Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Chapter Prospectus

Chapter 7, From Monologism to Dialogicality, engages the fact that spontaneous spoken discourse is essentially dialogical and is correspondingly most clearly recognizable through the observable phenomenon of turn-taking. Language use is always a social and cultural engagement rather than some sort of solipsistic behavioral phenomenon, as it would appear to be in the reductionistic treatments and artificial experimentation of mainstream psycholinguistics. Empirical research dare not neglect this essential component of spontaneous spoken discourse.

Where Are We?

In our search for a concrete way to contrast monologism and dialogicality, we encountered an appropriate example in a presentation at a recent Psychonomic Society convention. A paper by Swets, Ferreira, and Altmann (2006, November 18, p. 29) manifests the differences more clearly than any example we might have concocted. The paper was entitled “Where was I?”: A psycholinguistic investigation of interruptions to language production. We include here their entire abstract:

When people communicate in a dialog, the speech stream of one speaker is sometimes interrupted by the speech stream of another. In such cases, it is often difficult for the interrupted interlocutor to return to the point where he or she left off. Hence, interruptions present an interesting problem concerning language production: How do speakers keep track of where they were before being interrupted? We report four experiments that investigate this unexplored issue. Experiment 1 used a seminatural dialog in which a confederate interrupted at predetermined narrative junctures. Measures of resumption difficulty reveal that interruptions with conflicting conversational goals are particularly disruptive, as are interruptions requiring long verbal responses before resumption. Experiments 2 through 4 investigated similar processes for sentence production. Results demonstrate that verbal and nonverbal interruptions early in sentence production are more disruptive than later interruptions. We discuss the implication of these results for theories of language production.

The empirical logic as evident in the experimental design is actually of more interest to us than the findings themselves. The “return to the point where he or she left off” is the first formulation of interest. The “point where” is a spatial metaphor, whereas the dialogue in question takes place in time. It is, then, really a question of the “time when,” not the “place where,” insofar as the post-interruption dialogue has moved forward in time. And, unless the interruption is totally trivial, irrelevant, or empty, it brings the dialogue necessarily to a new interactive moment of resumption. In short, at that moment, the post-interruption dialogical situation cannot be properly conceptualized as a regression to a previous moment in time. To consider the speaker’s goal at this moment to be exclusively the retrieval in memory of what he or she had been saying before the interruption took place is therefore a thoroughgoing misunderstanding of the nature of dialogue. The history of a dialogue is indeed relevant to continuation, but it is dialogically relevant only as part of the momentary situation to which the dialogue has now advanced in time. Swets et al. have insisted that the “return to the point where he or she left off ” is in accord with Levelt’s conceptualization of the resumption process post-interruption. But, their conclusion that “interruptions with conflicting conversational goals are particularly disruptive” simply confirms the ordinary finality of dialogue: To try to resolve the conflict so that the dialogue can advance in a coherent direction. In other words, conflicting goals move the dialogue in multiple directions at the same time and must be somehow redirected. Of course, this does not constitute genuine disruption at all, but simply points up the need for ongoing clarification of goals because the dialogue is moving forward in time. The moment-to-moment need for clarification is intrinsic to any dialogue. Or, as Lueken (1996, p. 88 f.; our translation) has put it:

My comprehension of what another person wants me to understand develops only in the course of the dialogical process, sometimes aided and abetted by the correction of misunderstandings.

But Swets et al.’s experiment conceptualizes interruption as synonymous with disruption, not as forward dialogical movement in time. Setting up an artificial situation in which “a confederate interrupted at predetermined narrative junctures” sets the speaker up as a monological agent and does indeed make the interruption trivial, irrelevant, and empty. The resumption needed in a post-interruption situation does not answer the question Where was I? or even Where am I?, but must instead always reflect Where are we now? Or to put it in another way – without the intrusion of the spatial metaphor, – the question for the researcher at any moment in a dialogue is: What is the dynamic movement at this moment in time, and how are the interlocutors to proceed from this moment on? Actually, for an interlocutor to revert to the “Where was I?” question constitutes a narcissistic withdrawal from dialogue; this would not only impede the forward movement of the dialogue in time, it could itself be highly disruptive. This is not intended as an assertion on our part that the interrupted speaker’s first psychological inclination in time might not indeed be to stay with the “Where was I?” or – later – to revert to it. The point to be made, however, is that, by perseverating in or reverting to that preoccupation, the interrupted speaker would be essentially neglecting the ongoing dialogue by failing to integrate the interruption into the ongoing discourse as a legitimate and indeed essential part thereof.

And so, the Swets et al. experiment has attempted to generalize to dialogue from a very unnatural situation in which the confederate has no role other than that of the stooge who interferes with the experimental subject, and the speaker in this “seminatural” setting may have no investment in the topic or procedure that the experimenters refer to as a dialogue. The experimental subject is not engaged in a genuine dialogue at all, but in a monologue which is to be disrupted at the behest of the experimenters. One should also note that their abstract begins with a concern about people communicating “in a dialog,” but ends with “the implication of these results for theories of language production” – a phrase that in itself carries monological overtones.

In Figure 7.1 , we have sketched the temporal relationships of the utterances of the speaker (S) and the confederate (C) and the interruption in the Swets et al. experiment. The horizontal axis is a timeline. The “speech stream” of S must be interrupted by the “speech stream” of C, in accord with the operationalization of the experiment: While S continues to speak “seminaturally,” C awaits a “narrative juncture” at which to articulate a preplanned interruption. Note that “narrative juncture” here does not imply a pause; the “speech stream” of S continues into the interruption. The abstract does not indicate that there is any necessary connection or semantic relationship between what S has said and what C says; apparently, they are both just following instructions. To label any of this as dialogue seems quite inappropriate. It is, in fact, monological insofar as the interest of the experimenters is only in what S says; the role of C is literally to affect what S says. What is abundantly clear in this setting is that the demand characteristics set up by the experimenters dictate what happens. There is certainly no spontaneous spoken discourse here insofar as a basic characteristic of spontaneous spoken discourse is absent – open-endedness (see our Chapter 21). The interruptions do not come from C’s paying attention to the substantive message of S. Note also that we have inserted a triple ellipse at the termination of the speech stream of S in Figure 7.1. This indicates that S is not able to complete her or his utterance precisely because of the interruption. Were this not to be the case, we would have a simple overlap, but no genuine interruption. The only operationalization of interruptions given by Swets et al. (2006, November 18, p. 29) in their abstract is that “the speech stream of one speaker is sometimes interrupted by the speech stream of another.” In other words, as diagrammed in Figure 7.1, C begins to speak while S is still speaking. This criterion for interruptions is actually neither necessary nor sufficient to define an interruption (for further details regarding interruptions, see our Chapter 16).

Figure 7.1 Interruption of a Speaker (S) by a Confederate (C) at a Narrative Juncture
Full size table

In summary, then, Swets et al. (p. 29) have claimed to be dealing with “a dialogue,” but have only extracted some information that seems quite irrelevant to dialogue insofar as it pertains only to an individual speaker from “a seminatural dialogue.” In a dialogue engaged under genuinely natural conditions, the question for speakers at the moment of resumption after an interruption is generally not “where they were before being interrupted,” but how the interactive dynamic has changed the post-interruption moment and how it is to be dealt with now – in real time. To ask the question solely about the pre-interruption setting is to treat the whole transaction of an interruption not as a transaction at all, but as a solipsistic behavior on the part of a speaker. An interruption is not something that goes wrong in a speaker’s own little world, not something necessarily “disruptive,” but rather one way of furthering the inevitable moving forward of a dialogue in real time. This inevitable moving forward involves typically an emotional reaction on the part of the speaker at being brought up short by the interruption; the psychological setting may be radically changed thereby for the speaker from what it had been a moment before.

Mainstream Psycholinguistics and Monologism

O’Connell and Kowal (2003, p. 191) have contended that mainstream psycholinguistics has been predominantly monologistic in its orientation ever since its inception in mid-twentieth century, i.e., that it “is concerned only with the person in whom cognition takes place and from whom communication proceeds.” This is essentially an asocial methodology or what Clark (1985, p. 179), as we have noted above, called “the individualist view of language use.”

It is hardly surprising to note that a psychology that historically specialized in laboratory experimentation and originally in introspection should view language use from the vantage point of the individual speaker. The approaches during mid-twentieth century and the advent of modern-day psycholinguistics were, after all, closely guarded by the twin powers of behaviorism and positivism. Neither had a penchant for the cultural or social and neither was modest enough to acknowledge competitors for the domain of language use. Hence, the early days of mainstream psycholinguistics were not exactly oriented toward the listener. After all, listeners could hardly be characterized as overtly and observably active in any verbally relevant way.

There had indeed been some nineteenth century signs that the dialogical was not to be entirely disregarded and neglected. But Lazarus’s (1879/1986) emphasis on the investigation of conversation went unheeded (see also Käsermann & Foppa, 2003, p. 767), and even Wundt’s (1900–1920) Völkerpsychologie was not able to arouse an interest in the socio-cultural nature of language use. Similarly, the social issues raised by World War II and the concomitant rise of clinical psychology failed to jar the psychology of language use loose from its individualist moorings. Instead, the formulation of information theory and the birth of generative grammar took central stage and won over psychologists interested in language use to a mainstream psycholinguistics which felt at home in the laboratory and comfortable with concocted and unrealistic strings of words, phrases, and sentences.

Even so, the monologistic orientation of mainstream psycholinguistics was astoundingly strong. The influence exercised by linguists such as Noam Chomsky and by psychologists such as George A. Miller was very powerful through the first decades of the new psycholinguistics. The prestige of both MIT and Harvard Universities also bolstered the new orientation to a psychology of language use. Linell (1998, p. xii f.) has summarized the spirit of those times among mainstream psycholinguists as the following theoretical orientation:

Thus the paradigm of dialogism must be understood in contrast to something else, namely ‘monologism’. The latter is the dominant theoretical framework in the language sciences. The term alludes to the tendency to identify the speaker alone as the origin of the utterance. Basically, such a framework adopts some version or other of the following theories; cognition as individually-based information processing, communication as information transfer, and language as a code.

Note that all of these preoccupations are about the dealings of the individual speaker, without any advertence to the listener or to any other dialogical or socio-cultural consideration, and they do indeed constitute a veritable ‘monologism.’

The phrase “cognition as individually-based information processing” contains most succinctly the monologistic notion. Cognition inheres within an individual; only individual human beings know, and even matters known by many individuals are not eo ipso dialogically known, i.e., known by dissemination from individual to individual. The terms cognitive and cognition have, in the course of time, taken on almost mystical importance. And cognitive psychology has become the home base of mainstream psycholinguistics. It excludes reference to dialogical and socio-cultural variables, and insofar as it engages only individual knowledge, it also excludes the domains of human intentionality, motivation, affect, intersubjectivity, and volition. By the beginning of the twenty-first century, the term cognitive had been reduced to a meaningless good-old-boy designation, a sort of Good Housekeeping or an Underwriters’ Laboratory (UL) stamp of approval – and nothing more.

The implication in a cognitive and monologistic approach to language use that communication is simply “information transfer” has been most clearly expressed by Norman and Rumelhart (1975, p. 4): “People use language to convey information.” Such a conduit theory of human communication “distorts the act of communication beyond recognition” (Ong, 1982, p. 176). It disregards the fact that understanding is a creative act: “The utterance in itself does not convey any information to the hearer: it only guides the hearer to creating the information for himself” (Hörmann, 1981, p. 308). “Information transfer” is clearly not an adequate basis for a psychology of language use.

Finally, language is more than a code; word meanings remain mere potentialities for meaning and are not automatically decoded without further ado, i.e., without the intervention of thinking and understanding human beings. The meaning of the code itself is neither self-sufficient nor automatic.

Is All Human Speech in Principle Dialogical?

In a section entitled Monological speech and thought, Linell (1998, p. 267) has taken up the obvious problem that some speech and thought appear isolated and hence monological. He has argued that dialogism “is supposed to be a theoretical framework valid for monological speech and thought as well” and that “thinking is largely arguing with other dimensions of one’s self… . The thinker is, according to the dialogistic theory, not a Cartesian ego, but a profoundly social being.” Furthermore, “monological speech is thus intrapersonally dialogical, though interpersonally it exhibits only limited dialogicality.”

In short, a sort of distancing from oneself is necessary in order to dialogue effectively within oneself, and this reflexive dialogicality incorporates the monological within its theoretical ambit. It is undoubtedly true that even our most private thinking is contextualized both epistemologically and metaphysically by an ambient reality which is eminently socio-cultural. The ontogenesis of language use must also be said to be pre-eminently dialogical; it is doubtful that a child could learn language and language use from the sole presence of TV, without the intervention of speaking adults who interact with the child. We might even consider for a moment the extreme case of a letter written to a correspondent in the privacy of one’s room. This is truly an extreme case that appears to be legitimately monological. But the ultimate rationale even of such isolated, solitary composition is not at all monological, but dialogical: We write in answer to letters and in turn request answers to our letters. The very use of the terms correspondence in English and Briefwechsel (literally letter-exchange) in German are eloquent witness to the socio-cultural embeddedness of letter writing (and receiving).

Hence, dialogism is to be embraced not as a supplement or appendage to mainstream psycholinguistics, but as a radical departure from its monologism. We would accordingly be in complete agreement with Linell (1998, p. 23) that mainstream psycholinguistics is “strongly misleading if presented as a full theory of communication through spoken interaction.”