Robert T. Craig

In press,  Communication Theory.

Version of 10/20/00

Abstract

In a critical response to Craig (1999), Myers (this issue) develops two main points, to which I reply. First, in reply to the charge that my proposal is misrepresented and self-contradictory, I argue that it is neither. The constitutive metamodel does not pretend to avoid the reflexive paradox intrinsic to communication, and it does open a common ground for discussion about the practical implications of otherwise incommensurable theories. Second, in reply to the claim that the proposal will not eliminate the proliferation of communication theories, I argue that it is not intended to do so. However, the rejection of an absolute truth criterion for adjudicating among theories does not preclude reasoned judgments about their relative usefulness as resources for practical metadiscourse.

Minding My Metamodel, Mending Myers

In "Communication Theory as a Field" (Craig, 1999) I envisioned a forum in which to deliberate about the practical implications of communication theories. I took as a working hypothesis that all theories about communication, whatever their disciplinary origins or underlying assumptions, do have practical implications and therefore are potentially relevant to such a field. Each theory provides a specific conceptual vocabulary with which to constitute communication problems and practices as objects of reflection. Each theory draws upon certain commonsense beliefs about communication (metadiscursive commonplaces) while challenging other such beliefs. Thus, each theory can potentially be incorporated into practical metadiscourse (the ordinary metatalk that goes on continually throughout society) as a resource for articulating and challenging ideas about communication that become relevant in particular situations. Theories agree and disagree with each other according to the metadiscursive commonplaces that each relies upon and challenges. Debate within the field (theoretical metadiscourse) can focus on these points of agreement and disagreement in order to weigh the relative merits of different theories as resources for practical metadiscourse. To illustrate the approach and provide a starting point for further discussion, I showed how seven major traditions of communication theory across the field can be arrayed in a matrix that highlights practical implications of each tradition and key issues among the traditions.

Implicit in the above, and further elaborated in Craig (1999), are two principles. The first, which I called the constitutive metamodel, holds that different theoretical vocabularies constitute (or model) communication differently for practical purposes. Rhetorical theory, for example, typically constitutes communication as a practical art requiring skill and good judgment, whereas phenomenology constitutes communication as an experience of dialogue requiring openness and authenticity. The two theoretical vocabularies imply significantly different practical orientations to communication, and those differences call for clarification and discussion among communication theorists. The second principle formalized this practical function by positing that communication theory is a metadiscursive practice, a theoretical discourse that arises from, reflects upon, and informs the mundane practical metadiscourse that constitutes and regulates communication in society.

The avowed purpose of my article was to get a conversation going in the field of communication theory. In that spirit I can only welcome David Myers’ critical reply (this issue) as an invitation to continue the dialogue.

Myers offers two main criticisms. First, he charges that I misrepresented the constitutive metamodel by claiming (falsely and inconsistently) that it is something other than a first-order constitutive model of communication. Second, he argues that my proposal will not prevent the continued proliferation of communication theories because it provides no truth criterion sufficient to adjudicate among competing theories. Neither criticism directly addresses what I take to be the key issue that underlies both, which has to do with our respective notions of truth and the proper role of empirical truth as a criterion for adjudicating among theories. I reply to each of the two criticisms—and the underlying issue—in the following sections.

Myers opens his critique of the constitutive metamodel with a brief summary of my argument that is generally accurate enough for present purposes. On a matter of terminology that he raises in an endnote (mss. p. 15, note 2), I agree that what I call a metamodel could just as well be described as a metatheory or a commonly understood set of assumptions for discourse in a theoretical field. Theory for me is a broad term that includes any relatively abstract, systematic discourse, whether about conceptual models, scientific explanations, normative philosophical principles, or metatheoretical reflections. Metatheory (theory about theory) is unquestionably the subject of my article. However, my term "theoretical metadiscourse" is not synonymous with metatheory. Rather, it is a term within the specific metatheory that I propose. It describes what communication theory is (or would become) according to the constitutive metamodel: a relatively abstract, systematic discourse about discourse, distinct from but richly connected to the practical metadiscourse of everyday life.

Although it is easy enough to define communication theory as theory about communication and communication metatheory as theory about communication theory, in practice it is hard to do much of either without also doing some of the other. Arguments about particular theories are always metatheoretical by definition: To say anything about a theory is to adopt a standpoint logically "meta" to it, using some vocabulary beyond the theory itself. As Myers suggests, any theory can play the role of a metatheory; you just abstract concepts or principles from the theory and use them to say something about other theories or theory in general (as I have admittedly done with the constitutive model qua metamodel). By the same token, metatheoretical arguments are inherently theoretical insofar as they pre-constitute a theoretical subject matter. To posit a metatheoretical principle like "Communication theories should describe causal generative mechanisms," or "Communication theories should have emancipatory potential," is already to make a huge theoretical claim about how communication can best be understood (as a causal process, or as a self-reflexive practice). Moreover, it is hard (and in any case would be rather pointless) to construct a sustained metatheoretical argument without using examples and thereby slipping into a first-order theoretical discourse.

Slippage between logical levels is an inherent feature (or bug) of communication, and we should not forget that theory is, among other things, communication. Any communication theorist acquainted with Bateson (1972) will recognize the essential form of the problem. Although the Theory of Logical Types forbids mixing different logical levels in the same proposition so as to avoid paradoxes like "the set of all sets that are not members of themselves is not a member of itself," practical communication necessarily does exactly that. Every message is also a metamessage about the context or frame that determines what the message means. For example, to communicate a request is necessarily also to metacommunicate about the situation and the social relationship between the requester and requestee, usually such that the request makes good sense and is socially appropriate (Searle, 1969). Any attempt to include the metamessage explicitly within the message would not eliminate the metamessage. It would merely generate another, different metamessage, perhaps implying that the social relationship is in doubt or cannot be taken for granted. ("May I borrow your car tonight? I’m asking because I assume we are still friends.")

From the standpoint of formal logic, communication is thus inherently fraught with paradox. How do we resolve these logical paradoxes in practice? Actually, we don’t. There is no way to resolve a logical paradox in practice. One can think about it or ignore it; laugh about it or change the subject. Often one is well advised to ignore the paradox, do whatever seems best in the situation, and move on.

Communication may be rife with paradox and uncertainty but that, of course, is not a valid excuse for outright fibbing. Myers accuses me of misrepresentation and inconsistency because my so-called constitutive metamodel is really just a first-order constitutive model of communication in disguise. While claiming to advocate argumentation among competing theories in the field, according to Myers, I actually tried to subsume all theories of communication under one grand constitutive theory—and worse, I explicitly denied doing so. A Trojan horse!

Well, maybe I should have been clearer about certain matters. Myers correctly points out that the constitutive metamodel implies a first-order constitutive model, even though I argued that the two must be clearly distinguished. I would add that it also implies a third-order constitutive model (a constitutive metametamodel, a model of metamodels) and so forth at still higher levels—if you really want to go there. Myers, therefore, is quite right that my reflexive turn does not ultimately avoid paradox, as I explicitly acknowledged (Craig, 1999, p. 155, note 12). What Myers has failed to see (perhaps because I failed to make the point clearly) is that the constitutive metamodel radically transforms the constitutive model by acknowledging, rather than denying (as a first-order constitutive model implicitly does) the reflexive paradox. Far from making the constitutive model into a grand theory of the field, this acknowledgement of paradox effectively deconstructs it. To wit: If the constitutive model is uniquely true, then communication exists only as constituted in communication; but communication can be constituted in communication in many different ways; hence there can be no uniquely true way of constituting communication; hence the constitutive model is not uniquely true. Faced with this bit of sophistry, we can dwell upon it (not a good idea) or else simply acknowledge it (laughter optional), do what seems best in the situation, and move on.

My proposal is to adopt the constitutive metamodel and move on. The reason is not that we are logically compelled to do so, for we are not. The constitutive metamodel is a practical option, not an empirical truth or a logical necessity, and I recommend it for merely practical reasons. As I wrote, "I favor a pragmatic interpretation" of the constitutive model "that opens up an conceptual space in which many different theoretical models of communication can interact" (Craig, 1999, pp. 126-127). The constitutive metamodel accomplishes this in principle. If we choose to adopt it, a field of communication theory becomes possible because all theories of communication, whatever their substantive claims or underlying assumptions, however contradictory or even incommensurable with each other they may be in other respects, can still be regarded as different ways of constituting communication for practical purposes. A chief advantage of this option is that it puts at our disposal the sum total of practical value derivable from every communication theory.

This argument may be subtle, and certainly it is subject to criticism from other metatheoretical standpoints, but Myers has not convinced me that it is inconsistent or misrepresented. Does the constitutive metamodel, as he claims, "rip the theoretical bowels from" other models (mss. p. 16, note 4)? (Ugh!) Does it "animate the hollow bodies" of communication theories while depriving them of their "paradigmatic souls" (mss. p. 7)? Not at all! It merely highlights an aspect of theories that other metatheoretical perspectives tend to obscure: that every theory has practical implications. Choosing to focus on practical implications, we need not deny that theories have other kinds of differences or can be evaluated from other points of view. Specifically, the constitutive metamodel does not deny that theories differ with regard to empirical testability or plausibility based on empirical evidence. I have long argued that empirical evidence, appropriately used in conjunction with other relevant criteria, ought to play a large role in the evaluation of practical theories (Craig, 1983, 1984, 1989, 1993, 1995a, 1995b, 1996).

Admittedly, to adopt the constitutive metamodel implies that we also adopt a stripped-down first-order constitutive model of communication (along with a constitutive metametamodel and so forth), but only as a practical expedient for limited purposes, not as a foundational necessity. For these limited purposes, paradigmatic commitments can be acknowledged while translating them into practical terms (for example, by asserting the practical benefits of causal prediction, while acknowledging the practical limitations of ecological validity, associated with controlled laboratory experiments). The constitutive metamodel does not impose a grand unified theory upon the field. It assumes that deep differences and inconsistencies among theories within the field will continue to exist. It merely seeks, by turning to the practical lifeworld that all of us share, common ground on which to discuss some of those differences.

Myers is probably right about one thing. Given that the constitutive metamodel is a constitutive model, shot through with social constructionism and bristling with alien vocabulary words like "theoretical metadiscourse," some communication theorists are likely to be put off by it. If you really dislike the constitutive model of communication, you may not like the stripped-down metamodel version much better. Not only does it offend objectivist sensibilities, it deflates the status of theories by treating them all as merely different ways of constituting communication for practical purposes, whereas advocates for theories often prefer to make more grandiose claims for them. Myers is right to suggest that this situation threatens to complicate, if not entirely defeat, the project of promoting dialogue among different theoretical traditions.

The field of communication theory that I envisioned can exist only if theorists are willing to address practical problems across the gaps between different traditions and paradigms. Argumentation across traditions requires an attitude we might call theoretical cosmopolitanism. (Not "compromise," mind you. I stand firmly with Myers: A pox on all compromise!) Rather, the cosmopolitanism I have in mind involves a willingness and ability to participate in more than one theoretical conversation. Discourse within a particular theoretical tradition can be conducted in the restricted code of that tradition with all its locally shared paradigmatic assumptions. Discourse across theoretical traditions must be conducted in a cosmopolitan language that allows us to engage over practical issues while acknowledging paradigmatic differences. The constitutive metamodel can provide such a cosmopolitan language, but only if we are willing and able to use it.

Myers’ second criticism is that my proposal will not eliminate the proliferation of communication theories because it provides no criterion for adjudicating among competing theories. Caricaturing my vision of the field as a "Mad Hatter’s tea party," he concludes: "Though a constitutive metamodel may well allow all to participate in this party of discourse, it seems particularly ill-suited to inform any of the participants when it is time to leave" (mss. p. 14). By way of illustration, he speculates that my proposal would not have prevented the revivals of technological determinism that have occurred among CMC (computer-mediated communication) researchers in recent decades. This argument not only misrepresents my proposal, it largely refutes itself.

The argument misrepresents my position in at least two ways. First, it wrongly implies that the constitutive metamodel was intended to stop the proliferation of communication theories. Nowhere in the article did I call for fewer theories. Quite to the contrary, I wrote that "the goal … should be … theoretical diversity" (Craig, 1999, p. 123), that "the field is logically open to new traditions" (p. 150), that "we should not hope, nor need we worry, that the work of creating communication theory will ever be completed" (p. 152). The essential problem of communication theory in my view is not the great number and diversity of theories but the absence of a coherent field.

Second, Myers wrongly implies that my proposal offers no basis for adjudicating among theories. Does "anything go" in this metamodel? In short, not at all, though I will confess (as I did in Craig, 1999, pp. 151-152, and especially p. 156, note 28) that a methodology for testing theories within this scheme has not yet been worked out in full. It has, however, been worked out in parts in a long series of essays (Craig, 1983, 1984, 1989, 1992, 1993, 1995a, 1995b, 1995c, 1996a, 1996b, 2000; Craig & Tracy, 1995). A consistent theme of these essays has been that, if we want to build really useful theories, we must consider, along with empirical criteria, normative criteria such as moral and political principles, aesthetic values, and models of excellence derived from particular, situated practices. This is not a view in which anything goes. It does not eliminate criteria for evaluating theories but rather expands them. Empirical truth claims are often quite relevant, but the essential question from a practical standpoint is that of usefulness in praxis, and usefulness in the fullest sense involves other considerations beyond those traditionally associated with empirical science (Craig & Tracy, 1995).

Although he is less than explicit about his own position, Myers apparently thinks we need an "absolute truth" criterion in order to adjudicate among theories. Quoting my disclaimer that "This field of communication theory is not a repository of absolute truth" (Craig, 1999, p.154), he responds: "Yikes" (mss. p. 3). I need hardly point out that "yikes" is not an argument. In the following paragraph he dismisses my "constructivist notions" as "trendy" (mss. p. 4), again without offering any arguments to show what is wrong with them.

Myers’ next mention of truth reflects a misunderstanding of my position. Although I agree with Myers that "the constitutive metamodel does not, strictly speaking, assert the falseness of … other models," I must dispute his claim that "it denies the common validity of true-false distinctions" (mss. p. 7). We need to distinguish between vocabularies that constitute communication in discourse, which cannot be true or false in themselves, and particular empirical claims that might be expressed within those vocabularies, which can be judged true or false on the basis of empirical evidence in ordinary reasonable ways. My disclaimer that the "field is not a repository of absolute truth" but "claims no more than to be useful" (Craig, 1999, p. 154) was not intended to deny the usefulness (indeed, the unavoidability) of reasoned judgments of truth and falseness. What it was intended to deny (but only as a goal of communication theory) is absolute truth.[1]

In order to take this argument any further, we would need to know more about Myers’ theory of truth, but he says no more about it. Instead, he digresses into a critique of technological determinism. His point, apparently, is that my constitutive metamodel, because it lacks an absolute truth criterion, gives us no way to combat theoretical errors like technological determinism. I have no reason to question Myers’ judgment that technological determinist theories are empirically weak, especially when tested by conventional social-psychological methods (as opposed to the broad-gauged historical-critical methods favored by some media theorists). The example fails to support Myers’ point, however, and for two reasons.

First, the constitutive metamodel in no way precludes the judgment that technological determinist theories are empirically weak—which would be rightly regarded as a factor limiting their usefulness, although critics might argue that other factors are more important for some purposes.

Second, the example ironically illustrates how a community of social scientific researchers, despite their shared commitment to an empirical truth criterion, can embrace theories that one might think they should have rejected out of hand based on past research. After all, the technological determinist revivals in CMC research that Myers laments have occurred largely among scholars who regarded themselves as scientific social psychologists. If the rigorous disciplinary standards of social psychology failed to prevent their errors, what standard of absolute truth would have succeeded? This is a problem for the sociology of science, and perhaps for the hope that an empirical truth criterion of any kind can protect us from error (to err being human). How it supports a critique of the constitutive metamodel is less than apparent, however. Myers implies that things will be even worse if we embrace the constitutive metamodel. I could just as well reply that things will be better if we embrace a field of communication theory: More dialogue, more criticism from more points of view, therefore less persistent error. However we choose to speculate about the future, it remains that Myers’ attempt to illustrate the danger of giving up an absolute truth criterion tends, instead, to support an opposite conclusion: Practically speaking, his illustration suggests that we would gain little by holding on to such a criterion.

In answer, then, to Myers’ two "nutshell" questions:

Re: the proposal’s internals. The constitutive metamodel differs from a first-order constitutive model (although it implies such a model) because, having acknowledged the reflexive paradox, it can only be proposed as a practical option. It can retain fundamental, paradigmatic distinctions among theories because it finds, in the practical lifeworld that all share, a common ground for discussing all theories, but not a lowest common denominator for reducing all theories.

Re: the proposal’s externals. The constitutive metamodel may not provide a mechanism for reducing misrepresentation and mistake below current levels. Neither will it close the Antarctic ozone hole or solve other problems for which it was not designed. It can, however, in principle, improve the current situation in at least two ways. While expanding the range of criteria for adjudicating among theories, it makes possible a field of communication theory that can inform the practice of communication in society.

Finally, a brief comment on the false dichotomy implied by Myers’ closing remarks. It is quite possible to get along as colleagues while engaging in argument, debate, and even the occasional academic snipe. As I wrote, "The goal should not be a state in which we have nothing to argue about, but one in which we better understand that we all have something very important to argue about" (Craig, 1999, p. 124). Within a field of communication theory, our solidarity derives from a shared conviction that communication is important and that communication theory can inform the practice of communication. So united, we are obligated to read each other carefully, interpret each other charitably, and argue vigorously over differences that matter. And as for occasional academic sniping, well, that’s only human.[2]

References

Bateson, G. (1972). Steps to an ecology of mind. New York: Ballantine Books.

Craig, R. T. (1983). Galilean rhetoric and practical theory. Communication Monographs, 50, 395-412.

Craig, R. T. (1984). Practical criticism of the art of conversation: A methodological critique. Communication Quarterly, 32, 178-187.

Craig, R. T. (1989). Communication as a practical discipline. In B. Dervin, L. Grossberg, B. J. O'Keefe, & E. Wartella (Eds.), Rethinking communication; Volume 1: Paradigm issues (pp. 97-122). Newbury Park, CA: Sage.

Craig, R. T. (1992). Practical communication theory and the pragma-dialectical approach in conversation. In F. H. van Eemeren, R. Grootendorst, J. A. Blair, & C. A. Willard (Eds.), Argumentation illuminated (pp. 51-61). Amsterdam: SICSAT.

Craig, R. T. (1993, Summer). Why are there so many communication theories? Journal of Communication, 43(3), 26-33.

Craig, R. T. (1995a). Applied communication research in a practical discipline. In K. N. Cissna (Ed.), Applied communication in the 21st century (pp. 147-155). Mahwah, NJ: Lawrence Erlbaum.

Craig, R. T. (1995b). Forward. In W. Leeds-Hurwitz (Ed.), Social approaches to communication (pp. v-ix). New York: Guilford.

Craig, R. T. (1995c). The normativity of practical theory: Two approaches. In F. H. van Eemeren, R. Grootendorst, J. A. Blair, & C. A. Willard (Eds.), Perspectives and approaches. Proceedings of the third ISSA international conference on argumentation (Volume I, pp. 173-182). Amsterdam: SicSat.

Craig, R. T. (1996a). Practical theory: A reply to Sandelands. Journal for the Theory of Social Behaviour, 26, 65-79.

Craig, R. T. (1996b). Practical-theoretical argumentation. Argumentation, 10, 461-474.

Craig, R. T. (1999). Communication theory as a field. Communication Theory, 9, 119-161.

Craig, R. T. (in press). Dewey and Gadamer on practical reflection: Toward a methodology for the practical disciplines. In D. K. Perry (Ed.), Pragmatism and communication research (pp. 131-148). Mahwah, NJ: Lawrence Erlbaum Associates.

Craig, R. T., & Tracy, K. (1995). Grounded practical theory: The case of intellectual discussion. Communication Theory, 5, 248-272.

Dance, F. E. X. (1970). The 'concept' of communication. Journal of Communication, 20, 201-210.

Dance, F. E. X. (Ed.). (1967). Human communication theory: Original essays. New York, NY: Holt, Rinehart & Winston.

Dance, F. E. X. (Ed.). (1982). Human communication theory: Comparative essays. New York, NY: Harper & Row.

Dance, F. E. X., & Larson, C. E. (1976). The functions of communication: A theoretical approach. New York, NY: Holt, Rinehart & Winston.

Searle, J. R. (1969). Speech acts: An essay in the philosophy of language. Cambridge, England: Cambridge University Press.

Endnotes

[1] Absolute truth does not seem to be a popular concept even among scientists, who generally seem quite comfortable with the idea that truth claims at the theoretical level can be assessed only relative to the empirical evidence available, not absolutely. Myers, however, wonders whether my position “borders on broader anti-foundationalism” (mss. p. 16, note 7). I would say that it does, at least in the technical sense that it is inconsistent with a strict foundationalist epistemology like Cartesian rationalism or logical positivism. I don’t know about “Rorty’s brand of all-inclusive relativism” (mss. p. 16)—or whether that is a fair description of Rorty. Rather than harp in general terms on themes like anti-foundationalism and social constructionism, I have chosen to address my arguments to a particular field of discourse, namely communication theory. This is what anti-foundationalists and social constructionists usually advocate doing but too seldom actually do.

[2] The phrase “academic sniping” (but not the practice to which it refers!) is due to Frank E. X. Dance (1970, p. 210), who played a large role in laying groundwork for a field of communication theory (see Dance, 1967, 1972; Dance & Larson, 1976).