Science as a Rhetorical Transaction: Toward a Nonjustificational

Science as a Rhetorical Transaction:Toward a Nonjustificational Conceptionof Rhetoric

Walter B. Weimer

This essay reconceptualizes rhetoric, logic, and dialectic within anonjustificational metatheoretical framework. It argues thatscience, in its various functions as a mode of inquiry into and ex-planation of phenomena, is a Thetorical transaction. The "logic" ofscientific inference, the nature of theoretical understanding andexplanation, even problems of theory choice and paradigmallegiance are intrinsically rhetorical if one abandons the traditionaljustificationist conception of rhetoric as a "second rate" art whichpales by comparison to logic and dialectic. Indeed, a non-justificational approach to scientific thought makes logic anddialectic compaTatively restrictive facets, or aspects, of rhetoricwhich have applicability to limited domains compared to the fullrange of rhetorical transactions.

Let us begin by viewing the context in which science came to becharacterized as the accumulation of truths proven by logic (ordialectic, which employs formal logic). Here, logic and dialecticwere identified as the methods of truth and proof, hence of science,in opposition to rhetoric. Based upon justificational tenets whichignore the pragmatic aspects of human conceptualization and in-quiry, this view produced a model of science so remote from actualscientific practice that it is unobtainable even in principle. If one isto understand how science is rhetorical (rather than logical-dialectical) it is necessary to see how and why the justificationalframework is inadequate and should be replaced.

Traditional Conceptions of Logic, Scientific Inference, and Ex-pianation

For over two milennia mathematical systems have representedideals of knowledge in Western thought. When the Greeks sought

Philosophy and Rhetoric. Vol, 10, No, 1, Winter 1977, Published by Thc Penn-sylvania State University Press, University Park and London,

WALTER B, WEIMER

to understand human knowledge they studied geometry, because itrepresented the highest, most perfect form of knowledge available.Compared to the beauty and perfection of eternally valid andcertain Euclidean systems, mere empirical knowledge seemed flaw-ed and degraded. Not surprisingly, science for the Greeks was firstand foremost mathematics, and the resulting conception of scien-tific knowledge that has dominated Western epistemology hasequated knowledge with mathematical derivation. That is,knowledge has been identified with proof and certainty: genuine or"valid" knowledge became proven assertion. Knowledge, proof, andtruth became defmitionally fused concepts: Knowledge is truthwhich is proven (assertion). This is the essential tenet ofjustificationism.'

But not all putative knowledge claims are genuine or valid: onlysome inferences are (or can be) proven true. How can we determinewhich knowledge claims, i.e,, inferences, are genuine knowledgejustified as proven truth? In geometry the answer is easy: specifiableprocedures exist which enable one to prove that a conclusion is atrue or certain inference from given premises. Valid derivations canbe shown to follow deductively from premises; they aredemonstrative truths. If a conclusion does not fit the pattern of validderivation, it is not proven true, which is to say from ajustificational point of view, is not genuine knowledge at all. Logicbecame a separate entity when proof procedures were taken fromgeometry and applied as syllogisms of valid reasoning. Logicb<u;ame nothing more, nothing less, than valid proof proceduresdivorced from any particular content or subject matter. Logicbecame a codified set of principles of "right reasoning" whichguaranteed the validity or truth of a conclusion; it became theorganon which certified knowledge claims as genuine truths anddismissed as erroneous conclusions formally invalid.

Thus, methodologists held that what one does in testing aputative knowledge claim, in proving it true, is to justify it. Allgenuine knowledge is by definition demonstrated to be valid. Logicis the exemplification of such demonstration: when a conclusion isjustified, it follows logically from its premises. Thus, logic ex-emplified the laws of thought (correct thinking), or, more properly,the laws of valid reasoning: iogic became the method of science in

SCIENCE AS A RHETORICAL TRANSACTION

inference, the method of science in assessment, and the method ofscience in explanation. Science became the epitome of logic, andvice versa. This is the framework in which justificationism becameenshrined as the metatheory of the scientific endeavor.^

But justificationism is an inadequate metatheory, and the con-ception of science and the logic in science that it proposes is in-correct. The logic of science is not found in either assessment, ex-planation, or inference, at least as these areas are usually construed.Consider that justificationism ignores the pragmatic context inwhich science is embedded.

Pragmatics, explanation, and the function of language. Logic is oneof many possible language systems, and all human language isembedded in pragmatics. Thus before looking at logic in relation toscience we must consider language. I maintain that explanation isdependent upon the argumentative function of language.

Language is not often analyzed from the standpoint of purepragmatics; perhaps the best functional analysis stems from KarlBuhler, over half a century ago. Bahler analyzed the com-municative function of language into three hierarchically relatedfunctions: (1) the expressive or symptomatic function, whereinlanguage serves to express the speaker's state of mind (thoughts,emotions, etc.); (2) the stimulative or signal function, in whichlanguage triggers or releases effects and meanings in the hearer;(3) the descriptive function, which is present to the extent thatlanguage describes some state of affairs. In Sprachtheorie^ Buhlerasserted that these three functions are hierarchical in that the higherfunctions cannot be present unless all lower ones are present, butnot vice versa: the lower functions do not entail the presence ofhigher ones. One may express without signaling, or express andsignal without describing. But the descriptive function of languagecannot occur without both the signaling and expressive functionsbeing present. To illustrate this consider a sound in spokenlanguage, which may be used by a speaker to express himself, and ifthere is no audience (neglecting soliloquy) that is the only functionit can serve. If the spoken sound is received by a listener it functionsas a signal or appeal, but not necessarily in the manner intended bythe speaker. The sound may also, in the descriptive function, sym-bolize some state of affairs (often independent of either speaker or

WALTER B, WEIMER

hearer). This is a pragmatic hierarchy in the sense that it explicatesfunctional aspects of the domain of intentionatity (in Brentano'ssense).

Karl Popper, one of Buhler's students, extended this analysis to afourth hierarchical function: the argumentative function. Popperholds that argumentation cannot occur unless language is beingused to describe.

An argument, for example, serves as an expression insofar as itis an outward symptom of some internal state (whetherphysical or psychological is here irrelevant) ofthe organism. Itis also a signal, since it may provoke a reply, or agreement. In-sofar as it is about something, and supports a view of somesituation or state of affairs, it is descriptive. And lastly, there isits argumentative function, its giving reasons for holding thisview, e.g., by pointing out difficulties or even inconsistencies inan alternative view,*

A child at the stage of just naming things, or a map, are cases oflanguage which is descriptive but not argumentative. Popper alsodescribed the argumentative function as "the presentation andcomparison of arguments or explanations in connection withcertain definite questions or problems,"*

The point to note is the equation of argumentation and ex-planation. Popper tells us that the explanatory function of languageis identical to the argumentative function because of "a logicalanalysis oi explanation and its relation to deduction (or argument}.''''''Equating explanation with deduction, which is to say with logic, re-quires equating argument, as the defense of a conclusion, withdeduction, i.e., with logic. These two equations guarantee the third:argument is explanation, and both are deduction.

In this abstract anatysis predicates applicable to deductive syntac-tic systems, such as truth or falsity, consistency, derivability and im-plication, logical consequence, etc., are being used as predicates todescribe a function of language. The explanatory framework, thedistinguishing feature of both refined common sense and scientificinquiry, is thereby given the conceptual predications of formallogic. Popperians staunchly defend this line of reasoning:

If we abandon logic, we lose the power of argument: forargument consists, in essence, in showing that two claims are


incompatible. And if we abandon logic we also diminish enor-mously our powers of description: for to say what something is,to describe it, is at the same time to say a great deal about whatit is not. And if we allow contradictions to be introduced, wepermit all descriptions at once. Thus one defensive move — thescuttling of logic — forces us step by step down the hierarchyof linguistic functions. We lose the power to argue, we lose thepower to describe; we are left with the powers to express and losignal.^

I agree with Popper that explanation is inherently a matter ofargument, that explanatory science is an instance of the argumen-tative function of language within the context of pragmatics. But Ithink it clear that the equation of explanation, as a pragmatic con-cept, with deduction as a syntactic concept, mars contemporaryphilosophy. Explanation in growing science (actual practice) is notdeduction as the latter is defined within formal logic. But if that isso, then the "logic" of science cannot be formal deduction anymorethan it is induction.'

Theoretical understanding and argument. To explain aphenomenon is to conjecture a tentative theory, but the theory willnot deduce the phenomenon: rather, it will argue that thephenomenon must occur given the correctness of the theory's pic-ture of reality. It will not deduce the event in question, but demandthat it occur. Only theories can explain, and ali explanation isargument within a pragmatic context. Theories are argumentsbecause they are conceptual points of view that organize obser-vations or "data" into meaningful patterns. Theories argue for aparticular pattern or way of seeing reality. In organizing a subjectmatter, they render it intelligible by representing the structure ofphenomena with which they deal. Theories function as instrumentsof understanding because they are structural representations of adomain; or rather, theories are postulations that the structure of adomain is represented by a given pattern or configuration,

Russell Hanson's Patterns of Discovery^ arnply illustrated theconceptual nature of facts in science by considering "the same"phenomenon as quite different pheSomtna when viewed as differentconceptual patterns. Thus Kepler and Tycho Brahe, althoughstimulated by the same "optical sensibilia," perceived quitedifferent "facts" as they watched the sunrise. Facts are, on such a

WALTER B, WEIMER

view, deep structurally ambiguous: one and the same surface-structure entity is seen to have two (or more) deep structuralrepresentations or meanings (factual attributions).

Conceived in this manner, it becomes obvious that talk aboutpatterns is qualitatively different from talk of, say, iines and dots.Patterns are detectable, but they cannot be drawn. They can be seen,but they are not visible (as are lines and dots, etc). Consider thedifference in our perception of Fig, 1 when it is labeled "Mexicanon a bicycle" instead of "approaching missle," These differentpatterns are not elements of the array of "optical sensibilia."

Conceptual ambiguity: Mexican on bicycle, or approaching missle?(Reprinted from The Nature and Function of Scientific Theories, edited by Robert G,Colodny, by permission ofthe University of Pittsburgh Press, © 1970 by the Univer-sity of Pittsburgh Press)

Twenty-twenty vision is no guarantee of seeing the Mexican onthe bicycle. Patterns are not elements in an epistemic con-figuration. Rather, the pattern is the configuration itself. Byanalogy, the plot of a novel is not another cluster of words; theform of a sonata is not just another cluster cf notes; thearchitectural design of a building is not merely more bricks andbeams; the aerodynamic structure of an aircraft wing — its air-foil section — is not just more ribs and skin plates; indeed, themeaning of a proposition is not another articulated term.'^

Hanson's thesis is that theories are conceptual patterns the waythat Fig. 1 illustrates perceptual patterns. Scientific theoriesrepresent the phenomena of their domains by being structuralpatterns that are isomorphic to the structural relations that obtainin the phenomena themselves.

But what is representation in scientific theories? What is therelation between theory and "representation" and "state of affairs"and "understanding?" Hanson put it this way:


States of affairs, that is, constellations of phenomena, are oftenrendered understandable and intelligible and comprehensiblebecause some objective, structural component of thosephenomena is duplicated in a corresponding structural com-ponent within some scientific theory. Scientifically under-standing phenomena x. y, and z consists in perceiving whatkinds of phenomena they are — how they relate one to theother within some larger epistemic context; how they aredependent upon, or interfere with, one another. Insights intosuch relations 'out there' are generable within our perceptionsof the structure of theories;-these theoretical structures functionvis-a-vis our linguistic references to x, y, and 2 in a wayanalogous to how the scene stands to the tree and hill 'outthere' and also to the painted patches on canvas. Thus, Isuggest that in contrast to the delineation of theories as 'ideallanguages' or 'Euclidean hypothetico-deductive structures,' theimportant function of scientific theory is to provide such struc-tural representations of phenomena that to understand how theelements in the theoretical representation 'hang together' is todiscover a way in which the facts ofthe world 'hang together,'"

Theories are conceptual gestalts that enable one to understand byrepresenting a potentially infinitely extended domain of datapoints: that is, they represent the structure inherent in actual andpotential data. States of affairs are represented by theories, and thatrepresentation very literally is understanding.

But how are we to construe "representation" such that scientifictheories can be properly said to represent their subject matters?Representation is a generic concept; it includes picturing or iconicrepresentation and other variables. Scientific theories are abstractrepresentations — and the correlation of particular theoreticalwords with objects is inherently conventional, as is the correlationbetween the symbols on a map and the terrain it represents. Thus,theories are not like pictures; instead, they are like graphs or fiowcharts.

He who can understand the conventions of (or read) a graph orchart representing structural characteristics of phenomena is in aposition to infer to unobserved cases. With the parameters at hand,the graph plot represents an infinitude of data points that need nothave been observed. Thus, theories are generative conceptualschemes; they sanction inferences to unobserved cases. Laws of

WALTER B, WEIMER

nature are not just data summaries; they are also rules of inference.To see what a theory says of a domain is to know how to infer struc-tural properties as yet unobserved in that domain.

Now we can see the relation between representation and under-standing. Theories enable us to understand a domain because theyenable us to represent, in a compact and economical manner, thestructural properties that obtain in the domain "out there," Notonly do scientific theories represent phenomena by being structuralrepresentations isomorphic to the phenomena, but insofar as theyare explanatory they simultaneously argue for the correctness oftheir way of seeing (or, more correctly, conceiving) the domain inquestion. Theories as explanations argue that their account is a trueportrayal of the nature of reality.

Theoretical explanation is thus intrinsically argumentative. Italways has modal force. Theories say that reality must of necessitybe the way their patterns represent it to be. To reject a theory as in-correct is to say that its postulated representation does not obtain,but that does not obviate the modal force inherent in the theory.The scientist's commitment to a theory is reflected in this modalaspect: in embracing a theory the scientist is saying, in effect, "Thisis the way reality has to be." And this commitment is more thanlogic, in the form of hypothetical (if-then) reasoning, can provide.

Explanation and Logic

Explanation is, as Popper emphasized, a matter of conjecturingthe nature of reality. But it is equally a matter of asserting that thestructural characteristics of reality must be those which the theorypostulates. The pragmatic context of human inquiry and concep-tualization is intrinsically modal, and it guarantees both thattheories are arguments and that the logic of science is a modal logic.If we wish to account for scientific inference in terms of logic, weshall have to abandon the traditional hypothetical, the concept ofmaterial implication, for once and for all. Science requires adifferent inferential schema, or mood, to capture its argumentativeforce.

Inference in science may be logical, but not the logic of materialimplication. Indeed, fragments from the Stoic logicians who


developed propositional logic in opposition to Aristotelian classlogic hint at a modal inference schema that seems to embody the es-sence of scientific inference: the adjunctive conditional may be theschema of theoretical, which is to say explanatory andpostulational, science.

The history of Stoic logic remains obscure, because ourknowledge of Stoic doctrines derives entirely from commentators(such as Sextus Empiricus, Diogenes Laertius, and the RomanGalen) who mentioned it primarily to refute it. The best singlesource is Stoic Logic, by Benson Mates,'^ The propositional formswe are concerned with originated with the Old Stoa, the founder ofwhich was Zeno of Citium, Zeno was interested in language andgrammar, and appears to have originated case grammar.̂ ^ Hissuccessor, Cleanthes, preserved Zeno's doctrines with apparentlyno innovation, and was succeeded in turn by Chrysippus (c.280-205B.C), a master logician who wrote voluminously (705 books, ifDiogenes' list can be trusted). Chrysippus may have been thegreatest logician of ancient times, and no doubt he is most responsi-ble for codifying and elaborating Stoic logical theory.

The logicians of Old Stoa were concerned to find the laws ofthought. They looked at logic from a pragmatic perspective andsaw that language is in an argumentative mood when it is used inexplanation. The Stoics were thus concerned with a propositionalcalculus (rather than a logic of classes, as was Aristotle) thatrepresented the argumentative form of explanatory reasoning.Logic, for the Stoics, was always modal logic, and the laws ofthought were cast in various moods or schemata. Inferentialpatterns were modal forms: the various inference schemata werejust examples of modal force in reasoning. Thus their logic differedfrom Aristotle's in two major respects: first, it was a propositionalcalculus instead of a iogic of classes; second. Stoic logic is a theoryof inference schemas, whereas Aristotelian logic is a theory oflogically true matrices.

Stoic logic becomes relevant for scientific inference when we ex-amine the types of propositional connectives Stoics allowed (sincefor every valid argument form there is a corresponding conditional)to see what alternatives were available to the hypothetical con-ditional or "material implication." Like contemporary logicians.

10 WALTER B, WEIMER

the Stoics acknowledged implication (inciuding Philonian im-plication, which is identical to modern material implication), dis-junction (in both exclusive and inclusive senses), and conjunction.But they also acknowledged what was called an inferential, or ad-junctive proposition. The adjunctive propositional form ischaracterized by "since-necessarily":

Since the first, the second necessarily and is valid oniy when boththe antecedent and consequent are true. Its force is extremelystrong, and it seems to capture the argumentative force of scientificreasoning:

Since my theory is true, the world is necessarily this way. The ad-junctive connective asserts that the consequent follows necessarilyfrom the antecedent, and that the antecedent is true.

The adjunctive proposition is a good model for science for tworeasons: first, well-known inadequacies of the hypothetical con-ditional; second, the ability ofthe adjunctive proposition to capturethe modal force of theoretical statements. Consider these in turn.The hypothetical makes a poor model for science because it can bevalid under conditions which intuitively we reject as modelingscience: for instance, it is valid when both the antecedent and conse-quent are false, and also when the antecedent is false. Further, thereis no necessity in the relationship between antecedent and conse-quent in the hypothetical, and intuitively we fee! that laws of natureare necessarily dependent upon theories. If our theories are true, thelaws of nature are necessarily what they are. In short, attempts to fitscience into hypothetical propositions lead to paradoxes and incon-sistencies. This is, in essence, the reason that most theorists give forthe contention that science is not logical: that its inference does notfit the amodal hypothetical conditional schema.

In favor of the adjunctive proposition as a model for scientific in-ference is its demand that the antecedent and consequent cor-respond; i.e., that the ccHiclusion follow with necessity and that thepremises be true. Thus it seems to capture the inexorability withwhich true theories capture relationships among events: sincetheoretical law X. then outcome Y necessarily. Note also that obser-vation of the consequent does not confirm an adjunctiveproposition. Affirming the antecedent is still a fallacy due to thenon-commutativity of the proposition. On the other hand, obscr-

SCIENCE AS A RHETORICAL TRANSACTION 11

vation of a single refuting instance, i.e., failure to obtain the conse-quent, is ground for rejecting an adjunctive proposition. Thus atleast in principle, falsifiability is retained.

It is important to note that only a theory can be the antecedentterm of an adjunctive proposition. Theories can entail that a conse-quent necessarily obtain, but neither laws (whose derivative modalforce stems from the theory in which they are embedded) nor fac-tual propositions are strong enough to exhibit this conceptualnecessity. Indeed, it is possession ofthis conceptual necessity whichseparates explanation, as the argumentative force ofa theory, fromdescription. This is why only theoretical science is explanatory, anddescriptive science is not. The adjunctive form is the logic of ex-planatory science.

Thus conceptual necessity, which requires a modal propositionalschema and entails an explanatory and argumentative framework,makes theories quite different from taws and facts. Even the truthof a theory must be construed differently from the truth of factualpropositions. Truth is relative to conceptual frameworks, and thisrelativity, in combination with the analytic character of theorems ofa theory within a given conceptual framework, results in theoriesstanding or falling en bloc. A theory stands or falls en bloc in thatwhen one of its theorems (which has the status of a synthetic, apriori statement within the conceptual scheme which is the theory)is endorsed, the remainder must aiso be, and when one of them ischailenged they are ali challenged (even though the attemptedfalsification may be directed at only one theorem). No theory cansurvive if a singie one of its analytic consequences is falsified: suchfalsification introduces a contradiction into the theoretical system,which falsifies it in toto because whatever implies something false isfalse. The only successful defense of a theory so chailenged is toargue that the assumed theorem is not in fact a consequence (validderivation) within the theory at all.

Injunction' and the communication of argumentative discourse.Traditional accounts portray the language of science as purelydescriptive. Beginning from the truism that theories describe thenature of reality, they ignore the argumentative nature of ex-planatory discourse and assume that representation, the essentialcharacteristic of science, is just description. But not only is such an

12 WALTER B, WEIMER

account deficient with regard to the argumentative force oftheoretical explanation, it cannot account for communication inscience, either among established practitioners or between appren-tice and master.

At the observational level, pure description fails to provide a suf-ficient specification of factual data. One cannot see a fact in anydescriptive sense; to see facts as such requires an interpretativeframework.To be sure, facts can be described within a conceptualframework, but how does one learn and then communicate the con-ceptual framework?

In both communication between initiates within the scientificcommunity and in the instruction of apprentice researchers, theargumentative mode of discourse requires injunction rather thandescription. Communication in science is primarily a matter of com-mands: correct description will not be attained unless an injunctionis obeyed. Both scientific articles and research training given tonovices enjoin their audiences to behave in a certain way. In thisregard science very literally is a "cookbook" endeavor — it is amatter of recipes for conceiving, perceiving, and doing, and therecipes are given as injunctions. Scientific communication becomesa set of commands that will enable the researcher to have theappropriate experience: "Do this, and you will experience the worldcorrectly!" The descriptive commentary of the scientific reportbecomes a description only within this injunctive framework, G,Spencer Brown conveys this with admirable clarity:

Natural science appears to be more dependent upon injunctionthan we are prepared to admit. The professional initiation ofthe man of science consists not so much in reading the propertextbooks, as in obeying injunctions such as 'look down thatmicroscope.' But it is not out of order for men of science, hav-ing looked down the microscope, now to describe to each other,and to discuss among themselves, what they have seen, and towrite papers and textbooks describing it. Similarly, it is not outof order for mathematicians, each having obeyed a given set ofinjunctions, to describe to each other, and to diwuss amongstthemselves, what they have seen, and to write papers and text-books describing it. But in each case, the description isdependent upon, and secondary to, the set of injunctions hav-ing been obeyed first.'*


Rhetoric in adjunctive claiming and injunctive tuition. Since scien-tific discourse is inevitably adjunctive at its theoretical base and in-junctive in its specifications, a theory of scientific discourse mustnot be based on traditional logic. To understand science as anongoing endeavor requires a theory of the rhetoric of adjunctiveargument and injunctive tuition. What is required is a theory whichexplains the rhetorical nature of the various uses of the argumen-tative mode of discourse, which is literally the essence of science.Such a theory must be a theory of rhetoric simply because thedomain of rhetoric is all of pragmatics (rather than semantics orsyntax) and the argumentative use of behavior (includinglanguage).

This conception of rhetoric emphasizes the intrinsic valuationalor ethical dimension inherent in all scientific activity. Knowing andvaluing are inseparable: "The primary and pervasive significance ofknowledge lies in its guidance of action: knowing is for the sake ofdoing. And action, obviously, is rooted in evaluation. For a beingwhich did not assign comparative values, deliberate action wouldbe pointless; and for one which did not know, it would be im-possible."'^ In the past both scientists and rhetoricians haveaccepted, seemingly as revealed truth, the idea that science andlogic are vaiue-free, in contrast to properly rhetorical concerns(where values were grudgingly admitted), Methodologists andsociologists of science have known otherwise for some time. Manyhave emphasized that the growth of knowledge and the practice ofresearch, since they are embedded in pragmatic action, entailethical and valuational considerations at (literally) every step,"Knowing, doing, and valuing can and must be reunited. Insofar asall involve the argumentative function, rhetoric is the domain inwhich that unification must occur.

The object of communication in science is always to educate theresearch communities involved. Science instructs us in how to con-ceive the universe in which we find ourselves, and it does sorhetorically since the adjunctive claims of scientific theory can onlybe given support by argument. The adjunctive reasoning scientistsengage in is inextricably linked to injunction in tuition addressed toother scientists (and students). Scientific description can oniy bedescriptive in virtue of being embedded in an explanatory

14 WALTER B, WEIMER

framework, explanations can only be explanatory in virtue of beingargumentative, and arguments can only be learned by injunction.Hence, to state a fact is to argue for the warrant ofthe theory whichnecessitates it, and to teach others that it is a fact is to enjoin themto see reality as that theory commands one to see it. Whenever thescientist communicates, even the most mundane and seemingly in-nocuous descriptions, he is persuading his audience, literally com-manding them, to adopt his point of view.

The rhetorical nature of theory choice and revolutionaryreconceptualization is easy to see (and is noted below); but whatmust be emphasized is that there is absolutely no difference betweenrevolutionary and normal science in this regard: tuition and com-munication are injunctive in both. Indeed, Spencer Brown'sremarks refer specifically to normal science tuition, and Kuhn,Polanyi, and others have discussed this aspect of scientific commu-nication at length," Thus it is sufficient to note that this aspect ofscientific tuition was known to Plato, as the interchange betweenSocrates and Meno's slave in the Meno makes clear.

Science as a Rhetorical Transaction

The above conception of the "logic" of science in conjunctionwith the argumentative and injunctive nature of scientific discoursemake explicit several reasons why the scientific endeavor is a matterof rhetoric rather than of either logic or so-called dialectic. Scienceis a rhetorical transaction in the classic Greek sense ofthe term: it isa nonjustificational, argumentative, and persuasive art form whosereasoning is adjunctive and whose tuition and communication areinjunctive. In this regard it is instructive to examine contemporaryapproaches to rhetoric, in light of classic justificationist concep-tions of rhetoric and dialectic. It will become clear that themethodologist interested in explaining scientific research has muchto learn from a nonjustificationai rhetorical theory, and that suchrhetoric is as necessary a supplement to philosophy in the under-standing of science as are the psychological sciences.

Infiuential classical doctrine held that rhetoric can at best providetrue opinion, whiie genuine knowledge, which is certain truth un-shakeable by persuasion, can be achieved only by dialectic. Plato's

SCIENCE ASA RHETORICAL TRANSACTION 15

Republic (533-34) makes this clear: "Dialectic, and dialectic alone,goes directly to the first principle and is the only science whicb doesaway with hypotheses in order to make her ground secure: the eyeof the soul , , , is by her gentle aid lifted upwards. , , , Dialec-tic , , . is the coping-stone of the sciences, and is set over them; noother science can be placed higher, , , , " Thus, rhetoric producesonly belief hy argument and persuasion, and belief can at best betrue opinion; but genuine knowledge is produced by dialectic,Socrates is a master dialectician in the maiutic technique, andPlato's dialogues are cast to reflect dialectic rather tban rhetoric.From this framework Aristotle gave rhetoric the classic definitionof a methodological technique of persuasion independent of anysubstantive content area.

By establishing an ideal of reasoning that cannot conceivably beattained, the justificationist framework forced theorists who haverecognized the futility of "dialectic" or "scientific method" toattain certainty and proof to abandon a rational conception of con-tingent disciplines. It has fostered the idea that rhetoric represents aretreat from ideal, but unfortunately unattainable, critical stan-dards. Rhetoricians have often accepted this second-rate status ofthe discipline in comparison to both logic and dialectic. Some, forexample, many renaissance humanists, concluded that rhetoricalargument is not rational. Today it is commonly held that rhetoric isnot as rational as the "sciences" which employ logic and dialectic,which are assumed to be different from rhetoric.

No such conclusions are warranted, however, for they rest uponan untenable conception of rationality and an unwarrantedseparation of dialectic and rhetoric. Consider the rationality ofrhetorical argument and the nature of criticism. The acceptedpremise is that if one cannot be critical, then argument is notrational. But what constitutes criticism? According to the mostprevalent view, criticism consists in the search for contradictionsand attempts to eliminate them. The rationality of science is thuscriticism, which is the motive force of theory change:

Criticism invariably consists in pointing out some contradic-tion; either a contradiction within the theory criticized, or acontradiction between the theory and another theory which wehave some reason to accept, or a contradiction between the

16 WALTER B, WEIMER

theory and certain facts. . . . Criticism is, in a very importantsense, the main motive force of any intellectual development.Without contradictions, without criticism, there would be norational motive for changing our theories: there would be no in-tellectual progress.'*

Popper's reason for endorsing this strong stand on the nature ofcriticism and its centrality to rationality is that "if one were toaccept contradictions then one would have to give up any kind ofscientific activity.. , . This can be shown by proving that if two con-tradictory statements are admitted, any statement whatever must beadmitted: for from a couple of contradictory statements anystatement whatever can be validly inferred,"'*

But despite the validity of this proof, this identification ofcriticism with the search for contradictions, and both with scientificrationality, is incapable of accounting for the growth of science. Noone has delighted in detailing the growth of science on inconsistent"foundations" or theories as much as Paui Feyerabend,̂ ® but Kuhnand Lakatos^' have also convincingly documented this charge.Paradigm exemplars of rational scientific progress discussed by vir-tually every historian have utilized inconsistent premises, and thuscould have "logically deduced" any conclusion whatsoever.

Rhetoricians have reached a simiiar conclusion about the role ofcriticism (defined according to Popper) in argument. "I used tothink of philosophical criticism primarily as the attempt to exposean internal inconsistency in the position criticized, forcing theholder of the position to revise or abandon it, , . , This in-terpretation assumes that consistency is the highest aim of anyonetaking a philosophical position. . . . "̂ ^

But having seen the inadequacy of identifying rationality withcriticism, and criticism with consistency, what is to be done? John-stone provides one answer: the abandonment of rationalism.

My rationalism . . . is now a thing of the past. I no longer seephilosophical argumentation as an attempt to ap(»a] to a stan-dard of consistency in order to get an interlocutor to revise orabandon his position.... It is an attempt to evoke on his part afuller consciousness of presuppositions that may have beenmerely implicit in his philosophical position, . . , Thus theabandonment of rationalism has made it possible for me toadopt a view of phiiosophical argumentation according to


which rhetoric has a naturai and proper place in the discussionsand debates of philosophers,"

Johnstone, like the majority of contemporary rhetoricians andphilosophers, takes critical rationalism to be the only conception ofrationalist identity. Like other neojustificationists, he assumes aforced choice between two exhaustive alternatives: criticalrationalism or irrationality. Faced with this choice he abandonsrationalism in favor of an existential-phenomenologicalirrationaiism, in a retreat to commitment.^*

But a more adequate conception of rationalist identity, com-prehensively critical rationalism, allows philosophical argument(and rhetoric) a place within rationalism. Within comprehensivelycritical rationalism the concept of criticism has no fixed, a prioridefinition: what constitutes criticism, or better, appropriatecriticism, can be determined only in the given case. There is nostatute law of criticism, only case law. Further, even in the givencase, a particuiar identification of criticism can never be justified,but only defended by adducing good reasons, which is to say,arguments, in its favor. That is, what constitutes criticism may besubjected to critical evaluation. It may, in a given case, be perfectlyrational either to conjecture a new theory on the "basis" of incon-sistent foundations or to manifest the commitment to a paradigmso characteristic of normal science practice.

The point with regard to criticism and consistency is that sciencecan be perfectly rational even though logically anything at ail couldhave been deduced from any set of inconsistent premises that itemploys. Scientific creativity can be completely illogical, uncon-scious, even inconsistent. So can the assessment of theories; neitherproof nor consistency need enter into acceptance of a theory. Scien-tific theories are defended by adducing good reasons for them, notby judging them consistent,̂ ^ If the inconsistency between theoryand background knowledge (and competing theories) is not suf-ficient to warrant rejection of either, then it is hard to see why eithercriticism or rationality should be identified with consistency. Bothcriticism and rationality can be reinterpreted within a non-justificational framework so that consistency is not identified witheither. An adequate explanatory theory of rhetoric ought to explorein detail how this can be accomplished in case law determination.

18 WALTER B, WEIMER

where (as many rhetoricians have noted) it is uniquely focused todeal with the specifics of a given case.

The reunification of rhetoric and dialectic. But if we abandon thejustificationist conception of knowledge as proven (or probable)assertion in favor of the nonjustificational notion of warrantedassertion, the character of dialectic changes completely. This hasbeen implicitly recognized by many philosophers:

The Greek expression He diatektike fiechne}' may betranslated '(the art of) the argumentative usage of language,', , , One at least of its ancient meanings is very close to what Ihave described above as 'scientific method,' For it is used todescribe the method of constructing explanatory theories andof the critical discussion df these theories, which includes thequestion whether they are able to account for empirical obser-vations, , , , ̂ '

With this admission dialectic and rhetoric merge into one another,equally utilizing comparable concepts of argument and criticism.Critical appraisal of a theory (which must always be the appraisalof competing theories rather than a single theory appraised againstatheoretical facts) becomes at heart a rhetorical process; it is"dialectic" only insofar as it involves competing alternatives." Ascontemporary rhetoricians have put it:

Anytime an affirmation is challenged, that action is rhetorical,. , , Whenever and wherever claims occur and those claimseither are or might be challenged, a situation exists which canbe fully explained only if rhetorical explanation is provided.

The rhetorical dimension is unavoidable in everyphilosophical argument, in every scientific discussion which isnot restricted to mere calculation but seeks to justify itselaboration or its application, and in every consideration on theprinciples of any discipline whatever , , , whenever [men] acton other men by means of a discourse or are acted upon, theyare engaged in an activity which is of interest to the rhetorician,Man thus appears to be an essentially rhetorical animal."

Much of what has been said thus far is consonant with views putforward in the last two decades by an increasing number oftheorists who have held that rhetoric is a way of knowing.Disenchanted with the sharp separation of logic and dialectic (as


the methods of science and therefore knowledge acquisition) fromrhetoric, these theorists have focused attention on the social, psy-chological, and communicationa! aspects of the acquisition ofknowledge, arguing that the human epistemic situation is inevitablyrhetorical. Thus these theorists link rhetorical theory, construed asthe theory of argument, to theory of knowledge: rhetoric becomesan activity of ideational discovery, and "in many circumstanceshumans can 'know' only through engagement in rhetorical ac-tivity,"^" The central claim in this group of theorists is thatargument (both as theoretical principles and applied practice) isfundamental to knowledge. Insofar as man is engaged in the ac-quisition of knowledge he is engaged in argumentation. Thus manas knower is intrinsically rhetorical, and this obviously in the prac-tice of science. As one article put it:

Unless one is willing to argue that that [scientific] statement in-cludes everything that could conceivably be said, and said fromevery conceivable vantage point — unless one is willing toargue that — one must grant that what has been told is a par-tial story. And the partial story told is obviously one that is ad-vocated, for it is senseless to choose a certain story about acertain object/event and to claim at the same time that someother story is preferable. Hence, by definition, the scientificstatement is rhetorical,^'

Theorists ofthis "school," as Arnold'^ has called them, would nodoubt resonate to both identification ofthe domain of rhetoric withthe argumentative function of action and discourse and to the claimthat the logic of science is adjunctive and its tuition injunctive. Butwhile I agree with many of their claims, I do not wish to be iden-tified with this school, and it would be a mistake to take this essayas another instance of its growing impact. The reason for this is thatwithout exception these theorists remain justificationists. Theirusual formulation argues against a comprehensive rationalist,classic justificationist picture of science (amodal logic, decisiveproof or refutation, absolutistic and infallible method) and en-dorses a critical rationalist, neojustificationist account of rhetoric(probabilistic inference in place of certainty, situational relativismof truth and knowledge),'* "Often, this 'school' asserts, no humanthinking can be expected to achieve certainty. When certainty is not

20 WALTER B, WEIMBR

possible, men can only act rhetorically, , , , "'^ Thus rhetoric is en-dorsed in a retreat from higher, but unfortunately unobtainable,critical and epistemic standards. Rhetoric is identified as the onlyalternative available in an uncertain world in which the apodicticproof of logic plays no role (recall Johnstone's reasoning in aban-doning "rationalism"). Even though rhetoric is given "primacy" inthese accounts, it still has second-rate status, as indeed it must havein a justificationist account, no matter how sophisticated it mayotherwise be.

The justificational underpinnings of the "new rhetoric" needthorough exposure and documentation. Although worthwhile, thistask is beyond my purpose; I note only that while this essaythoroughly endorses the interdetermination of rhetoric andepistemology (rather than the primacy of either) and the iden-tification of rhetoric with argumentation, it does so within the con-fines of a nonjustificational metatheory of rationality and science.Rhetoric can be a way of knowing only if knowledge is a matter ofwarranted assertions rather than proven or probable assertions.Further, rhetoric's primary application is in the realm of con-tingence, in so-called empirical endeavors; logic remains themethod of reasoning to conclusions in formal disciplines.

The scientific research community in the rhetorical transaction.The production of knowledge indispensably involves a producer in-teracting with an audience, if for no other reason than thatargument presupposes a plurality of points of view and therefore(neglecting arguing with one's self) a community of knowers. Thecommunity structure of science has been extensively studied of lateby sociologists of science, and the work of Merton, Hagstrom,Mulkay, Ziman, and others" has begun to sketch an informativepicture of the scientist as a social being. When combined withsocially oriented accounts from other perspectives, such as those ofPolanyi (philosophy) and Kuhn (history), the picture is con-siderably enhanced. But as yet there is little on the audience inscience. Accounts have focused upon the research communityprimarily in reference to its impact upon the individual scientist.But the scientific community as a rhetorical audience, as an activeconstructor of the scientific dialogue and the meaning which it


manifests, has received little attention from any discipline orperspective.

As one illustration of ways in which rhetorica! study of theproduction of knowledge can illuminate science in a novel and in-formative manner, consider how the research community in whichhe operates becomes a persona to whom a scientist must address hisfindings. In communicating his results the researcher is, quiteliterally, engaged in a dialogue with the persona representing hisresearch community. The initial process of formulating (to himself)what he thinks he knows, what he can argue for and against, in-volves the researcher in an internal dialogue with what he perceivesto be the point of view of his research community. Subsequent com-munication (from informal discussion with colleagues, throughcolloquium or convention presentation, to and includingpublication) requires the researcher to sharpen his presentation tobe maximally effective to the audience persona in question. Inperiods of Kuhnian normal science, much of the metaphysicaland/or sociological paradigm finds concrete expression in views theresearcher attributes to the persona he is addressing, and this per-sona typifies a community of "like-minded" scholars. Inrevolutionary periods the persona is often hostile, opaque, dis-inclined to listen to "reason," etc, and the entire nature of the com-munication process changes accordingly.

The personae of science render inteiligible another phenomenonthat historical research has disclosed to be characteristic of science:the enormous discrepancy between a scientist's preachment and hisresearch practice. When queried about their research, scientistsusuaiiy respond that it conforms to one or another methodologywhich they take to be the personification of "good science." ThusNewton, a consummate visionary and theoretician, would trtter"hypotheses nonfingo" to a community of inductivist researchers inabsolutely convincing fashion, because he really believed in it.Historians examining his research have found Newton to be muchmore speculative than he would have admitted, and one way to un-derstand the discrepancy is to consider Newton both as a personaand aiso as addressing one. The ideal and the real interact in theconstruction of thc products of science, and a rhetorical theory of

22 WALTER B. WEIMER

personae could be enormously informative of how that interactionoccurs.

Note that the persona of ideal science, as it appears injustificationist accounts, has been a severe drawback both to thedevelopment of rhetorical theory and to the rhetorician desirous ofstudying science. Virtually to a man, rhetorical theorists whoemphasize rhetoric as a way of knowing have been misled in theirotherwise informative accounts by the justificationist persona of"objective" science coldly marching toward proven truth inamodal, value-free, and "logical" fashion,̂ * Were these theoristsmore inclined to look at science and its history rather than at thepersonification of science in justificationist philosophy and text-book history, there would be no need to apologize for rhetoric, orto assign it second-rate status compared to "dialectical" science.There is no need for sentiment such as this: "It was said [atWingspreadj that when subjects, data, and their meanings can beinterpreted as asserted to simply by distinguishing facts from non-facts the 'method of rhetoric' is inapplicable, but issues on whichmeaning cannot be determined by fact-nonfact judgments areprecisely those on which decisions can be reached only throughrhetorical considerations of the choices available to men,"^^

Another neglected aspect of the research community as aproducer of knowledge in virtue of its participation in the rhetoricaldialogue of ongoing research sheds Hght on the time-boundedcharacter of "truth" in science. Research communities, in theircapacity as auditors of scientific formulations, change their viewsover time. Often the change is not straightforwardly related to theaccumulation of new "data," or even to significant theoretical ad-vance. After a period in which formulations are guided by one oranother cluster of directives, situations that otherwise appear iden-tical will be treated entirely differently. An example (borrowed,with acknowledgment, from Carroll C, Arnold) from the law mayhelp. It sometimes happens that legal precedent A, after havingbeen the basis of judges' decisions in cases for considerable time(perhaps a generation or more), will suddenly give way to precedentB in similar cases. The switch from A to B comes about not becauseof new "facts," but because some feature of precedent B was ex-plored and seen to be more applicable to these cases than the prior

SCIENCE ASA RHETORICAL TRANSACTION 23

precedent A. The point to emphasize is that the criteria ofevaluation were rhetorically created.

The ascendancy of precedent B occurred as a result of rhetoricalinterchange between judges, juries, and attorneys. Precedent A wasabandoned not because it was seen to be inapplicable (disconfirmedor falsified) — indeed it need not have been tested at all — butbecause it was seen as less applicable, plausible, relevant, etc, thanB.

Scientific ideas have a similar time-boundedness, and it is easy tocite examples of switches in theories due to rhetorical factors suchas those in the hypothetical legal example. One example is thevarious metaphysical directives at the heart of scientific researchprograms which J, W, N, Watkins called "haunted universedoctrines."^* For example, the Cartesian directive "the universe is aclockwork mechanism" and the ether doctrine that "all matter iscontinuous" were replaced by Newtonian directives such as "theuniverse is matter in motion" and "all matter is compounded ofatoms," not because they were disconfirmed (because, asmetaphysical statements, they are confirmable but not refutable),but because Newtonian "theory" came through rhetorical in-teraction to seem a more adequate description of reality. The scien-tific audience, in cases such as this, is as much a factor in shapingthe character of knowledge as is the theorist who initiallypropounds an idea. The scientific "auditor" is not an impartialbookkeeper checking the arithmetic of the scientific record book(the CPA model of logical empiricism and most rhetorical ac-counts); the "auditor" is rhetorically engaged, an active constructorof the products that enter the record book of the scientificendeavor.

Incomplete though they are, examples such as these indicate theextent to which science, as a rhetorical transaction, is interplaybetween theorist or researcher, on one hand, and the research com-munity as audience, on the other.

From the perspective used in this essay, the vexing problem oftheory choice in science takes on a new dimension. Most debate has-concerned the adequacy of philosophical versus socio-psychological reconstructions of the scientific endeavor, and such

24 WALTER B, WEIMER

dashes as the Kuhn-Popper-Lakatos debates have been interpretedprimarily in that light. But theory choice is a subject for rhetoricalanalysis, and an adequate rhetorical theory could add much to ourunderstanding of science as an argumentative form of discoursewithin a nonjustificational framework. Kuhn's recent remarks ontheory choice lead straight to rhetorical analysis, as comments suchas this make clear:

Nothing about that relatively familiar thesis [that theory choiceis not susceptible to logical proof] implies either that there areno good reasons for being persuaded or that those reasons arenot ultimately decisive for the group. , , , If two men disagree,for example, about the relative fruitfulness of their theories, orif they agree about that but disagree about the relative impor-tance of fruitfulness and, say, scope in reaching a choice,neither can be convicted of a mistake. Nor is either being un-scientific. , , . To understand why science develops as it does,one need not unravel the details of biography and personalitythat lead each individual to a particular choice, though thattopic has vast fascination. What one must understand,however, is the manner in which a particular set of sharedvalues interacts with the particular experiences shared by acommunity of specialists to ensure that most members of thegroup will ultimateiy find one set of arguments rather thananother decisive,''

Kuhn's last sentence summarizes the function of rhetoric in the un-derstanding of science: rhetoric can provide a framework andtheory for the explanation of decisions that result from theargumentative use of discourse, in both revolutionary reconcep-tualization and normal science practice,^" In both cases, scientistsface arguments (in print and in experimentation), never data.Rhetoric will be as indispensibie to the understanding of science associal and cognitive psychology.

Department of PsychologyThe Pennsylvania State University


NOTES

' I have discussed justificationism as a metatheory of science in detail in, for ex-ample: "Thc History of Psychology and Its Retrieval From Historiography: I, TheProblematic Nature of History, and II, Some Lessons for the Methodoiogy of Scien-tific Research." Science Sludies 4 (1974). 235-58; 367-96: "The Psychology ofInference and Expectation: Some Preliminary Remarks," in Induction. Probabilitvand Confirmation, ed, G, Maxwell and R, M, Anderson, Ir, Vol, 6, MinnesotaStudies in the Philosophy of Science {Minneapolis: University of Minnesota Press.1975), pp, 430-86; Notes on the Methodology of Scientific Research (HilSsdale, N,J,:Lawrence Eribaum Associates) (in press): and Structural Analysis and the Future ofPsychology (Hilisdale, N,J,: Lawrence Eribaum Associates. 1976),

' Justificationism may be characterized as attempting to provide a theory ofrationality according to which science is a rational source of knowledge. Within thisframework oniy those claims which can be justified are accepted as genuineknowledge and, hence, as rational. Logic was the organon, or tool, which provedknowledge claims "followed" from whatever epistemological authority was accepted(usually sense experience). When it was gradually realized that empiricalpropositions cannot be proven, the classic justificationist quest for certainty gaveway to the neojustificationist quest for near certainty (as the next best thing). Nearcertainty was identified with the calculus of probability, and the "logic of science"was assumed to be a probabilistic inductive logic, or confirmation theory, which wasto be the method of science. For a devastating critique of justificationist rationality,see W, W, Bartley, HI, The Retreat To Commitment (New York: Knopf, 1962), es-pecially Ch, 4, For the futility of inductive logic as the inferential metiiod of sciencesee K, R, Popper, The Logic of Scientific Discovery (New York: Harper, 1959) andConjectures and Refutations (New York: Harper, 1963); J, Agassi, "Towards AnHistoriography of Science" in History and Theory. Bieheft 2 (Middletown, Conn,:Wesieyan University Press, 1963); I, Lakatos, "Changes in the Problem of InductiveLogic," in The Problem of Inductive t-ogic. ed. 1 Lakatos (Amsterdam: NorthHolland Publishing Company. 1968). pp, 315-417.

' Karl Buhler, Sprachtheorie: Die Darstellungsfunktion der Sprache (Stuttgart:Gustav Fischer Verlag, 1965),

Popper. Conjectures, p, 295,Ibid,, p, 135,Ibid,, p, 135,W, W, Bartley, III, Morality and Religion (London: Macmillan, 1971). p, 33,Popperian arguments against inductive logic convince all who do not believe that

science cannot be rational unless it is inductive (see references in Notes 1 and 2,above). Convincing arguments against deduction as the logic of science in either in-ference or expianation are found in Paul Feyerabend, "Explanation, Reduction, andEmpiricism," in Minnesota Studies in the Philosophy of Science, Vol, 3, ed, H, Feigland G, Maxwell (Minneapolis: University of Minnesota Press, 1963), pp, 28-97; andP, K., Feyerabend, "Reply to Criticism." in Boston Studies in the Philosophy ofScience, Vol, 2, ed, R,S, Cohen and M, W, Wartofsky (New York: HumanitiesPress, 1965), pp, 223-61, The question then arises: having eliminated traditional in-terpretations, is science illogical? Beiow we argue that science can be both logical(and truth functional) and rational if justificational notions of logic and rationalityare replaced by more acceptable ones,

' N, R. Hanson, Pattems of Discovery (Cambridge: At the University Press, 1958),'" N, R. Hanson, "A Picture Theory of Theory Meaning," in The Nature and

Function cf Scientific Theories, ed, R, Colodny (Pittsburgh; University of PittsburghPress, 1970), p, 236," ibid,, p, 240,

26 WALTER B, WEIMER

'2 B, Mates, Stoic Logic (Berkeley: University of California Press, 1953)." B, Russell, A History of Western Philosophy (New York: Simon and Schuster,

1945),'* G, Spencer Brown. Laws of Form (New York: Julian Press. 1972). p, 78." C, I. Lewis. An Analysis of Knowledge and Valuation (La Salle, 111.: Open Court,

1946), p, 1," For example, see T, S. Kuhn, TTie Structure of Scientific Revolutions (Chicago:

University of Chicago Press. 2nd ed,, 1970): J, Agassi, "Historiography": 1. Lakatos,"Falsification and the Methodology of Scientific Research Programmes," inCriticism and the Growth of Knowiedge, ed, I, Lakatos and A. Musgrave (Cambridge:At the University Press, 1970), pp, 91-196: W. B. Weimer, "The History ofPsychology and Its Retrieval From Historiography,"" M, Polanyi, Personal Knowledge (New York: Harper, 1958), and The Tacit

Dimension (Garden City. N.J,: Doubleday, 1966)." Popper, Conjectures, p, 316," Ibid., p. 317." For example, "Against Method." in Minnesota Studies in the Philosophy of

Science. Vol, 4, ed. M, Radner and S, Winokur (Minneapolis: University ofMinnesota Press. 1970). pp, 17-130: "Consolations for the Specialist," in Lakatosand Musgrave: and "Problems of Empiricism, Part II," in Colodny, pp, 275-353,

" I, Lakatos, "Falsification"; and "History of Science and Its RationalReconstructions," in Boston Studies in the Philosophy of Science. Vol. 7, ed. R, Buckand R. S, Cohen (Dordrecht: D. Reide! Publishing Co., 1971), pp, 91-136.^' H, Johnstone, "Some Trends in Rhetorical Theory," in The Prospect of Rhetoric,

ed, L, F. Bitzer and E. Black (Englewood Cliffs, N.J,: Prentice Hall, 1971), pp, 87-88," Ibid,, p. 188.^' The proliferation of "new rhetorics" in recent years has reached epidemic

proportions. Consider the switch in stated editorial policy in PhUosophy and Rhetoricto save space for more likely to succeed articles. !n its first years, editorial directivewas brash, confident, even starry-eyed: the journal "publishes definitive articles onthe nature, scope, and limits of rhetoric, as well as incisive an.swers to the que,stionwhether man necessarily engages in rhetoric , , , ," (See PdR. Vols. 1-6,) Tlien themessage became one of resignation: "The editors . . . believe that the nature and.scope of rhetorical activity are philosophically un,settled matters. Papers which ex-plore rhetorical action, rhetorical art, or rhetorical experience without presupposinga priori conceptions of rhetoric are solicited" (See P&R. Vols. 7-9). Such positionsrepresent a concession to a perceived dilemma in the justificationist conception ofrationality which limits rationality by rendering it impossible to achieve. It beginswith this simple C)ue,st!on: Can we rationally defend the superiority ofthe rationalway of iife, of such rational procedures as the employment of scieniific reasoning?According to the argument, the answer is "No," because of the limits of rationalityper se. Bartley develops the argument this way in The Retreat To Commitment:

No matter what belief is advanced, someone can always challenge it with: 'Howdo you know?' and 'Give me a reason.' Unless this procedure is to go on forever,it must be halted at a 'standard,' 'criterion,' 'ultimate presupposition,' 'end,' or'goal' whose authority is simply accepted. If all men do not cease their question-ing at the same point, however, 'ultimate relativism' results. For there is noArchimedes' lever with which to decide among competing sets of ultimate stan-dards. Even if everyone did happen to stop at the same place, there would be noway to determine whether this universal subjective standard led to objectivelytrue statements about the world, , . , Obviously, a man cannot, without arguingin a circle, justify the rationality of his standard of rationality by appealing tothat standard, \e\. if he holds certain beliefs — for example the standarditself — to be immune from the demand for rational justification and from thequestion 'How do you know?' he can be said to hold them irrationally or


dogmatically. And, so it is claimed, argument among men about the radicallydifferent beliefs they hold in this way is pointless. For rational argument consistsin mutual criticism, with each man supporting all his beliefs with good reasons.The limits of rational argument within any particular way of life, then, seem tobe defined by reference to that object or belief in respect to which commitment ismade or imposed, in respect to which argument is called to a halt (pp, 90-92).

The skeptical critic of rationalism has a tu quoque argument ia justify rationally hiscommitment to irrationaiity. If this argument form is valid, then one has. so to speak,a "scientific" excuse for being unscientific! The mystic, theologian, or existentialist.etc. who does not accept the validity of scientific reasoning is now provided with a"scientifically valid" justification for saying "1 do not accept the findings ofscience!" If this argument were valid, there would be now way to justify science as ameans to knowiedge, and hence no rational reason whatsoever to prefer it to otherputative sources of knowledge, including the most irrational ones!

And what is this tu quoque argument? It is the combination ofthese three premises:(1) For the logical reasons above rationality is limited so thai everyone must make adogmatic, irrational commitment.(2) Therefore, one has a right to make whatever commitment he pleases: the scientistto science, mystic to mystical relevation, etc,(3) Therefore, no one may criticize anyone else's commitment, nor can anyone becriticized for making that commitment.

Those who understand the force of the tu quoque adopt what Bartley calU criticalrationalism, which consists in acknowledging that rationality is limited in the sensethat it cannot be rationally justified. From my cursory examination, the "newrhetorics" invariably instantiate one or another form of critical rationalism, with aninevitable retreat to commitment in order to evade the tu quoque. The usual retreat isto conventionalism, which gains immunity from criticism at the expense of giving upthe ability to criticize rivals. Thus journal editors quietly ask for an end to"theoretical" papers whose sole critical function is to clog up the journals, and re-quest instead the atheoretical exploration of what rhetoric is. This is exactlyanalogous to Nelson Goodman, in Fact. Fiction and Foreca.it (Cambridge: HarvardUniversity Press, 1955), proposing the "new riddle" of induction, that of explicatingwhat an inductive inference is, in place of the "old riddle," Hume's problem of itsjustification. But critical rationalism is no more successful in rhetoric than it is inphilosophy of science.

What is the alternative to critical rationalism? Bartley's answer is comprehensivelycritical rationalism, which abandons the quest for justification, as well as theauthoritarian basis for knowledge and rationality. This permits us to characterize arationalist as one who holds al! his beliefs open to criticism all of the time, A com-prehensively critical rationalist will never hold a belief or fundamental standard im-mune to criticism; he will never resort to faith or irrational commitment to justifyacceptance of any belief. Most especially, he will never cut off criticism of his con-ception of rationality: comprehensively critical rationalism is itself an approach thatis open to criticism. That is, this conception of rationalist identity does not lead to acrisis of integrity: it is, according to its own formulation, rational.

This conception of rationalist identity, which locates rationality in criticism ratherthan justification, is sufficient to break the power ofthe tu quoque: it is broken whenthe structure of the infinite regress is dissolved by identifying the essence ofrationality with criticism rather than justification. One can be critical_/brfver, nevershutting off debate, without ever justifying or failing to justify a position or belief. Ifcriticism is the essence of rationality, it is rational to hold an unjustified belief, andeven an unjustifiable one, so long as criticism is never cut short! Indeed, justificationis not always desirable (or undesirable) for the comprehensively critical rationalist:its merit in any given case Is always "open to criticism,"

This makes rationality rhetorical, in both its essential formulation and its practical

28 WALTER B, WEIMER

pursuit. Criticism, as the essence of rationality, is a rhetorical process which neverseeks proof or ultimate commitments. Instead, it seeks to articulate and assess themerits ofa position within Its own unique context, thus making rhetoric a "way ofknowing" without retreating to commitment or abandoning rationality.' ' See Weimer. "The Psychology of Inference and Expectation."'» Popper, Conjectures, p. 313," Robert Persig's Zen and the Art of Motorcycle Maintenance (New York:

Morrow, 1974) has occasioned considerable interest among rhetorical theorists.From the standpoint of this essay two things are particularly noteworthy in thiswork: Persig's attempt to put rhetoric in the spotlight usually reserved for dialectic(esp, pp. 338 ff,); and the Intensity of his response to the crisis in integrity faced byjustificationist conceptions of rationality, Persig's intellectual honesty in pursuingjustificationist rationality to its self-stultifying end led to his psychotic breakdown —unable to cope with the crisis in integrity, he abandoned rationality altogether, Per-sig also saw that dialectic need not dominate rhetoric, and that Aristotle'sjustificationist ghost need not always direct the field. Unfortunately he neversucceeds in totally abandoning justificationism: his revolution overthrows dialecticfor rhetoric without uniting them in a nonjustificationa! framework,™ C, C, Arnold, "Reflections on the Wingspread Conference," in The Prospect of

Rhetoric, ed. L, F, Bitzer and E. Black (Engiewood Cliffs, N.J,: Prentice-Hall, 1971).p. 195.» Ch, Perelman, "The New Rhetoric." ibid., p. 119." C, C. Arnold, "Inventio and Pronuntiatio in a 'New Rhetoric,' " p. 5, Paper

presented at Central States Speech Association Convention, 1972,'' P, N. Campbell, "The Personae of Scientific Discourse," Quarterly Journal of

Speech 61 (1975), 393," Arnold, "Inventio." p. 1,J' Wayne C, Bootb, in Modem Dogma and the Rhetoric of Assent (Chicago: Univer-

sity of Chicago Press, 1974) is an illustration. After chastising Bertrand Russell forbeing a comprehensive rationalist. Booth endorses a weak form of criticalrationalism, in opposition to "modern dogma" (comprehensive rationalism)," Amoid, "Inventio" p, 6," R, K, Merton, Social Theory and Social Structure {'Hew York: Free Press, 1968):

The Sociology of Science: Theoretical and Empirical Investigations (Chicago: Univer-sity of Chicago Press, 1973); W, O. Hagstrom, The Scientific Community (New York:Basic Books, 1965); M, J, Mulkay, The Social Process of Innovation (London: Mac-millan, 1972): J, Ziman, PiUilic Knowledge: The Social Dimension of Science(Cambridge: At the University Press, 1968). Ziman is also aware of the rhetoricalside of science: "The familiar 'method' of science, whatever its logical status andepistemologicat virtue, aiso has tremendous rhetorical power. If applied correctly, ithas overwhelming persuasive force" (p, 31)." A beginning attempt at redressing the balance is taken by Campbell, in "The Per-

sonae of Scientific Discourse," which is one of the few reference* to personae inscience," Arnold, "Reflections on the Wingspread Conference." p, 198,5* "Confirmable and Influential Metaphysics," Mind 67 (1958), 344-65;

"Metaphysics and the Advancement of Science," British Journal fortke Phitosophy ofScience 26 (1975), 9I-12!.» Kuhn, Jlie Structure of Scientific Revolutions, pp, 199-200.*" Normal science discourse, although it has not been emphasized in this essay. Is

equally susceptible to rhetorical aniysis. To cite oniy one point, note the epidicticnature of normal science, Perelman and Olbrechts-Tyteca, in The New Rhetoric(South Bend. Ind,: University of Notre Dame Press, 1969). p, 51, could be taken asdirectly addressing Kuhn's conception of normal research as conservative,enshrined, status quo preservation of the values of the paradigm:


, , . the argumentation in epidictic discourse sets out to increase the Intensity ofadherence to certain values, which migbt not be contested when considered ontheir own but may nevertheless not prevail against other values that might comeinto conflict with them. The speaker tries to establish a sense of communioncentered around particular values recognized by the audience,, , , The very con-cept of this kind of category , , , results in its being practiced by those who, in asociety, defend the traditional and accepted values, those which are the object ofeducation, not the new and revolutionary values which stir up controversy andpolemics.

Documents

Science as a Rhetorical Transaction: Toward a Nonjustificational