimensional denotational “space” not unlike the phonetico-phonologic
“space” in which sound segment-types (segmental phonemes) are locatable according to val-
ues of dimensions of contrast (“features” in the case of phonology).
11. See Hanks’ (1992) useful phrasing of the distinction between indexical presupposition of
features of contextual “ground” and relationally picked-out referred-to “figure” in sensorially-
based systems of deixis.
12. Study of “naming systems” in sociocultural anthropology or of “systems of address” in
social psychology and interactional sociolinguistics is all about delimiting the universe of
metapragmatic regimentations on baptism and subsequent events of usage of particular par-
adigms of pragmatic alternants that together constitute an individual’s “name,” “title,” or
equivalently usable linguistic form. Braune 1988 has a good bibliography on this point.
13. Observe that the whole area of refiguring human categorization in terms of prototype
extensional objects-of-reference of terms, of interdiscursively emergent stereotypes about the
class of denotata of such terms, et cetera is consistent with this approach to meaning. See now
Taylor 2003 and Aarts et al. 2004, especially parts II and V.
14. In Silverstein 1998:128-135 and Silverstein 2003a, I have elaborated this dialectic of ex-
tensionalization—(re)intensionalization via the abductive power of linguistic ideology in ex-
plaining how all social indexicality works. Indexicals are not “natural” but are sociocentrically
emergent emblems (conventional iconic indexicals) to the extent they are always already sub-
ject to naturalizing essentializations in the processes of usage.
15. Cited from The Oxford Shakespeare paperback edition (1998), ed. H[arold] J[ames] Oliver
(Oxford: Oxford University Press), from which I was required to teach. The play itself likely
originated ca. 1592, though the text cited comes from a critical collation of quarto and folio ver-
sions of the early 17th century.
16. Benveniste’s (1966:277–285) term in French is verbes délocutifs, and his translator for the
1971 English edition of Problems of General Linguistics, Mary Meek, follows the French deriva-
tional morphology in creating the solecism *delocutive. However, it is clear from Benveniste’s
own account that the underlying nominal base from which he derives his—morphologically
conforming—French term is the French locution, by an analogical parallel, as he explicitly says,
to the Latinate linguistic terms [verbe] dénominatif < nom, [nom] déverbatif < verbe. The English
noun locution-, on the other hand, yields the whole parallel set of adjectives already in exis-
tence in linguistic-pragmatic register based on its English adjectival derivative, locutionary,
viz., illocutionary, perlocutionary. This dictates a conforming adjective delocutionary deriva-
tionally parallel to denominal [verb or adjective] and deverbal [noun].
17. There is another metapragmatic descriptor, in verbal form (to) cluck the/one’s tongue,
in nominal generic form tongue-clucking, that seems to describe the performance of this
sound as a performative of “disapproval” on the part of the tongue-clucker.
18. The Bush-II administration has gone back to the pre-Eisenhower-era practice of redact-
ing and cosmetically editing “official” versions of the president’s “spontaneous” utterances,
carefully negotiating their public quotation and broadcast with media outlets that must, after
all, be admitted into the Executive presence for continued access.
19. The New York Times cheered “Mr. Bush’s New Gravitas” on 12 October 2001, the day fol-
lowing an East Room press conference. “He seemed more confident, determined, and sure of
03.JLIN.15.1_06-22.qxd 5/12/05 12:09 PM Page 20
his purpose and was in full command of the complex array of political and military challenges
that he faces,” gushed the editorial page: “Mr. Bush clarified and sharpened his positions on
several important issues.” A not uncharacteristic empirical sample is quoted above.
20. I have indicated each of the most glaring of grammatical, lexical, and stylistic problems
with a distinct font treatment in my citation (not present in the Times transcript).
References Cited
Aarts, Bas, David Denison, Evelien Keizer, and Gergana Popova, eds.
2004 Fuzzy Grammar: A Reader. New York: Oxford University Press.
Baker, Mark C.
2003 Lexical Categories: Verbs, Nouns, and Adjectives. Cambridge: Cambridge University
Press.
Bakhtin, Mikhail M.
1981 The Dialogic Imagination: Four Essays. Michael Holquist, ed. Caryl Emerson and
Michael Holquist, trans. Austin: University of Texas Press.
1986 Speech Genres and Other Late Essays. Caryl Emerson and Michael Holquist, eds. Vern
W. McGee, trans. Austin: University of Texas Press.
Banfield, Ann
1993 Where Epistemology, Style, and Grammar Meet Literary History: The Development of
Represented Speech and Thought. In Reflexive Language: Reported Speech and
Metapragmatics. John A. Lucy, ed. Pp. 339–364. Cambridge: Cambridge University Press.
Benveniste, Emile
1966 Problèmes de linguistique générale. Paris: Editions Gallimard.
Braune, Friedericke
1988 Terms of Address: Problems of Pattern and Usage in Various Languages and Cultures.
Berlin: Mouton de Gruyter.
Bühler, Karl
1990[1934] Theory of Language: The Representational Function of Language. Donald Fraser
Goodwin, trans. Amsterdam: John Benjamins Publishing Co.
Bybee, Joan L., and Paul J. Hopper, eds.
2001 Frequency and the Emergence of Linguistic Structure. Amsterdam: John Benjamins
Publishing Co.
Clark, Katerina, and Michael Holquist
1984 Mikhail Bakhtin. Cambridge: Belknap Press of Harvard University Press.
Goffman, Erving
1976 Replies and Responses. Language in Society 5:257–313.
Halliday, Michael A. K., and Ruqaiya Hasan
1976 Cohesion in English. London: Longman Group Ltd.
Hanks, William F.
1992 The Indexical Ground of Deictic Reference. In Rethinking Context: Language as an
Interactive Phenomenon. Alessandro Duranti and Charles Goodwin, eds. Pp. 46–76.
Cambridge: Cambridge University Press.
Haviland, John B.
1993 Anchoring, Iconicity, and Orientation in Guugu Yimithirr Pointing Gestures. Journal of
Linguistic Anthropology 3:3–45.
1996 Projections, Transpositions, and Relativity. In Rethinking Linguistic Relativity. John J.
Gumperz and Stephen C. Levinson, eds. Pp. 271–323. Cambridge: Cambridge University
Press.
Hill, Jane H.
1995 The Voices of Don Gabriel: Responsibility and Self in a Modern Mexicano Narrative. In
The Dialogic Emergence of Culture. Bruce Mannheim and Dennis Tedlock, eds. Pp.
97–147. Urbana: University of Illinois Press.
Hopper, Paul J.
1996 Some Recent Trends in Grammaticalization. Annual Review of Anthropology
25:217–236.
1997 Discourse and the Category ‘Verb’ in English. Language and Communication
17:93–102.
Hopper, Paul J., and Sandra A. Thompson
1984 The Discourse Basis for Lexical Categories in Universal Grammar. Language
60:703–752.
03.JLIN.15.1_06-22.qxd 5/12/05 12:09 PM Page 21