April 22, 1998
 

ABSTRACT

We argue the case for a combination of rhetoric and pragmatics in the analysis of economic discourse. We contend that such a rhetorico-pragmatic approach is a viable alternative to both the excesses of the “rhetorical turn” and to the over-reaction of methodologists who discard as irrelevant a careful textual analysis and stick to the belief that economic knowledge progresses only through the conscientious application of Method. We examine two privileged grounds where the rhetorico-pragmatic approach does better than either Pure Rhetoric or Pure Methodology:  first, a study of economic controversies as the framework within which economic works take shape; secondly, a study of the cognitive role of metaphors in economic discourse.
 

Marcelo Dascal
Dean, The Lester and Sally Entin Faculty of Humanities
Tel-Aviv University
Tel-Aviv, ISRAEL
Fax: +972-3-6407839
E-mail: dascal@post.tau.ac.il

Sergio Cremaschi
Facoltà di Lettere e Filosofia
Università “Amedeo Avogadro”
Via G. Ferraris 109
13100 Vercelli, ITALY
Fax:  +39-161-228229
E-mail: CREM@ZEUS.VC.UNIPMN.it
 
 

0. Introduction

Most economists still have not fully recovered from the mid-Eighties' rhetorical turn: its origins are still identified with an indigestion of "constructivism", stuffed with "anti-modernism, anti-foundationalism, post-structuralism, hermeneutical deconstructivism, discourse analysis, radical relativism, end-of-philosophy critique" (Blaug 1994, p. 130). This paper has been prepared as an appetizer for the anorexic. We believe that such an antidote may be of some use, granted that so-called constructivism is neither a "scourge" (ibid.) nor the expression of "a form of outright intellectual nihilism" (Hutchison 1994, p. 30). In our view, the recent wave of reconstructions of economic discourse inspired by  the social studies of science, post-empiricism, and rhetoric, is, as a whole, quite promising. We believe that the real problem is that, for a number of reasons, economics remained largely unaffected by an awareness, growing in other fields of the social sciences, of the role of the language we use in shaping knowledge (see Backhouse 1993b, pp. 1-3). In fact, most economists still stick to an old para-positivist image of science according to which "there is simply a world 'out there' and good science represents that world accurately" (Hands 1994, p. 97), which can be done by "letting the data decide between theories - confirmation for some authors and falsification for others - but 'the facts' in either case" (ibid.). And, of the mentioned wave, only the crest, namely McCloskey's numerous books and papers, lays itself open to the criticism of flirting with the nihilistic aspects of postmodernism. McCloskey's rhetoric, as Backhouse notes, "has dominated discussion in economics to such an extent that for most economists it is probably the only approach to analyzing economic discourse" (1993b, p. 7); but this is not the only available approach: Klamer and Henderson use rhetoric in a more sober way, and other constructivists such as Mirowski, Hands, and Weintraub do not focus primarily on rhetoric.
And yet, the 'discovery' of economic rhetoric was a great discovery (albeit made by somebody who never realized what had been discovered, and followed by the usual choir of "we already knew that", with the final litany of "So, what?"); postmodernism (as all kinds of scepticism) deserves being taken seriously; and the study of the use scientists make of language is basic for science studies, even if McCloskey's approach is a tool still inadequate to the task and, accordingly, Hutchison's and Blaug's criticism may well apply once its target has been restricted to the inadequacies of this particular tool.
Our main contention in this paper is that the rhetorical turn in economics (and elsewhere) can be vindicated if one stresses the cognitive aspects typical of Aristotelian rhetoric, and combines them with the insights of a Gricean-style pragmatics. The resulting rhetorico-pragmatic framework thus combines the study of communication as a cooperative rational enterprise with  the study of means of persuasion, often used in conflictual situations such as debates, and including logically invalid moves, such as fallacious arguments. We support this contention by examining two aspects of economic discourse from a rhetorico-pragmatic perspective. First, we argue that an analysis of the polemical environment, where many (if not most) economical works belong, reveals that argumentation in such works is neither a straightforward application of logic, as if it were conducted in Popper's Platonic Third World, nor a pure battle of wits or of power, where any rhetorical trick goes. Viewed as texts produced in the context of criticism-in-action, they are subject to communicative and cognitive constraints characteristic of the dialogical practice of criticism. As such, their contribution to the “growth of knowledge” cannot be accounted for in terms of a logic of “acceptance”, “refutation” or “confirmation”, nor can it be reduced to the imponderable effects of a polite conversation. Second, we take up the role of metaphor, rhetoric's master trope, in economic discourse – a topic already treated by others. Metaphor is indeed a way of "constructing scientific facts by words" as claimed by McCloskey, and it is sometime used as a deliberate means for inducing persuasion. But the analysis of metaphors can and must go well beyond that. We try to show how, among its unintended effects, the choice of metaphors may reveal the roots of divergent conceptions of what a science of economics is or should be for its practitioners. Both, the analysis of controversies and of metaphors, illustrate and support our claim about the dividends of a rhetorico-pragmatic approach.
We begin with a diagnosis of the multiple sources and facets of the current malaise in economic methodology (section 1). We then criticize “Pure Methodology” (section 2) and “Pure Rhetoric” (section 3), and proceed to sketch our proposed via media (section 4). After an examination of an example of economic controversy (section 5) and of some of the metaphors involved in it (section 6), we conclude with a peroration in favor of our proposal (section 7).
 

1. Methodologists, rhetoricians, and cross-purposes

Since the Seventies, "interest in economic methodology has grown dramatically, to the extent that it is now possible to view economic methodology as a clearly identifiable subdiscipline within economics" (Backhouse 1994, p. 1). Before the Seventies, the discussion of methodological issues, a subject with a grand past in the classical age (and with brief revivals at the time of the Marginalist revolution and of General Theory), had not been particularly hot for some time. After 1968 there was a revival. One of the reasons, indicated by Backhouse, was "the breakdown, around 1970, of the Keynesian-neoclassical synthesis that had dominated economics since the 1950s" (p. 2); another was the shock provoked outside the natural sciences by changes introduced in the image of natural science by the post-empiricist philosophy of science (see Bernstein 1983, pp. 51 ff.), changes that had turned generally shared prejudices unquestioned by both camps in the Positivismusstreit of the Sixties between Adorno and Popper; see Hesse 1980, pp. vii-viii); one more reason was the attack against neo-classicism launched, sub specie methodologiae, by the various non-mainstream trends in economics, an attack that took advantage of  neo-classicism's internal troubles at the time.
The interest in economic methodology was revived in the Seventies first by the discovery of Kuhn and Lakatos, and only later on, around 1980, by the adoption of  Popperianism. The hard-liners Blaug and Hutchison stuck to this position, whereas several moderates, such as Boland, Caldwell, and Backhouse subsequently moved closer to an emerging mid-of-the-way trend of critics of Popperianism who defended some other more updated methodology (De Marchi and Hausman). The ensuing reaction against the anti-methodologists included the “rhetoricians" (e.g., McCloskey, Henderson, and Klamer) as well as the "constructivists" (e.g., Mirowski, Hands, and Weintraub) (see Salanti, 1989; see also Salanti and Screpanti, 1997).
Mirorwski raised the question why the alignment of the "methodologists" became so important among economists, and why it succeeded in living its own separate life (in what an economic methodologist is different from a philosopher of the social sciences?) in a waterproof shell, impervious to the deep currents that were stirring the   history, philosophy and sociology of science. He pondered whether economic methodologists were actually ignorant of the contributions of Toulmin, Hesse, Granger, Hacking, Latour, and Pickering or did they ignore them on purpose. An answer might well be that most of them studied carefully post-empiricist philosophy of science and the social studies of science literature, and decided that “all that kind of stuff” was irrelevant. After all, Mirowski concluded, "something must be done about the fact that no one seems to care about Popper any more, except a few economists" (Mirowski 1994a, p. 54). Perhaps the decisive reason for the revival of Popperianism (or of a Lakatosianized version of the same) was the justification it offered for the possibility of one Method, understood in rather crude positivistic terms, as a procedure for doling out grades to various lines of inquiry according to some universal standard of excellence. In Blaug's own words, what methodology can do
 

is to provide criteria for the acceptance and rejection of research programs, setting standards that will help us to discriminate between wheat and chaff (Blaug, 1980, p. 264).

It is precisely the refusal of Blaug's claim that unifies the, otherwise rather heterogeneous, opposition. Let us come now to the alleged standard-bearer of this united front.
McCloskey's Rhetoric of Economics (1985) aroused a turmoil. The reasons for this were probably that the book had given expression to a widespread malaise, and that the author was a respectable insider. The book finally became the first instalment of a trilogy, including If You're So Smart (1990) and Knowledge and Persuasion in Economics (1994), where the author's claims were modified in important respects while still claiming victory over critics.  In the third book, McCloskey explains how in the Eighties a few economists realized that such facts as the "character of the audience", the "style of the customary medium", the "practical purpose to be achieved from the communication" do influence scientific communication while not necessarily distorting it; he suggests that, in order to account for such factors, economists need rhetoric, which is not "an ornament, or what is left over after logic and evidence have done their work" (1994, p. 35). Rhetoric in its broad definition is the art of argument, including what is presently called logic, one corner of the rhetorical tetrad (which also includes, for McCloskey, metaphor, story, and fact).
McCloskey notes that in our century, after three centuries of eclipse, there was a revival of rhetoric, with a parallel growing awareness in various disciplines of the role language plays in shaping knowledge. He believes that some of this awareness is hindered by mixed loyalties: so, any attempt to admit a role for scientific metaphor while keeping a distinction between the artistic and the scientific uses of metaphor presumes after all a naive scientific realism, and its proponents "are adopting without realizing it a romantic literary criticism that puts the poet outside the routines of conversation" (McCloskey 1994, p. 45).
Since economists too tell stories, adopt a person, use metaphors, also Twentieth-century economic literature as a whole has, no less than Monsieur Jourdain, a style, and this style is "scientistic". Style in the writing of economics followed a path parallel to that of the modern novel: the suppression of the authorial "I" in the latter corresponds to the suppression of the "I" in science, yielding "represented reality". The implied author of economic literature, once the Philosopher or the Historian, has become the Scientist. The scientific paper depends on a theory of writing – the dominant one being based on “strange commonplaces” such as that style may be separated from content, or that "invention" (one of the parts of classical rhetoric that allegedly consists in “the framing of arguments worth listening to”) is a self-standing and all-encompassing process, supposedly independent from the metaphors and stories through which the economists’ world is created, and the style and arrangement in which their arguments are couched (see McCloskey 1994, pp. 124-5).
The apex of McCloskey's attack is his dismissal of the very idea of a Method. Methodologists understand method as something intermediate between common-sense 'methodology', i.e. the practioner's tool-box, and an "ethics of discourse"; as such it is a chimera, the dream of an apparatus or tool for producing "justified true belief", that is, for producing something that would be "an admirable ideal... if we could get it in a finite conversation about something controversial", while "effective persuasion" is all that we have (1994, p. 188). This attack ends in a rejection of the "modernist" dichotomy Subjective vs. Objective. A place of honor should be given, instead, to the Conjective (what we know together), since economics is not so much about subjective meaning and intention as it is about intersubjective patterns of action (and of belief and conversation)" (1994, p. 378).
The shock provoked by The Rhetoric of Economics was, to say the least, an over-reaction. The idea of a rhetoric of science would not have been so shocking for the average philosopher of science in the the mid-eighties. In fact, forty years ago, shortly before the crisis of the standard view of science was to blow up, two books by Perelman (1958) and Toulmin (1958) had already sketched out an idea of scientific argument wider and looser than the one described by the covering-law account, and the contributions by Hesse (1980) and Granger (1967, 1982) had acknowledged a serious role to metaphor in science. An image of science alternative to both its positivist or 'modernist' image and the total erasure of boundaries between science and "conversation" was, by then, already available; and more recent developments have sketched out the idea of a rhetoric of science (the 'hard sciences' included), understood as a 'dialectics' or a 'topics', that is, a third way between Method and conversation (see Gross 1990; Pera 1991). In these accounts of scientific rhetoric and scientific metaphor no dismissal of the cognitive dimension of scientific discourse is implied as an exit-toll out of the strictures of Method.
 

2. A critique of Pure Methodology

Following a suggestion by Weintraub we distinguish between two senses of methodology: "methodology", understood as referring to the methods practised by economists, the interaction between programs and practices, ways of referring to facts and of appealing to logical entailment; and "Methodology" as "the attempt to govern appraisal of particular economic theories by an account of theorizing in general" (Weintraub 1989, p. 478). We argue not only that Methodology is impossible, and that if it were possible it would be useless, as argued by Weintraub, but that the Methodologists’ reading of economic texts depends on an “absolutist” approach that ignores the role of context in their production and understanding (see Backhouse 1994; Screpanti 1994), and consequently commits the “semantic fallacy”. Furthermore, Methodologists continue to believe in a miraculous capacity of 'Method' to decide all issues by the application of some 'algorithm' that would rule out certain explanations and accept others on the basis of "Facts". Although qualifications of this belief abound in their writings, 'Method' is conceived as a matter of the correct application of a few rules of logic, any deviation from which amounts to opening one's door to irrationality  (see Dascal 1997).
Pure Methodology overlooks tthe insights of the new history of economic thought, postempiricist philosophy of science and the sociology of scientific knowledge (SSK), all of which demonstrate the importance of frameworks (linguistic, conceptual, textual, social, technological, etc.) that always mediate between the inquirer and the objects of inquiry, whether the latter are theories, texts, events, or processes. According to these insights, a text cannot be considered as a plain presentation of a theory, understandable per se, independently of its co-text and con-text, as the methodologists’ practice of reading indicates they believe. Their main defence of this approach has been the belief that the appeal to contextualization is supposed to replace a serious consideration of  “content”. But this is not necessarily the case, pace SSK’s claims. Context and text, pragmatics and semantics, are equally important in the construction and interpretation of meaning. Economic works contain speech acts of various kinds, not only of the purely assertive or expository kind. These acts are addressed to specific audiences in specific circumstances. The utterance of a sentence may have radically different meanings in different contexts. Its meaning can only be properly reconstructed if one takes into account its context of utterance, including the audience – e.g., the community of inquirers – it is addressed to. All this is, by now, well-known and generally accepted. What it implies is that (a) a text's understanding cannot be decontextualized by a priori fiat; (b) no part of the text can be excluded by a priori fiat as a potential contribution to its 'content'; (c) no sharp and fixed separation between the 'logic' or 'methodical' and the 'rhetorical' components of the text can be presumed. Decontextualization, that is typical of the methodologists' approach, may occassionally be useful for some purpose. But then, rather than being presumed to be "the universal norm of reading", it must be viewed as the exception, and must be carefully justified.
The methodologists’ approach is unable to make sense of the history of economic thought. Methodologists stick to a whig history of economics, according to which the past is reduced to a history of errors and partial (either continuous or discontinuous) progresses towards the true theory.  From this point of view, the history of economic thought is largely irrelevant: why should the practising economist bother about past (mistaken) theories if he can avail himself of a certified method? We believe – along with Weintraub (1989) and Samuels (1974) – that economists do not possess one certified method, in Blaug’s sense. And this is the reason why they cannot simply rush forward, forgetting the previous moves in the conversation among economists, for in this conversation – foremost in its episodes of controversy – alternative conceptions of method emerge, are confronted with each other, and evolve (see Cremaschi and Dascal 1996). Economists, thus, must be able to read economic texts from the past, which cannot be understood in isolation from the conversation where they belong, and especially from the controversies within which they function as major moves. Although Samuels called attention for the interest and importance of controversies in the history of economic thought, they are almost unfailingly overlooked by whig histories of the discipline – which ignore also other co-textual and con-textual factors.
In order to substantiate this claim, let us consider one example: the treatment of Adam Smith in Blaug’s Economic Theory in Retrospect. The reader of this book will certainly note that:

1) It examines The Wealth of Nations in isolation from Smith’s other works, unpublished manuscripts, and correspondence (see Blaug 1985, pp. 34-5);
2) It treats Smith’s voice in isolation frm that of James Steuart, David Hume, the Physiocrats, the economic pamphleteers, the early interpreter Robert Pownall (ibid.);
3) It considers The Wealth of Nations as a contribution to – and only in so far as it is a contribution to – economic science (ibid.);
4) It recognizes that Smith was the first to understand the dramatic implications of the fact that “the price system is a mechanism that imposes orderly rules of behavior on economic agents” (p. 60); and yet, it adds that his demonstration of the optimizing characteristics of a market economy “was incomplete and unsatisfactory” (ib.; italics added);
5) It ascribes to The Wealth of Nations a concern more with a “theory of economic development, if one can use such a phrase” (p. 61) than with a static of allocative efficiency in competitive markets;
6) It notes Smith’s concern with the institutional arrangements necessary to make the functioning of markets possible, notwithstanding the entrenched myth that ascribes to Smith an Harmonienlehre (p. 62), but it leaves this circumstance without any explanation;
7) It ends without being able to solve Schumpeter’s paradox: why was Adam Smith important at all if, firstly, nothing was new in his work in terms of analytic tools (p. 63) and, secondly, even his knowledge of technological improvement was far from accurate (pp. 36-7);
8) It needs to contrast “analytical concepts” and “technique”, on the one hand, with “superior economic insight” on the other (p. 63), in order to account for Smith’s worth – but what is precisely “insight”?

3. A Critique of Pure Rhetoric

We said that our approach is similar to what goes currently under the label of Rhetoric-cum-Constructivism in so far as we believe in the importance of discursive strategies as constitutive of theory formation. We have no objections, for example, to the kind of rhetorical analysis practised by Klamer and Henderson, e.g., in their treatment of metaphor (see below).  Yet,  we are critical of McCloskey’s “rhetorical turn” – both in its initial formulation and in its more recent developments – because of  the basic ambivalence that characterizes it. On the one hand, by appealing to the tradition going back to Aristotle, Cicero and Quintillian, McCloskey can claim that his rhetorical analysis is not meant to be “a verbal shell game, as in ‘empty rhetoric’ or ‘mere rhetoric’” (McCloskey 1985, p. 29). On the other hand, however, it is possible to identify another strand, which deserves the name “Pure Rhetoric”. Its main characteristic is a reductionistic tendency that purports to explain all the features of economic theorization and economic discourse as nothing but persuasive efforts – whether they use  indiscriminately the whole arsenal of rhetorical means or only a ‘moralized’ subset thereof.
The sources of McCloskey’s ambivalence lie perhaps in the peculiar mix of two basic ingredients in his original proposal -- the lessons of the new rhetoric, learned from colleagues in the English department, and a philosophical cocktail where the influence of Rorty’s relativistic neo-pragmatism looms large (See Weintraub 1992, p. 53).  McCloskey is aware that the mix is far from harmonious: though at one point (p. 27) he suggests that James’s pragmatist conception of truth and Feyerabend’s anarchist theory of knowledge may be equivalent to “rhetoric”, a few pages later he acknowledges that the “literary, epistemological, and methodological strands have not yet combined into one chord” (p. 30). The reader familiar with the philosophical literature would rightly suspect that there is more dissonance than harmony in McCloskey’s particular choice of philosophers and its attempted coupling with the new rhetoric – a dissonance that explains a good deal of the project’s ambivalence.
McCloskey (in 1985) seems to be denouncing the arrogance of scientism, or the qui-pro-quo of confounding one particular rhetoric or style with the absence of rhetoric or style tout court (pp. 69-74; see also McCloskey 1994, Ch. 9).This may be the quite plausible lesson learned from the new rhetoric movement. But he is also advancing another claim, that goes far beyond the former, without even noticing the difference between them. More than once he suggests that there is no real problem of appraisal of assertions, since “the standards of ‘good’ reasons and ‘warrantable’ beliefs and ‘plausible’ conclusions are to come … from the conversation of practitioners themselves” (p. 29). To this he adds that the “conversation” is able by itself to select good arguments by a kind of “market mechanism” – a blind and impersonal mechanism that would also succeed in imposing a kind of Socratic morality to conversation (p. 24):
 

One recognizes with ease when a conversation in one’s own field is working well …The conversations overlap enough to make one almost as sure about neighbouring fields: examinining the overlap is what editors, referees, and members of research panels do. The overlaps of overlaps … keep all honest if some try to be …. There is no need for philosophical lawmaking or methodological regulation (p. 28).

Why – one could ask – such an optimism? Couldn’t the same mechanism, unconstrained by any criteria other than conversation-driven ones, condone “irrational means such as propaganda, emotion, ad hoc hypotheses and an appeal to every kind of prejudices” (Feyerabend, 1975, p. 125) as “good reasons for warrantable beliefs and plausible conclusions”? After all, such means could well be very efficient for keeping the conversation going or for achieving persuasion. The reference to Feyerabend is McCloskey’s, not ours. And Feyerabend, as McCloskey knows, is attacking epistemology as such, not proposing to replace it by a moralized rhetoric. Nevertheless, the 1985 book repeats over and over again that the rejection of scientism and modernism and the rhetorical turn do not imply the demise of reason, or “an invitation to ‘replace careful analysis by rhetoric’, or to abandon mathematics in favor of name-calling or flowery language” (p. 36). For
 

[t]he good rhetorician loves care, precision, explicitness, and economy in argument…A rhetorical approach is not an invitation to irrationality in argument. Quite the contrary. It is an invitation to leave the irrationality of an artificially narrowed range of argument … It brings into the open the arguing that economists do anyway (ibid.).

But are the preferences of the good rhetorician – of this vir bonus dicendi peritus of a rhetorical tradition imbued with morality, as McCloskey reminds us (p. 28) – sufficient guarantee against epistemological relativism? Can a Sprachethik that makes persuasion more gentle than propaganda, but results from the self-regulating mechanism of conversation, obviate the need for “philosophical lawmaking and methodological regulation”? Even overlooking the fact that many conversations – and rather important ones – do not “work well”, that fields do not easily overlap, and that even within a given field conflicting research programs may turn out to be incommensurable, it is hard to see how McCloskey’s discourse ethics, with all its gentleness and good intentions, would be able to fill the epistemological gap left once Methodologism is rejected.
The debate triggered by the 1985 book did not leave everything as it was. In his following two books (1990 and 1994) McCloskey introduced several changes. First, a more careful treatment of metaphor and the way in which it molds the economists’ world, lending weight to the suggestion that economics consists mainly in “allegory” (McCloskey 1990, Ch. 6; see also 1985, Ch. 5). Second, a more articulate treatment of style and the way in which an implicit modernist theory of style influences the production of bad prose and unconvincing arguments by economists (1994, pp. 124-5; 34 ff.). Third, some attention is devoted to controversies. With the help of the sociologists of science, McCloskey realizes in 1994 that a monological reading of (economic) texts/theories is inadequate, because theories take shape in a dialogical context of argument, objections, and replies (1994, Ch. 8).
 These modifications no doubt render the picture presented in the 1990 and 1994 books more palatable. But they do not suppress the basic tension described above. On the ond hand, McCloskey repeats his anti-relativistic mantra: “[s]ome arguments are better than others, anything does not go. Recognizing that nonetheless they are all arguments does not entail slipping into a hot tub of ‘relativism’” (McCloskey 1994, p. 293). On the other, he continues to maintain that “effective persuasion [is] all we’ve got” (p. 188) and “justified true belief” is something impossible to reach “in a finite conversation about something controversial” (ibid.).
Thus, the weak point in McCloskey’s proposal as a whole does not lie in defending ‘mere’ or ‘empty’ rhetoric – which he does not. It lies, rather, in a conflict between a dualistic and a reductionistic trend in his thought. At one and the same time he maintains (a) that economic discourse is ruled by two independent kinds of (rational) constraints – say, the ‘rationality’ of science and the ‘reasonableness’ of rhetoric – and (b) that, ultimately, the former reduces to the latter. Although he seems to be more inclined towards the second claim, i.e., towards what we have called ‘pure rhetoric’, he does not provide convincing justification for it. Both the conflict between (a) and (b) and the lack of justification for (b) are apparent once the overall project is put into practice, as the selection of problems discussed in what follows shows.
A first vice in McCloskey's approach is overemphasis on design and intended results. In fact, a text by an economist is not only, or primarily, an exercise of  persuasion. True, great economists of the past did not apply Popper's rules of falsification; , but nor did they simply set out deliberately to “persuade” somebody. An economic work is a major move or counter-move in an ongoing dialogue, comprising letters, oral discussions, lectures, book reviews, etc.; and these moves belong to a partially cooperative and partially conflictual enterprise, regulated by certain norms.
Theoretical progress, a sore point since the crisis of the standard view, is the ground where overemphasis on intended results reveals its limits. The scientific enterprise may yield progress, and not be limited to producing incommensurable world-views, even without demanding too much from the individual scientist's rationality and morality. In fact, while the individual scientist may still try as hard as ever to win a discussion, to gain time, to impress the audience, to stick to his favourite theory, he cannot do that merely by shouting more than the opponent, and is forced to provide ever more sophisticated answers to objections. Better arguments and more precise concepts often result from criticism by emergence, not by design.
The study of controversies between economists of the past like Malthus and Ricardo shows that a controversy is indeed a deviation from an idealized and 'methodic' 'right path to truth'. It is true that economists do "make things with words" as McCloskey contends, but, by means of words, they cannot do whatever they like because their 'making' is dialogical, not monological. What they are compelled to do is at once something less and something more than winning the day by "persuading" each other by whatever means. Controversies are sequences of communicative exchanges, obeying certain pragmatic norms. They display a 'logic' which, though not reducible to formal logic, regulates the 'rational' aspects of their dynamics. When viewed in this dialectical context, theoretical change begins to make epistemological sense, for, although not reducible to a strict methodology, it is neither reducible to the intentions of scientists or to the external circumstances of their work.
A second vice in McCloskey's approach is overemphasis on the perlocutionary vs. the illocutionary, that is, "a tendency to explain too much in terms of persuasion, without going into detail on why certain arguments are more persuasive than others" (Backhouse et al. 1993, p. 8). This emphasizes the effects an author is able to achieve by manipulating his audience. But communication involves more than manipulation. In communication, one has to choose, out of a range of possibilities, certain moves that are appropriate responses to the demands of the exchange. That is, one has to give pertinent answers to some question, or claim, or counter-claim made by the partner in the communicative enterprise. And the move is to be judged against some shared standard of propriety. McCloskey was right in insisting that economics is social action, but was wrong in describing such a social action in terms of a naive idealized sociology, which overlooks the role of norms in enabling actors to perform any social action (cf. McCloskey 1985, p. 41). As Mäki notes quite à-propos, for McCloskey "rhetoric is situated in a weak social context characterized by a persuader attempting to persuade an audience by using arguments consistent with a given set of rhetorical conventions" (Mäki, 1993a, p. 45); what we need is, instead, an idea of the specific social organization of the discipline. And McCloskey, when invoking persuasion of "well-educated participants in the conversation" (1985, p. 46),  is in fact appealing to social constraints in order to qualify somehow "persuasion" as an end for economic discourse. But, in order to avoid circularity, we would need some independent way of grounding such social constraints (See Mäki, 1995, p. 1310).
A third vice consists in overlooking micro-effects. McCloskey's rhetorical analysis focuses on the effect of economic works as a whole on the the community of economists, not on changes on a small scale produced on economists' theories by their colleagues' criticism (Klamer's use of rhetoric focuses precisely on the missing links in economists' arguments). No doubt sometimes economists perform ‘mere rhetoric’ and make use of the repertoire of means highlighted by Feyerabend (and recommended, for example, by James Mill and MacCulloch to Ricardo against Malthus; see Cremaschi and Dascal 1998) or its McCloskeyan gentle version in the design of their economic treatises – which are also, after all, literary works. But what they did mainly was to face specific arguments, at specific points of an ongoing controversy (or simply conversation), as against which they had either to provide (self- and other-) convincing and reasonable replies, or else to modify their views.
McCloskey rather emphasizes the macroscopic dimension of style. He highlights devices such as  metaphor, typically studied by classical rhetoric, and insists on the idea that economic texts triumph or succumb by virtue of their literary merits to the point of believing that, for example, that one determining reason for the acceptance of Galileo's theory was the fact that the Dialoghi were a masterpiece of Italian literature (see McCloskey 1991, p. 151; cf. Bellofiore 1994b). This reveals an idea of style that embodies, itself, a rhetorical option, for it "remains thought of in classical terms, involving virtues such as purity, clarity, ornament, and decorum" (Backhouse et al. 1993, p. 8).
Finally, a fourth vice is overemphasis on  the socio-political aspect, at the macro-level of analysis. The examples discussed in the three books are basically case-studies of the way in which one approach was (globally) able to impose itself. In these examples, disparate kinds of assets all seem to play a role in leading to the success of a certain approach, theory, or paradigm. But, when McCloskeey considers individual cases, what he has to say that is specific to these cases – apart from the repeated claim that success was due to rhetorical factors – does not depart from the usual (and rather trivial) kind of critique based on “the old-fashioned – but plain and simple – distinction between logical and material truth” (Salanti 1997, p. 242). Here is what McCloskey says, for example, on “the absence of empirical contentin general equilibrium theory or the impossibility of establishing the existence of exploitation in capitalistic economies by means of mere algebraic manipulation” (ibid.; cf. McCloskey 1994, Part IV):
 

The only lesson to be learned, after all, seems to be that all standards accepted are deriving from a conversation: the assertion of plain fact derives its force … from the conventions of conversation in which it takes place … A rise in the American money supply, it is said, will cause a significant amount of inflation, albeit with a long and variable lag. The “significance” here must be relative to some experience in conversation that American economic historians and economists have had (McCloskey 1985, p. 151).

So, it is some kind of “common” or “tacit” knowledge (an idea due to Oakeshott and Polanyi) that makes assertions of fact “plausible”. But this relationship holds in a somewhat holistic way, and the reader will look in vain for a more detailed analysis showing how it actually occurs in this particular case.  The general conclusion of this example is that
 

Standards for quantitative statements, then, have to be conversational. It is only because a conversation about nuclear particles or market integration has arrived at a certain point that calculations of rates of decay or of correlations of prices are to the point (1985, p. 153).

The upshot of this critique of 'pure rhetoric' is not to deny that persuasion plays an important role in economic texts – and McCloskey is to be praised for highlighting it. Our point is that it is not the only factor that shapes these texts. Once the reductionistic thrust of 'pure rhetoric' is left aside, it is possible – we believe – to perform a different kind of rhetorical analysis of economic discourse, where the relationship between the communicative context’s constraints and the content of the theories adduced is the focal point. We will illustrate two directions in which such an analysis may be developed.
 

4. A via media

We mentioned that most economists believe they have to choose between Methodology and Rhetoric. Such and alternative is a misrepresentation. There is a third way between Popper and Rorty (or the Popperian Blaug of 1980 and the Feyerabendian-Rortian McCloskey of 1985), a way leading beyond objectivism and subjectivism (Bernstein, 1983) or towards conjectivism (McCloskey 1994). Our North-West Passage is what post-empiricist philosophers of science have been scouting around since 1962, and we navigate it following some of the rhetoricians’ kayaks that travel this way.
The context of economic works, we argued, is provided primarily by economic controversies. These, far from being annoyances forcing one to deviate from the 'right path to truth',  provide instead the immediate context wherein the "meaning" of a theory or of theory change can be properly grasped (cf. Dascal 1990, 1995). Controversies are no Popperian Third-world affairs: they develop in time, and each participant has to choose reasons and objections complying both with the controversy's demand at each point and with temporal constraints (that is, before all the possible pertinent reasons in favor or against any claim have been duly weighed, or, in McCloskey's words, some time before the Final Judgment); thus, the scientist has to make choices, underdetermined by either facts or logic, and these are made on the basis of some kind of softer rationality, something similar to Aristotelian phronesis.
Besides, the rationality of the controversy's game is wider than the participants' rationality: it is embedded in its own rules, not in the players' consciousness. Ironically, controversies thus exemplify the "unintended results" principle that the Scottish grandfather of today's economists formulated (see Hands 1994, p. 97). The contenders' purposive behavior depends, of course, on their intentions; but most of the contenders' intentions are opposed (both intend to win the debate, both intend to win the partner's respect if not to convince him, both intend to increase their reputation) and the shared intentions, such as the intention of contributing to a dispassionate examination of issues, are insufficient to determine the outcome.
It is our contention that scientific controversies such as those in economics can be best studied within the framework of rhetorico-pragmatics.  Pragmatics, as it has developed so far, is the theory of the social uses of natural languages, mainly for the purpose of communication. It views communication as a form of cooperation, guided by certain norms. The application of such norms ensures, in most cases, the intelligibility of discourse, i.e., it allows for the reasonably correct interpretation of utterances and texts, provided their context is taken into account. Rhetorico-pragmatics is a study of communication as a cognitive (i.e., non manipulative) and (to a point) cooperative enterprise. It broadens the scope of pragmatics, but its domain is more restricted than that of rhetoric at large (Dascal and Gross, forthcoming). What is left out are those rhetorical devices that seek to induce in the addressee beliefs and behavior not by providing 'reasons' (however fallacious) for such beliefs and behavior, but by somehow generating them in a 'direct', causal way – as commercial ads usually do.
Persuasion of the first type, like communication, is predicated upon the individual's ability to recognize and consciously interpret communicative intentions and meanings, whereas 'persuasion' of the second type bypasses such an ability. To be sure, many cases display features of both types. For example, rhetorical persuasion through 'reasons' may involve deception concerning the speaker's persuasive goal. While in other communicative practices a presumption of 'transparency' holds, in such 'rhetorical' communication the opposite presumption is likely to hold. Still, this kind of communication differs from usual communication only in that it relies on a different set of maxims for interpreting the speaker's meaning, but they are alike in that both address primarily the interpreter's cognitive powers. In controversies, transparency seems to be even rarer, since each speaker's goals may not only be occult but also different; but also here each side relies primarily on the audience's cognitive skills of interpretation, assessment, and inference.
Rhetorico-pragmatics studies, unlike logic, not only valid arguments, but also the reasonableness of fallacious arguments, whose presence in economics has been highlighted by McCloskey. But McCloskey’s ambivalence manifests itself also in this regard, for he leaves it unclear whether such arguments should be unmasked as ‘mere tricks’ or rather recommended (see Bellofiore 1994b; Cremaschi 1996). In fact, the use of such arguments is a distinctive feature of rhetorical practices. Enthymemes involving implicit questionable premises, analogical reasoning, inductions or generalizations based on clearly insufficient samples, appeals to character, emotion, authority are legitimate rhetorical tools. Rhetorical 'proof' (pistis) is not required - even by Aristotle - to be logical proof. And yet, rhetorical proof produces persuasion in a way that cannot be entirely irrational. For, in taking the form of inferences, such proofs take upon themselves some of the rule-governed, non-arbitrary character of inferences. An ad hominem argument is in general invalid because reliance on, say, personal characteristics of the proponent of a thesis is, strictly speaking, irrele-vant to its truth. But knowing something about the reliability of its proponent, in the absence of stronger reasons to accept or reject the thesis in question, may reasonably incline one to accept one of these options. Conformity with some interpretation of 'Newtonianism' - a typical ad verecundiam argument - may be, in certain circumstances (e.g., within a discipline where Newton's achievements set up the paradigm to be emulated), an acceptable reason to support a certain position.
Rhetorico-pragmatics provides a comprehensive theoretical framework for a context-sensitive analysis of the uses of language (an other semiotic systems) in concept- and belief-formation. Its epistemological significance lies in providing the means to investigate the ‘logic’ of conceptual change and belief-formation as manifested in the actual practice of scientific discourse, i.e. of a discourse that is produced – like any other discourse – in specific historical (intellectual as well as other) circumstances and addressed to a specific audience. The understanding of such a logic, afforded by rhetorico-pragmatics,  is essential for the clarification of a fundamental component of the ‘environment’ (internal as well as external) of scientific thought.
 

5. Controversies and the "growth of knowledge"

We have argued that the history of economic thought should be primarily a history of economic controversies because, if there is such a thing as the "growth of knowledge", then it results from a dialogical/dialectical process.  The most famous controversy between economists, the Malthus-Ricardo controversy, illustrates this claim (Dascal and Cremaschi forthcoming; Cremaschi and Dascal 1996; 1998; forthcoming). Other studies of economic controversies such as the controversy between Friedman and his critics (Backhouse, 1993; Dudley-Evans, 1993; Bloor and Bloor, 1993) and the controversy on the Phillips curve (Collins (1991) could also illustrate this point.
Methodology and the absolutist approach tend to leave controversies at the margin, concentrating on works as such: works - they assume - are presentations of theories; they speak by themselves, so, why bother with frameworks? This is precisely what Blaug’s Economic Theory in Retrospect does. But works are most of the time macro-moves in some controversy – as we argued above concerning The Wealth of Nations, and as could be shown even more easily regarding Ricardo’s Principles. Ergo, an understanding of the relevant controversy is necessary for an understanding of a given work.
Rhetoric in its turn tends to look at controversies as either juxtapositions of monologues or else as mere battles of wits. In either case they are considered quite irrelevant for the cognitive content of the positions in conflict. Furthermore, by emphasizing the overall goals of persuasive strategies, rhetorical analysis usually overlooks the local and unpredictable interaction between unexpected objections and counter-objections. These do affect substantially the cognitive content of the contenders’ theories. Ergo, controversies need to be studied in such a way as to account for their cognitive role, i.e. in a way that is beyond the scope of ‘pure rhetoric’.
In order to make the text of a controversy speak, we should consider at least (a) the context; (b) the participants' intentions; (c) the real audience addressed; (d) the strategical and tactical moves and countermoves performed by the contenders. We have referred the reader to publications where these analyses have been carried out regarding the Malthus-Ricardo controversy.  A few reminders of some of the results are sufficient here.
Malthus and Ricardo first met in June 1811 and from that meeting a correspondence originated which lasted 12 years and overlapped with the writing of their major works. The various positive issues discussed tend to show up in some phase of the controversy. These were: (i) the influence of the currency upon foreign exchanges; (ii) the Corn Laws and rent; (iii) the possibility of a "general glut"; (iv) the existence of an inverse proportion between wages and profits; (v) value, its distinction from wealth, and its relation-ship to exchange value; (vi) the search for an invariable measure of value. Recurrent methodological claims and counterclaims are grafted on the discussion of the above topics: the issue of the scope and aims of political economy is an overarching theme, encapsulating the issues of mono – or multi – causality, of realistic or unrealistic assumptions, of the alleged dangers of oversimplification, and finally of the nature of scientific language.
The phases of the controversy, each verging around one dominating positive issue, may be described in terms of 'cycles', i.e., periods of comparatively intense and continuous correspondence, usually triggered by some external event.
A good example of a cycle is the discussion on the Corn Laws which took place in 1813. The occasion was a public debate between supporters of agricultural protectionism and upholders of free trade about proposals for modifications of the "Corn Laws". To this debate Malthus himself contributed his Observations on the Effects of the Corn Laws (1814). Here he declares that, even if he is convinced of the soundness of the case for free trade in general, a number of arguments can be produced both in favour and against some kind of protectionist measure, depending not only on considerations of pure efficiency but on considerations at once political, moral, and economic. To this, Ricardo objects: that moral and scientific considerations should not be mixed up; that, since any economy always comes back to a state of equilibrium, moral and political considerations are superfluous, that supply always creates its own demand; that only simple causal links as can be detected in idealized cases can be treated scientifically; that only the real economy matters and the veil of money distorts our vision; that science considers only the permanent, not the temporary; and finally – somewhat disappointingly – that "permanent states" occur after "intervals" of about twenty, not six years as  Malthus believed.
If we consider the frequency of the letters exchanged and the number of issues, both substantive and metodological, raised or silenced, the structure of a cycle turns out to have a peculiar shape. Towards the middle of the cycle, more letters are exchanged and more new topics are introduced, while towards the end the interval between letters increases and many topics are dropped.
If we examine the course of this particular cycle we note that, from the middle of the cycle, the contenders become aware of the fact that they are not settling the issue at all; they soon realize that they diverge more than they first believed, and they begin to slow down the tempo of the correspondence as well as to silence topics, reaching finally an agreement on what the basic difference between them is. What is achieved – by way of closure of the cycle -- is neither persuasion nor victory; it is understanding (Dascal and Cremaschi, forthcoming). Rudolph Carnap's distinction between explicandum and explicans may help to clarify what happens in our cycle. Since it is a dialogical/dialectical process, the cycle does not end in an explication. What happens is that, on the one hand, the opponents do reach a clarification of the explicandum; on the other, they go on arguing without giving an explication. There is indeed a movement from the factual to the methodological level and back. At the very end of the cycle they do not reach an agreement on the positive/factual level, but they realize they need to come back to methodology - in fact to the very conception of the object of study: should the economy be considered in real or in monetary terms; should one focus on the permanent or on the temporary? But this return to a few basic issues is not simply a step back, since there has been a growth of - as they put it - mutual understanding and - we would add - self-understanding.
 A more detailed or “thick” description and analysis of this and other cycles of this controversy (and of other ones as well) would be required in order to show how the rhetorico-pragmatic approach actually works. This has been done elsewhere. All we can do here is to ponder about a few questions that arise even from the rather thin summary presented above.
 If controversies do not lead to the settlement of the issues raised, is it at all “rational” to engage in them? Arent’t those economists that avoid them – and the historians that consider them irrelevant – right? The answer is complex. If a controversy is viewed as a (collective) problem-solving enterprise, the lack of a straightforward solution to the problem at hand is indeed a sign that something may be wrong with the enterprise. But what is evident in controversies such as the one described above is that the “problem at hand” is only a trigger for a much broader and deeper range of problems, disagreement over which affect both the definition of what Popper calls “the problem situation” and its acceptable solutions. Furthermore, none of the contenders has full control over the development of the controversy. They are both players and pawns in a game that has its own dynamics and rules – a game where the opponent can always surprise with an unexpected move (cf. Hands 1993, p. 97; Dascal and Cremaschi, forthcoming). Under these circumstances, it is indeed not reasonable to expect straightforward solutions to problems as the outcome of controversies. Nevertheless, from the point of view of the overall gains – i.e., if one includes the clarification of alternative frameworks for characterizing and solving specific economic problems – there can be no doubt that the contribution of controversies, even if unintended by their protagonists, is extremely valuable (hence “rational”) for the development of knowledge (or at least of the discipline).
 Still, one may insist, if the results can be so meager as (unintended) “self-clarification”, and the price to achieve them involves resorting to invalid arguments, unjustified stubbornness, recurrent complaints of misunderstanding, and all sorts of stratagems (as both Malthus and Ricardo did), then controversies seem to be the prime example of the failure of reason and the triumph of rhetorical manipulation. What is remarkable in controversies, instead, is the complex interplay between logic and rhetoric. They actually show that the praxis of criticism and the activity of arguing cannot be easily idealized into either ‘pure logic’ or ‘pure rhetoric’, and that the ‘rationality’ of these activities – whatever it may be – lies in some combination of these two ingredients yet to be determined.
The moves in a controversy do not obey pure logic and their closure is not the result of a clearcut decision by the Supreme Court of Method (understood in Blaug's terms), for the simple reason that controversies call into question the very 'laws' and 'procedures' according to which such a Court is supposed to judge. But this does not mean that Method should give way to Mob Psychology or ‘pure rhetoric’ as the proper way to evaluate these moves. In this respect, McCloskey’s emphasis on the ethical standards of the good rhetorician is worth remembering. But, as the controversy we have discussed shows, such an ethical commitment can neither be always trusted nor does it replace all the other constraints – including those of consistency, of intelligibility, of reliability, of reasonableness, and of other forms of ‘weak rationality’ – that are undoubtedly operative in this particular economic controversy. We think it would not be unreasonable to conjecture that the close study of other economic controversies, as well as of controversies in other fields, would yield similar conclusions.
 

6. The cognitive content of metaphors
 

 6.1. Root metaphors and scientific discourse

One of McCloskey's astonishing discoveries was that also economists use metaphors. A typical response has been: Yes, thank you, and then? (e.g. Solow, 1988). In fact, this is the way 'orthodox' philosophers of science deal with the topic: metaphors/models are used by scientists as by anybody else, but they are useful at most in some preliminary phase - be it the context of discovery, or the introduction of new scientific terms - after which they are substituted by literal paraphrases. Metaphors may thus be present, and even ubiquitous, in science, but they have no durable cognitive value.  McCloskey on the contrary insisted that metaphor is ubiquitous in economic discourse; that model building, the most mathemathized part of the economist's job, which gives to economics the aura of a 'hard' science, is precisely the point where economy meets poetry; that metaphor is always a "way of doing things with words", and distinguishing its poetical and its scientific use means missing its specific function.
McCloskey, we believe, is right in giving metaphor as full a citizenship in science as in poetry and he is right in stressing the creative function of metaphor in science. But he is wrong in subordinating this function to one among other means of persuasion. For metaphor has – in economics as elsewhere – an important cognitive function as well. Economists do have recourse to metaphors in order to persuade and they occasionally invent a metaphor for this purpose. For the most part, however, they rely on a quite restricted number of basic metaphors that are delivered to them in a frozen state, shaped by the history of the discipline (cf. Klamer 1994, p. 43). That is to say, economists, like normal people, have a way of looking at the world that is molded by a few root metaphors: "constitutive metaphors are us" (ibid.; cf. also Lakoff 1987; Cremaschi 1988a; Dascal 1996). When an economist adds to this stock a new metaphor, he is likely to have initiated a significant conceptual change.
McCloskey's suggestions concerning details of the way metaphor works in economics are disapponting, boiling down to his general claim concerning 'style' and 'conversation' (i.e.., economic theories meet with success for aesthetic or sociological reasons). To find viable suggestions, one should better look at Klamer's, Henderson's and Mirowski's contributions. The latter's historical epistemology has granted metaphor a role much more basic than the one McCloskey admits of, looking beyond surface self-aware metaphorical expressions, and identifying deep basic metaphors, such as the physico-moral parallel underlying the modern world-view which molded the making of economic science (see Mirowski 1989). This line of inquiry is consistent with the one we are suggesting here, in so far as it takes seriously post-empiricist treatments of scientific metaphors.
The thesis that science employs anthropomorphic metaphors was advanced at about the same time by Nietzsche and Peirce, followed by the rhetoricians Richards and Burke, and radicalized by Stephen Pepper, who advanced the claim that the whole of the history of mankind is but a continuous recombining of a handful basic metaphors. Several post-empiricist philosophers of science have repeated this thesis. The Nietzsche-Peirce thesis contends that no genre of discourse is immune from tropes, scientific discourse included, and metaphor is the master trope of science, the basic tool for organizing and exploring the unknown, and thus - against Aristotle - it is more basic to science than to poetry. After the war waged by Galileo, the Royal Society, and Newton against metaphor and rhetoric in the name of 'plain discourse' and 'facts', and after the Romantic temporary vindication of metaphor within the framework of a somewhat subaltern rebellion against the Enlightenment, metaphor was believed by both the mainstream and the counter-culture of modernity to be the mark of the fields of poetry, religion, and myth, as opposed to the field of science where precision and literality rule (see Cremaschi 1988a).
Still a few decades ago, philosophers of science, ranging from logical empiricists such as Hempel and Braithawaite to metaphysical realists such as Boyd, were prepared to condone models and analogies either as tools for a scientific heuristics or as means of introducing new scientific terminology, but with caution and only during the preliminaries to real intercourse with facts. Mary Hesse, Donald Schön, Peter Achinstein, Marx Wartofsky, Thomas Kuhn, and Gilles-Gaston Granger argued for an indispensable, and non-provisional role for metaphor in scientific theory (see Cremaschi 1987). Here, we rely on a radicalized version of this view, as formulated by George Lakoff (1987), according to which metaphors are constitutive of languages and provide the conceptual underpinning of the "life-world". Domains such as the economy, politics, religion etc. are metaphorically constituted, and constantly mapped onto each other. To study the role of metaphor in science, then, is to analyze the structure and interplay of such mappings (see Cremaschi 1988; Klamer 1994).
 

 6.2. How to read metaphors in economic texts

Let us examine a few lines from the controversy we chose to refer to, in order to illustrate the work performed by root-metaphors. The following are passages from Ricardo's Principles:
 

Gold and silver are no doubt subject to <fluctuations> (p.14)

the correct language would be to say, that corn and labour have remained <stationary>, and all other things have <risen> in value (p. 19)

a <fall> in their value and not a <rise> in the value of the things with which they are compared (Ibid.)

labour, as being the <foundation> (p. 20)

The expressions highlighted embody frozen metaphors that evoke a spatial dimension in which the economy is placed, a universe of forces like that of Newtonian physics, and the idea of value as a physical object. If we follow the story told through ch. 1 of the Principles, the value-as-a-substance metaphor may be taken as a key to the chapter. The chapter's plot then becomes the following: Value is a thing, a physical magnitude, consisting at once of Corn and of Labour, and this thing remains invariable through its transformations into various commodities (Mirowski 1989, pp. 163-171).

Consider now a few lines from Malthus's Principles:

the <laws> which regulate the <movements> of human society (p. 11)

Man... is the primary <source> of all demand (p. 83)

the separation of rents... is a <law> as invariable as the action of the <principle of gravity> (p. 123, fn. 17)

the natural restrictions upon the importation of foreign corn during the war... may have directed the capital of the country into a <channel> more advantageous than that it would  otherwise have <flowed> (p. 328)

Here we have a set of statements embodying somewhat different metaphors. Not unlike Ricardo, the framework of Newtonian physics is assumed as a model, but, unlike Ricardo, the spatial/physical representation of the economy is subordinated to a higher-level 'legal' system whose "laws" or "principles" control the direction of 'flow' in the 'channels' of the economy. These motives are combined with others, deriving from older strata of metaphorical redescription. For example, the following simile (applied to Laissez Faire),
 

[t]he ablest physicians are the most sparing in the use of medicine, and the most inclined to trust to the healing <power> of nature (p. 15),

is a variation on Adam Smith's application of the older iatro-political metaphor to the economy. Another simile is based on the earth-as-a-machine metaphor:
 

The earth has been sometimes compared to a <vast machine>... for the production of food and raw material; but, to make the resemblance more just... we should consider the soil as... a <great number of machines>...  of very different original qualities and powers (p. 135).

This... effort of every man... [l]ike the unknown <principle of animal life>... frequently restores health... in spite, not  only of the disease, but of the absurd prescriptions of the doctor (Smith, 1776, I.iii.31, p. 343).

in the political body, the natural effort which every man is continually making to better his own condition is <a principle of preservation> (Ib., IV.ix.28, p. 674).

Every individual is continually exerting himself to find out the most advantageous employment for whatever capital he can command... and he is... <led> by an invisible hand to promote an end which was no part of his intention (Ib., IV.ii.4-9, p. 456).

the <fall> of profit in them and the <rise> of it in all others immediately dispose them to alter this faulty distribution... the private interests and passions of men

naturally <lead> them to divide and distribute the stock of every society (Ib., IV.vii.c.88, p. 630)

The natural price, therefore, is, as it were, the <central> price, to which the prices of all commodities are continually <gravitating> (Ib., I.vii.15, p. 75)

commerce, instead of running in a great number of <small channels>, has been taught to run principally in one <great channell> (Ib., IV.vii.c.43, p. 604)

And money is

 
the great <wheel> of <circulation> (Ib., II.ii.14, p. 289)

 In our reading, these examples of metaphorical statements are clues to wider and deeper metaphorical redescriptions. The national economy becomes for Adam Smith a whole, with its own 'laws' (laws that are derived metaphorically from legal and moral laws), precisely in so far as it is described not in terms of individuals and of physical goods, but in terms of the metamorphoses of a 'substance', exchange-value. In this way a world emerges that is analogous to the Newtonian physical world, with its own forces, action at a distance, mass and energy. This world may be conceived as an imaginary machine: the simple machine of 'gravitation' or the more complex machine of 'circulation'; but circulation is not only an analogon of the Copernican solar system: it is also, on a number of occasions, an analogon of Harvey's human body, with blood circulation, and of the Physiocrats' 'real' economy, with its circulation of bl?é (here the mixed metaphor celebrates its apotheosis). Finally, the "imaginary machine" of wealth may somehow reflect an (unknowable) unified world order like that conceived by the Stoics, with absolute causal necessity and a perennial cyclical movement.
 For Malthus and Ricardo the universe of The Wealth of Nations functions as a starting-point, and Smith's metaphors are almost frozen metaphors. But the preference by each of them of certain nuances within this cluster of metaphors, i.e. shifts in the barycenter of the cluster, is quite revealing. Thus, the organic metaphor is constantly (albeit not exclusively) preferred by Malthus; viewed as an organism, the economy should be explained in terms of a complex interplay of  multiple “causes” (social, cultural, institutional); consequently, the generalizations in political economy are hardly  without “exceptions”. Ricardo, on the other hand, emphasizes the mechanical metaphor; he is fond of the expression “law”, to which he assigns a mathematical sense;  consequently, his generalizations to refer to “permanent” states, and admit of no “exceptions”. In general, shifts in extant metaphors inform the understanding of key terms (themselves often metaphorically borrowed from other fields). Talk of “economic laws”, for example, may imply either that these laws are inherent in economic entities – as for the Physiocrats, following the blueprint of the immanent conception of natural law – or that they are superimposed to the original motions of such entities – as for Adam Smith, following the voluntarist conception of natural law (Cremaschi 1992).
Consider now the introduction of new metaphors. These are introduced into an already crystallized metaphorical (re)description of the world with the result of modifying it. A good example is the appearance of the notion of crisis in economic thought: the word 'crisis' in a political (and then economic) sense shows up at about the time of the French Revolution. 'Crisis' derives from the juridical, medical, and then theological lexicon and starts to be used at that time in order to describe the high risk of a sudden social change or a <revolution> (an astronomical metaphor). The first lost pamphlet by Malthus was in fact entitled The Crisis. In Malthus's political economy the permanent risk of under-consumption hints precisely at economic crises, the topic Marx, and then Keynes, would later on focus upon. In Smith we already had "cycles" (the legacy of the neo-Stoical view of history and of the Florentine neoclassical political thought); but talk of 'crisis', instead of 'cycle', conveys an image of historical trends as more precarious and unstable, making room for a prudent intervention of the government, if not in the role of a "man of System" who wants to impart an artificial order to society, at least in that of a helmsman who helps in avoiding the opposite shoals of underconsumption and of too sudden rises in public expenditure.
A similar reconstruction might be sketched out for other root-metaphors showing up in economic discourse at Malthus's and Ricardo's time. One is the idea of 'growth' of food and population, central to the Essay on the Principle of Population, that brings into the picture one biological metaphor that was still absent in Smith, waiting to be explored by later evolutionary approaches in economics, from Marshall to social Darwinism. Another is the idea of Land as the Limit. The introduction, at the time of the multiple discovery of the law of rent by Malthus and others, of the notion of Land, understood as a series of increasingly less efficient machines for the production of food, is yet another metaphor added to Adam Smith's vision. It modifies the idea of equilibrium. The smaller imaginary machine (the scales) of equilibrium - itself a mixed metaphor - was inserted by Smith in a wider imaginary machine: the wheel of circulation; here the economy is represented (along an astronomical, biological, and agricultural blueprint) as a wheel in perennial motion; it is indeed an ever-expanding wheel since at each new cycle new portions of land and labor are 'attracted' into the economy’s rotating wheel (Cremaschi 1984, pp. 189-192). In Malthus, with the principle of population and rent theory, we face a feedback loop, where the addition of one more unit of population and/or food, or the tillage of one more unit of land retro-act upon the whole system. Even if the kind of analogy is basically the same as in Smith's gravitation of prices, its role is quite different: in Smith the feedback acts within the economic system, whose development is governed basically by another logic; in Malthus the feedback governs the external boundaries of the economic system, and accordingly its chances of growth or stagnation.
Carrying out such analyses in depth would go beyond the scope of the present article (for examples see Cremaschi 1984; Mirowski 1989). What we want to show here is how metaphors, far from being just aesthetic or persuasive devices, are essential components of theory building and change, and as such carry much of a theory's cognitive content.
 

6.3 Metaphor, controversy, and the growth of knowledge

It should by now be evident that our two illustrations of a rhetorico-pragmatical approach – controversy and metaphor – are intertwined. A different set of preferential metaphors by contenders in a controversy may be the reason – other things being equal – for their preferences for different kinds of theoretical explanations. Methodological divergences between the contenders – e.g., on the relative importance of “permanent states” (Ricardo) vs. “intervals” (Malthus) – often hinge on the choice of different underlying metaphors. Such a difference may be responsible for a large number of actual and alleged “misunderstandings” that plague controversies. But it is also to be credited as contributing significantly to the “growth of knowledge”, at least on two levels.
First, because conflicting metaphorical visions are alternative sources of insights to be explored, eventually leading to the refinement and/or modification of each side’s theories: one may draw further implications of the vision of the economy as governed by the analogue of physico-mathematical laws, while the other may develop the vision of society as a living organism. Second, the confrontation between different visions, as expressed in the different metaphorical options of the contenders in a controversy, makes it possible to question the precomprehension of the field, and thereby to open it for new forms of theorization. In this way, it permits the deepening and radicalization of criticism, and paves the way for radical conceptual innovation. New metaphors are often the forerunners of such an innovation. The introduction of new metaphors (such as Crisis, Growth, Land as the Outer Limit to the Expanding Wheel of the Economy) may sometimes be more than a rhetorical stratagem designed to depict the facts in some preferred way (they would hardly help to persuade an opponent that lives by the old metaphors), and also more than a heuristic device that extends the insights of the older metaphors . They may express and lead to quite radical modifications of the below-the-waterline part of the tacitly accepted  precomprehension of the subject matter and of the categories through which it might be organized – in Foucault's words, they may have to with changes in the episteme. In short, they may play a significant role in what “scientific revolutions” are all about.
 

7.  Peroratio: The Third Way

The strategy adopted in this paper is also structured by a spatial metaphor. It reflects the polarization of meta-economic discourse into the extremes of Pure Method vs. Pure Rhetoric, and seeks to delineate a Third Way, between these extremes. We argued that our alternative, based on a cognitive rhetorico-pragmatical theory, does better than both its competitors because it is able to make epistemological sense of such phenomena as the role and conduct of controversies and the import of metaphors in economic thought.
We have criticized two approaches as exemplified by two extreme positions. No doubt the positions within both camps are more nuanced, and cross-fertilization has produced a variety of breeds within and between these camps. This is a part of the gain yielded by the cooperative-conflictual endeavour epitomized by controversy. The point of employing the traditional rhetorical move of the via media was to highlight what is distinctive in our approach. Instead of denying the role of metaphor and absolutizing the role of Method, we see both as dynamic components - "moves" - in the process of formation, evaluation, and modification of economic theories. Instead of denying the role of method and absolutizing the manipulative dimension of metaphor and other rhetorical devices, we see both as contributing to a cooperative-conflictual cognitive endeavour to reach an understanding, and through it to increase knowledge.
The discovery of economic rhetoric is no less important because its author did not realize precisely what had been discovered, namely the fact that the study of the partially cooperative enterprise of (conflictual) communication is an indispensable component of a proper account of actual theoretical progress in economics. The study of the reasonable arguments that make for the acceptability of beliefs is valuable even if it cannot be codified in so precise a way as logical rules; it is no substitute for epistemology (rhetoric and realism may even coexist, as argued by Mäki) and does not imply the demise of truth and rationality; it provides us with criteria for comparing the relative value of various kinds of reasons for accepting provisionally some claim; this holds for all those cases - and they are numerous - where we do not possess unquestionable 'logical proof' in scientific matters. In such cases, the constraints of cognitive rhetoric embody what we are left with: less than Method, but still more than conversation.
 
 

References

Backhouse, Roger E. ed. 1993a. New Directions in Economic Methodology. London and New York: Routledge.
--. 1993b. 'Introduction: New directions in Economic Methodology'. In Backhouse, 1993a, pp. 1-24.
Backhouse, Roger E., Tony Dudley-Evans, and Willie Henderson. 1993. 'Exploring the Language and Rhetoric of Economics'. In Henderson et al. 1993, pp. 1-22.
Bellanca, Nicol? and Marco E.L. Guidi. Forthcoming. 'Dove va la storiografia delle idee economiche? Una descrizione sistematica e alcune sue implicazioni'.
Bellofiore, Riccardo. 1994a. 'History of Economic Thought as a Problem'. In  Salanti, 1994, pp. 124-146.
--. 1994b. 'The Poverty of Rhetoric: Keynes vs McCloskey'. In A. Marzola and F. Silva eds., Keynes: Language and Method. Aldershot: Elgar.
Blaug, Mark. 1980. The Methodology of Economics. Cambridge: Cambridge University Press.
--. 1985. Economic Theory in Retrospect. Cambridge: Cambridge University Press.
--. 1994. 'Why I am not a Constructivist: Confessions of an Unrepentant Popperian'. In Backhouse, 1993a, pp. 109-136.
Bloor, Meriel, and Thomas Bloor. 1993. 'How Economists Modify Propositions'. In Henderson et al. 1993, pp. 15-172.
Collins, Harry 1991, 'The Meaning of Replication and the Science of Economics'. History of Political Economy 23/1: 123-42.
Cremaschi, Sergio. 1984. Il sistema della ricchezza. Angeli: Milano.
--. 1987. 'Granger and Science as Network of Models'. Manuscrito 11/2, pp. 11-136.
--. 1988a. 'Metafore, modelli, linguaggio scientifico: il dibattito postempirista'. In V. Melchiorre ed., Simbolo e conoscenza, Milan: Vita e Pensiero, pp. 31-102.
--. 1988b. 'Remarks on Scientific Metaphors', in M.L. Dalla Chiara and M.C. Galavotti ed. Temi e prospettive della Logica e Filosofia della Scienza contemporanee. 2 vols. Bologna: CLUEB. II: 114-116.
--. 1992. 'L'illuminismo scozzese e il newtonianismo morale'. In M. Geuna and M.L. Pesante eds., Interessi, passioni, convenzioni. Milano: Angeli, pp. 41-76.
--. 1996. 'Review of D. McCloskey, Knowledge and Persuasion in Economics'. Pragmatics & Cognition 4/2, pp. 425-429.
Cremaschi, Sergio, and Marcelo Dascal, 1996. 'Malthus and Ricardo on Economic Methodology', History of Political Economy 28/3, pp. 475-511.
--. 1998. 'Persuasion and Argument in the Malthus-Ricardo Correspondence', Research in the History of Economic Thought and Methodology 16, pp. 1-63.
--. Forthcoming. 'Malthus and Ricardo: Two Styles for Economic Theory', Science in Context.
Dascal, Marcelo. 1989. "Controversies as Quasi-dialogues". In E. Weigand and F. Hundsnurscher (eds.), Dialoganalyse II. vol. 1. Tübingen, Niemeyer, pp. 147-159.
--. 1990. 'The Controversy about Ideas and the Idea of Controversy', in  F. Gil. ed., Controv?rsias cient?ficas e filos?ficas.  Lisboa: Fragmentos, pp. 61- 100.
--. 1995. "Epistemolog?a, controversias y pragm?tica". Isegor?a 12: 8-43.
--. 1996. 'The Beyond Enterprise'. In J. Stewart ed., Beyond the Symbol Model: Reflections on the Representational Nature of Language. Albany, NY: Suny Press, pp. 303-334.
--. 1997. 'Critique without Critics?', Science in Context 10, pp. 39-62.
--. 1998. 'Types of Polemics and ytpes of polemical moves'. In S. Cmejrkova, J. Hoffmannova, O. Mullerova, and J. Svetla, Dialogue Analysis VI (+ Proceedings of the 6th Conference, prague 1996), Vol. 1. Tubingen: Max Niemeyer, 15-33.
Dascal, Marcelo, and Sergio Cremaschi. Forthcoming. 'The Malthus-Ricardo Correspondence: Sequential Structure, Argumentative Patterns, and Rationality'. The Journal of Pragmatics.
Dascal, Marcelo, and Alan Gross. Forthcoming. 'The Marriage of Rhetoric and Pragmatics', Philosophy and Rhetoric.
de Marchi, Neil ed. 1992. Post-Popperian Methodology of Economics: Recovering Practice. Dordrecht: Kluwer.
Feyerabend, Paul K. 1975. Against Method. Outline of an Anarchist Theory of Knowledge. London: Verso Edition.
Fiori, Stefano. 'Order, Metaphors, and Equilibrium in Adam Smith's Thought'. History of Economic Ideas 4/1-2, pp. 175-204.
Granger, Gilles-Gaston. 1967. 'Science, philosophie, ideologies', Tijdschrift voor Filosofie 29, pp. 771-780.
--. 1982. 'Postface: Form, Logics, and Reason'. In Formal Thought and the Sciences of Man. Dordrecht: Reidel
Gross, Alan G. 1990. The Rhetoric of Science. Cambridge, MA: Harvard U.P.
Gross, Alan G. and William M. Keith eds. 1997. Rhetorical Hermeneutics. Invention and Interpretation in the Age of Science. Albany: State University of New York Press.
Henderson, Willie. 1982. 'Metaphor and Economics'. In Backhouse, 1994, pp. 343-367.
Henderson, Willie, Tony Dudley-Evans, and Roger Backhouse eds. 1993. Economics and Language. London and New York: Routledge.
Hesse, Mary B. 1980. Revolutions and Reconstructions in the Philosophy of Science. Brighton: Harvester.
Jensen, Hans E. 1976. ‘Sources and Contours of Adam Smith’s Conceptualized reality in “The Wealth of Nations”’. Review of Social Economics 34/3: 259-274.
Klamer, Arjo. 1990. 'Towards the Native's Point of Vew: the Difficulty of Changing the Conversation'. In Lavoie, 1990, pp. 19-33.
Klamer, Arjo, and Thomas C. Leonard. 1994. 'So What's an Economic Metaphor?'. In Mirowski, 1994, pp. 20-54.
Lakoff, George. 1987. Women, Fire, and Dangerous Things. What Categories Reveal about the Mind. Chicago: University of Chicago Press.
Lavoie, Don ed. 1990. Economics and Hermeneutics. London: Routledge.
Lowry, S. Todd. 1974. ‘The Archaeology of the Circulation Concept in Economic Theory’. Journal of the History of Ideas 35/3: 429-444.
Mäki, Uskali. 1992. 'Social Conditioning of Economics'. In de Marchi, 1992, pp. 65-104.
--. 1993a. 'Two Philosophies of the Rhetoric of Economics'. In Henderson, Dudley-Evans, and Backhouse, 1993, pp. 23-50.
--. 1993b. 'Social Theories of Science and the Fate of Institutionalism in Economics'. In M?ki, Gustafsson, and Knudsen, 1993, pp. 76-109.
--. 1995. 'Diagnosing McCloskey'. Journal of Economic Literature 33 (September), pp. 1300-1318.
Mäki, Uskali, Bo Gustafsson, and Christian Knudsen. 1993. Rationality, Institutions and Economic Methodology, London: Routledge.
Malthus, Thomas Robert. [1820] 1989. Principles of Political Economy. Ed. by John Pullen. Cambridge: Cambridge University Press.
McCloskey, Donald N. 1985. The Rhetoric of Economics. Brighton: Harvester.
--. 1990. If You're So Smart: The Narrative of Economic Expertise. Chicago: University of Chicago Press.
--. 1994. Knowledge and Persuasion in Economics, Cambridge University Press.
--. 1997. 'Big Rhetoric, Little Rhetoric: Gaonkar on the Rhetoric of Science'. In Gross and Keith, pp. 101-112.
McNally, David. 1988. Political Economy and the Rise of Capitalism. A Reinterpretation. Berkeley: University of California Press.
Mayr, Otto. 1986. Authority, Liberty and Automatic Machinery in Early Modern Europe, Baltimore: John Hopkins University Press.
Mayer, Thomas. 1994. 'Why is there so much Disagreement among Economists?'. Journal of Economic Methodology 1/1, pp. 1-14.
Mirowski, Philip. 1989. More Heat than Light: Economics as Social Physics, Physics as Nature's Economics. Cambridge: Cambridge University Press.
--. 1992. 'Three Vignettes on the State of Economic Rhetoric'. In De Marchi, 1992, pp. 235-260.
--. 1994a. 'What Are the Questions?'. In Backhouse, 1994, pp. 50-74.
Mirowski, Philip, ed. 1994b. Natural Images in Economic Thought. Cambridge: Cambridge University Press.
Pack, Spencer J. 1991. Capitalism as a Moral System. Aldershot: Elgar.
Pera, Marcello. 1991. Scienza e retorica. Roma: Laterza.
Perelman, Chaim. [1958] 1983. Trait? de l'argumentation. La nouvelle rhetorique. 4th ed. Bruxelles: Editions de l'Institut de Sociologie, Universit? libre de Bruxelles.
Ricardo, David. [1817] 1951. On the Principles of Political Economy and Taxation. In The Works and Correspondence. 11 vols. Ed. by P. Sraffa with the collaboration of M.H. Dobb, Cambridge: Cambridge University Press, vol. I.
Salanti, Andrea. 1989. 'Recent Work in Economic Methodology: Much Ado About What?'. Ricerche Economiche 43\1-2, pp. 21-39.
--. 1997. ‘Review of D. McCloskey Knowledge and Persuasion in Economics’. The Economic Journal 107, n. 440, pp. 241-243.
Salanti, Andrea ed. 1994. "Minisymposium: History of Economic Thought: How and Why". History of Economic Ideas 2/2, pp. 109-205.
Salanti, Andrea and Eugenio Screpanti eds. 1997. Pluralism in Economics: Theory, History, Methodology. Cheltenham: Elgar.
Samuels, Warren J. 1974. 'The history of economic thought as intellectual history'. History of Political Economy 6/3, pp. 305-323.
Screpanti, Eugenio. 1994. 'Epistemic Relativism, the Post-Modern Turn in Philosophy, and the History of Economic Thought'. In Salanti, 1994, pp. 173-205.
Smith, Adam. [1776] 1977, An Inquiry into the Nature and Causes of the Wealth of Nations, ed. by R.H. Campbell, A.S. Skinner, and W.B. Todd, Oxford: Clarendon Press.
Solow, Robert M. 1988. 'Comments from Inside Economics'. In Arjo Klamer, D. McCloskey, and Robert M. Solow, eds. The Consequences of Economic Rhetoric. Cambridge University Press, pp. 31-37.
Toulmin, Stephen. 1958. The Uses of Argument. Cambridge: Cambridge University Press.
Weintraub, E. Roy. 1989. 'Methodology doesn't Matter, but the History of Thought Might'. Scandinavian Journal of Economics 91/2, pp. 477-93.
--. 1992. ‘Roger Backhourse’s Straw Herring’. Methodus 4(2): 55-7.
Worland, Stephen T. 1976. ‘Mechanistic Analogy and Smith on Exchange’. Review of Social Economy 34/3: 245-258.
 
 

Notes

  1. After 1985, McCloskey acknowledges the influence of Klamer, but he fails perhaps to realize how deep also the influence of his critics had been (see Cremaschi, 1996).
  2. For discussion of the state of the art in the history of economic thought we refer the reader to Samuels, 1974; Weintraub, 1989; Bellofiore, 1994a; Bellanca and Guidi, forthcoming. On what the social studies of science approach can contribute to the “new” history of economic thought, see Samuels, 1974; Cremaschi and Dascal, 1998; Maki, 1992, 1993b.
  3. In Samuels’ words, the historian of economic thought “must be interested in controversies, not so much because the history of the discipline is the history of controversy but because very often the controversies have been continuing and or replicated” and “the meaning of ideas and theories resides in no small degree in the intellectual matrix of the conflicts themselves” (Samuels 1974,  p. 314).
 4.  For example, among the (various  and non-decisive) reasons for Smith’s rejection of certain arguments, their morally or politically undesirale consequences loom large. Or, it is possible to explain his stressing one or another aspect of the economy in terms of the “pre-comprehension” of the economy and its place in society with which he had started his inquiry.
  5. The second ingredient seems to have been added at a stage when the basic recipe was already half-baked (See Bellofiore 1994b). In The Rhetoric of Economics, McCloskey acknowledges his philosophical debt to Paul Feyerabend, Michael Polanyi, Stephen Toulmin, Richard Rorty, and Thomas Kuhn (McCloskey 1985, p. xv). These, with the addition of James and Oakeshot (p. 27), are the authors explicitly included in the section of McCloskey’s philosophical library relevant to this book.
 6.  It is not surprising that the impact of McCloskey's rhetorical analyses on economists' actual practices is that of leaving everything as it was; that is, of ruling out criticism by a Kotuzow-like waste-land strategy (see Cremaschi 1996). Blaug’s suggestion that constructivism is an expression of "a drift towards the uncritical acceptance of the whole of modern economics, warts and all" (Blaug 1994, p. 130) may be thus endorsed once the target is restricted to McCloskey. From an opposite standpoint, this is also Bellofiore's (1994) diagnosis.
  7. As the controversy between Malthus and Ricardo appeared to McCulloch, the watch-dog of 'Ricardianism', or as disagreement among economists still appears to most methodologists (see Meyer 1992).
  8. For some results of this approach, see Cremaschi and Dascal (1996, 1998, and forthcoming) and Dascal and Cremaschi (forthcoming).
  9. For a discussion of 'progress' in science and the role of controversies therein, see Dascal (1995).
  10. We have been using so far the term 'controversy' to refer to all sorts of polemical exchanges. A typology of such exchanges can all sorts of polemical exchanges. A typology of such exchanges can be found in Dascal (1998). In this tpology the term 'controversy' denotes a specific type of polemical exchange, different from 'discussion' and 'dispute'. The Malthus-Ricardo debate is, in our view, a 'controversy' in this restricted sense.
  11. One reason for this reaction is perhaps that metaphor is most of the time unavowedly equated with vagueness, and science with precision (see Cremaschi 1988;  Klamer 1994, pp. 20-21).
  12. See Cremaschi 1984, pp. 147-48, 187-89. The same claims are repeated with a few developments in Fiori 1996, pp. 163-171.
  13. Unlike the Physiocrats, for whom economic laws were an expression of the physico-moral order of the world (Cremaschi and Dascal, 1996, pp. 499-500).
 

Acknowledgment