INFORMAL LOGIC XI.3, Fall 1989

Beardsley's Theory of Analogy

EVELYN M. BARKER The University of Maryland

1. Is the Argument from Analogy a consequences of an "abdication of Fallacy? statesmanship."2 Kissinger cites historical analogy as the only basis for sound judg­ In the various editions of Practical Logic ment in international diplomacy: and Thinking Straight Monroe Beardsley History is the only experience on which pioneered informal logic, the assessment of statesmen can draw. But it does not teach reasoning techniques employed in actual its lessons automatically. It demonstrates the human situations. 1 Although acknowledg­ consequences of comparable situations, but ing the prominence of analogies in reason­ each generation has to determine what situa­ ing, he contended that analogy has some tions are in fact comparable. heuristic uses, but an argument based on one Such analogical arguments, as Kissinger is inherently fallacious: points out, are directed to finding suitable Analogies illustrate, and they lead to premises for wise and prudent action. hypotheses, but thinking in terms of analogy Beardsley's intellectualist approach limits becomes fallacious when the analogy is us­ rational argumentation to asserting premises ed as a reason for a principle. This fallacy as evidence for the truth of a conclusion. is called the "argument from analogy." Consequently he overlooks or misconceives (PL 107) the character of some forms of analogical Beardsley's view implies that no serious argument prevalent in practical reasoning. reasoner will use analogies between things It is misleading to speak of "the argument in drawing conclusions, or be persuaded by from analogy," as Beardsley does, as if all another's introduction of them. According­ are of the same kind. 3 ly, an important source of human Historical analogy, or "learning from knowledge is proscribed: historical analogy, experience," most nearly resembles reasoning based on direct comparison of arguments of inductive analogy: From past situations with the present. observed facts about previous instances one In practical reasoning, however, one draws an inference about what is likely to must often reach a judgment about the be true of a new instance sharing salient crucial features of some particular situation characteristics with them. Beardsley denies, in order to determine an appropriate course however, that inductive analogy is "a suc­ of action with respect to it. Analogizing its cesful kind of inductive argument" (TS2 features to another whose favorable or un­ 66). favorable outcome is already known, far Beardsley's theory of analogical reason­ from being fallacious, is at least common ing exemplifies overformalization in reduc­ sense and at most wisdom. For example, ing diverse kinds of analogy arguments to in a recent newspaper commentary head­ a single form, then assessing them all lined "A Stable World: Lessons of World fallacious because of the logical War II", Henry Kissinger compares the characteristics of that form. This results in current international situation with that a "denaturalized epistemology" of leading to World War II, warning of the dire analogical arguments, one with inappro- 186 Evelyn M. Barker

priate concepts of knowledge, unrelated to define a fallacy as an argument that seduces reasoning techniques successfully employed one into thinking that the evidence provid­ in a range of actual situations. ed by its premises suffices for the truth of In order to track Beardsley's claim that its conclusion. an analogical argument is inherently fallacious, we must have in view his general theory of argument. Like many, he divides 2. Beardsley's Formalization of arguments into two exclusive kinds: deduc­ Analogical Reasoning. tive and inductive. A deductive argument makes the claim (correctly or incorrectly) Beardsley first formulated the principle that the truth of its conclusion follows of analogical argument in this way: "If X necessarily from the truth of the premises, and Y have a number of characteristics in because the argument is formally valid. He common, it is likely that any further calls all nondeductive arguments inductive: characteristics found in Y will also be found Such arguments make a lesser claim that in X." (PL 108) Few if any analogical "the evidence is sufficient to make the con­ arguments make such a sweeeping claim. clusion, at the very least, more likely to be He subsequently toned down the conclusion, true than false." (PL 197-203, TS228-29, schematizing the argument as follows: TS423-27) X and Y both have certain characteristics He insists that "any compelling argu­ a, b, c ... And it is known that Y also has ment must be one of these kinds, and another characteristic q. Therefore: X has moreover, it must conform to the standards the characteristic q. (TS2 65) of its kind." (PL 197). As Beardsley His reason for calling the analogical describes the argument from analogy, it is argument fallacious remains constant, a neither formally valid, nor does it fit his two thesis we may call the Principle of Infinite models of inductive reasoning, explanatory or Indefinite Variability, italicized by me hypothesis and inductive generalization. Not below: surprisingly, an analogical argument fulfills Beardsley'S conception of a fallacy: .. .if two things have a good deal in com­ mon ... the likeness may justify a further in­ An argument is a good one only if it pro­ vestigation to see whether they actually do ceeds in accordance with some rule of in­ have more in common. But it does not justify ference and thus can be justified by appeal our believing that they have more in com­ to that rule. If it purports to conform to a mon without the further investigation. No rule of inference, and thus acquires some matter how many characteristics a pair of plausibility, though in fact it violates that things have in common, there may be any rule, then it is said to be fallacious, or to number of other ways in which they are dif­ contain a fallacy. (TS4 26) ferent. (PL 108, TS2 66) As we shall see, the analogical argument With a little logic-chopping it is quite turns out not to have a justifying "rule of true that any two things, even identical inference" of its very own, or more exact­ twins, can readily be shown to have an in­ ly, its "rule of inference" contradicts a definitely large number of different principle Beardsley enunciates. In defining characteristics. But this logical point is ir­ a fallacy relative to its peculiar' 'rule of in­ relevant to whether there are significant ference" Beardsley betrays a formalistic similarities between the things analogized view of what constitutes "good argument" on which a reasonable judgment affirming as well as what constitutes a "fallacy". A an additional similarity may be based. In­ broader conception of both need make no ductive reasoning is useful because observ­ reference to the existence of a peculiar' 'rule ed similarities enable us to to make predic­ of inference." For example, one could tions about the future, which eventually are Beardsley on Analogy 187 confirmed or disconfirmed. tween blue-eyed parents and their offspring: In holding that analogical reasoning is fallacious because a conclusion may tum out John's parents both have blue eyes, and so do Jim's; John has blue eyes; therefore, Jim in fact to be false, Beardsley applies a must have blue eyes. (TS2 66) stricter standard to it than to ordinary in­ ductive reasoning-one appropriate to Beardsley's comparison of these two deductive reasoning. A "good" deductive arguments shows that they have the same argument has both 1) a valid logical form logical form, but that form is of such a kind and 2) true premises, so that the truth of that the premises may be true without the the conclusion follows necessarily from the conclusion also being true, that is, the argu­ truth of the premises. In effect, he accuses ment is formally invalid. In the first, we the analogical argument form of being in­ would not be surprised by the falsity of the valid, so that the truth of its premises does conclusion; in the second, we would be. If not guarantee the truth of its conclusion. But we insist on interpreting the argument from an analogical argument need not, and usual­ analogy as making a deductive claim, which ly does not, make that deductive claim. requires a valid argument form as well as Beardsley's formalization of the true premises, we would indeed have to call analogical argument tempts one to suppose it fallacious. the assumed resemblances a, b, c and the But the issue is, should we view' 'the inferred resemblance q are independent analogical argument" as making a deduc­ variables. Yet Beardsley's own definition tive claim that the conclusion follows of an analogy points out that they are all necessarily from the premises? Since interconnected: An analogy is an "extend­ analogical arguments are based on the ed simile" in which two things are com­ characteristics of the referents of the two pared in terms of the structural relationships terms analogized, the semantic interpreta­ holding between their characteristics or tion of the terms drives the argument rather part-whole relationships (PL 105-6). This than the logical form of the argument interrelatedness among the resemblances statements. The "must be" can reasonably mentioned in the premises tends to justify be understood as the less insistent "most an inference about a further resemblance. 4 likely" of an inductive claim. An invalid Like inductive reasoning generally, an deductive argument readily converts into a inductive analogy is not based merely on respectable inductive argument, as many in­ its stated premises but on all our knowledge troductory logic books say. If this is the about the world. The grounds for the argu­ case, an analogical argument, rather than ment include our independent knowledge being inherently fallacious, has more or less about the things analogized and the weight depending on what we know about characteristics attributed to them. Conse­ its actual terms and their known or infer­ quently the following argument, cited by red resemblances. Beardsley, would hardly be taken seriously: Another misleading feature of Beard­ sley's formalization of the argument from John's parents read Greek, and so do Jim's; analogy is the specification of its premises John likes horseradish; therefore, Jim must like horseradish. (TS2 66) as a comparison between a pair of things. In standard formulations of inductive We are aware that a person's eating analogy arguments, a number of previous­ tastes are quite independent of parental ly observed instances, YI .... Yn are com­ language skills, and would not credit any pared with a new instance X.5 So stated, such argument. On the other hand, the the truly inductive character of an argument following argument is supported by our is visible. The inductive canon applies to knowledge of the genetic correlation be- an argument reaching the singular conclu- 188 Evelyn M. Barker

sion that a new like instance has an addi­ them is increased, and the likelihood of the tional characteristic q as well as to one in­ conclusion that she is also a blue-eyed blond ferring a generalization that a class of things is more probable. If told that Mrs. Smith's with defining characteristic a has the fur­ sister has a different mother, however, then ther characteristic q. The canon includes, the degree of resemblance between sister for example, the rule that the greater the and Mrs. Smith, her mother and two number of previously observed instances, daughters is decreased. The likelihood of the more likely the conclusion is to be true; the conclusion that the sister is blue-eyed that a known connection between the pro­ and blond correspondingly decreases. perty q and characteristics a, b, c increases Beardsley denies this: "You can't even the probability of the conclusion. say that the more known resemblances there Once the inductive canon is seen to be are between X and Y, the more likely it is relevant to an argument from analogy, it is that X will have any further characteristic difficult to hold that it is by nature found in Y." (P L 108). His skepticism is fallacious. Like other forms of inductive due to the Principle of Indefinite Variability argument, its conclusion may turn out to be of individuals on which he grounds the false, even though the inductive reasoning fallaciousness of analogy. This diffidence contains no inductive fallacy. For example, goes counter to common sense and the in­ suppose you observe that my husband and ductive canon even with respect to a pair I both have brown eyes. By sound induc­ of things. tive principles, you would reasonably con­ In summary, Beardsley's rejection ofthe clude that our two sons also have brown argument from analogy is grounded on an eyes. In fact, the younger is blue-eyed. But abstract formalization of the argument and the falsity of the conclusion does not show an abstract logical principle. Since the argu­ any fallacy in the inductive reasoning. It is ment scheme is formally invalid, it falls out­ much more likely that his eyes would be side "good" deductive reasoning. Because brown than blue. of the abstract logical principle about the The inductive canon contains a special diversity of individuals he artificially limits rule for arguments of inductive analogy. induction from like instances to a conclu­ They reach a singular conclusion (not a sion about a class, throwing the argument generalization) that an individual Y, (not a from analogy outside inductive reasoning class of things with characteristic y) shares (TS 4 Ch. 2).6 Since deduction and induc­ a further characteristic q with the individuals tion are, in Beardsley's view, the only two XI .... X n. Now, the greater the resem­ "good" forms of reasoning, an argument blances between this individual Y and the from analogy has no place among legitimate other individuals XI .... Xn described in the reasoning strategies (TS2 26-31). In order premises, the stronger the inductive argu­ to be respectable, it must find rental accom­ ment. For the more characteristics known modation with a generalization. to be common to individuals Y and XI .... Xn, the more likely the inference that individual Y will have the further 3. The "Hidden Generalization" Thesis. characteristic q that individuals XI .... Xn are already known to have. Beardsley claims that a plausible argu­ For example, if you observe that Mrs. ment from analogy is a confused version of Smith, her two daughters and her mother some other argument form. He explains it are all blue-eyed and blond, you may well is often convertible into a deductive argu­ expect that Mrs. Smith's sister is also blue­ ment, whose major premise is a true eyed and blond. If told that Mrs. Smith's generalization. Both terms of the analogy sister is her twin, the resemblance between are instances of the same generalization. Beardsley on Analogy 189

The analogical argument itself is logically But Beardsley's account is mistaken: Here dispensable, once the unstated generaliza­ the generalization is not an implicit assump­ tion is recognized. tion of the reasoner, simply made explicit. If a reasoner has independent grounds for Suppose I say, 'John's parents both have blue eyes, and so do Jim's; John has blue supposing the generalization true, there is eyes; therefore, Jim must have blue eyes.' no point in offering weaker premises about Now, taken as it stands, this is an argument particular instances. The generalization can­ from analogy, and evidently a very feeble not be merely "hidden." Only when a one. (TS2 66, my italics) covering generalization is not known to be The reason for its being "feeble" is that true does it make sense to offer statements the conclusion, 'Jim must have blue eyes,' about particular instances as premises in lieu is "wildly jumped to" from the premises of the generalization. When its singular con­ about the eye color of John, his parents and clusion is derivable from a "previously Jim's parents. Beardsley defends this by tested generalization, " the argument from pointing out that it has the same form as the analogy rather than being reducible to a other analogy argument citing parental and "good" deductive argument, is child characteristics known to be unrelated, superfluous. that is, the parents reading Greek and the Beardsley's ' 'hidden generalization" son liking horseradish. According to Beard­ thesis underlies a recent text's presentation sley's analysis both arguments from analogy of "a new way to cast arguments by analogy are fallacious, because of their common in schematic form." 8 The argument is form. divided into two steps, the first an "induc­ But if an argument from analogy is not tive step", in which premises about one a deductive argument, their common form term of the analogy are evidence for an in­ is neither grounds for nor against the sound­ ductive generalization. In the second ness of the argument. In the case of a "deductive step", this generalization is ap­ nondeductive argument, the form only plied deductively to the other term of the serves to identify the kind of argument. analogy as a new instance falling under the Whether an argument is sound depends on generalization. substantive matters, not merely on the form According to this schematization, the of the argument. A reasonable person takes "hidden generalization," instead of being account of the content of an argument, not "previously tested" (as Beardsley holds), just its form; she would judge the arguments is itself proved in the course of the argu­ differently because of the varying contents ment, then is used to prove the conclusion of their premises and conclusion. together with the premise about the other Beardsley attributes the deceptive term of the analogy. Beardsley did not so plausibility of an analogical argument to a misread the argument claim: He correctly "hidden generalization": "When we make fixed the distinctive feature of the argument that generalization explicit, we can throw by analogy as the direct comparison of the away the rest of the analogy." (TS4 113).7 structural resemblances between two The original fallacious analogical argument wholes. is reduced to a "good one", a deductive Both views err in introducing a argument in which a previously tested generalization as an essential component of generalization is applied to a new instance, the analogical argument. The argument as follows: from analogy is useful in the absence of a John's parents both have blue eyes, and John "previously tested generalization" justify­ has blue eyes. ing the conclusion. The success of a series All people whose parents both have blue of arguments by analogy about particular eyes are people with blue eyes. instances could suggest some generalization 190 Evelyn M. Barker

about a class of things with some work has shown little aptitUde in the sub­ characteristic they share. Beardsley's exam­ ject; she is highly motivated, but is emo­ ple, 'All people whose parents have blue tionally unstable; the future assignments are eyes are people with blue eyes,' achieves not difficult but she is undertaking longer the status of "previously tested generaliza­ work hours, and so on. tion" partly because of verified predictions An individual (as Beardsley reminds us about the offspring of blue-eyed persons, in his Principle of Indefinite Variability) as well as Mendelian genetic theory of may show diverse characteristics whose in­ recessive and dominant heritable terplay in special circumstances must be characteristics. Few successful arguments assessed because they point different ways. from analogy are clumsy attempts to say This is not achieved simply by adding what a good deductive argument contain­ generalizations, for the applicable ing a previously tested generalization does generalizations would warrant different better. judgments, inconsistent with one another. Beardsley, I suspect, was influenced by The experienced instructor will often make contemporary philosophy of science in the an analogy between the student before her theory of explanation. In order to justify and past students whose academic success belief in the truth of a singular conclusion, and failure the instructor has observed. The he supposes there must be a true inductive analogy is the basis for a holistic judgment generalization (known or unknown) in­ about the present student's potentialities in tegrating singular premise statements about her situation. Analogical reasoning is effec­ particular instances with the singular state­ tive for dealing with variable combinations ment asserted in the conclusion. He appears of traits found in individual cases, which to be appealing to a Hempelian "covering­ exhibit "family resemblances," rather than law" model of explanation for singular con­ some constant characteristic. The com­ clusions, in judging an inductive analogy parison to other individuals cannot be read­ argument in need of a true "hidden ily replaced by some very complex generalization. " In his own view a true generalization, for the aim is to reach a con­ singular statment (not proved deductively) clusion fitting this particular individual in is inductively justified only as the best these special circumstances. explanation for a diverse set of facts The judgments of Art concerning in­ (PL Ch. 2, #10). Only a generalization is dividuals, who sometimes show an to be inferred from premises about like anomalous set of traits, or whose instances. characteristics may vary from a norm, will This theoretical bias generates a often make good use of inductive analogy, cramped view of a proven useful form of in conjunction with applicable generaliza­ reasoning. In practical life, we often make tions. Medical diagnosis and treatment often singular predictions without the help of involves such analogical reasoning. A pa­ "covering laws" or generalizations. Let me tient presents conflicting symptoms, or give a homely example familiar to us symptoms suggesting different diagnoses. academics. At some point in a course a stu­ Or her medical condition includes several dent will ask an instructor her prospects of problems; for one, a type of therapy is in­ passing, (or receiving A or B, and so forth). dicated, for another, that same therapy is The instructor has a plentiful supply of counterindicated. In distinguishing signifi­ generalizations for this purpose. But the cant from less important factors, a physi­ problem is that one and the same student cian's accumulation of experiences with instantiates a number of different analogous cases is often a patient's best generalizations leading to conflicting con­ security for accurate diagnosis and clusions: She is diligent, but her previous beneficial treatment. Beardsley on Analogy 191

3. The Phenomenology of Noninductive internal environment,' the bit of ocean ... that Analogy. the early creatures carried with them when they first undertook to live on land ... When Many arguments present an analogy be­ Nils Bohr constructed his model of the atom as a miniature solar system, with the elec­ tween two types of thing, sometimes from trons revolving around the nucleus, this pic­ different categories: the ancient analogy be­ ture led scientists to think of many ideas they tween the "body politic" and the human could test experimentally. (TS2 64) organism, and Judith Jarvis Thomson's con­ Noninductive analogy as "the favorite temporary analogy between a woman be­ in debate and polemic" plays a role different ing pregnant and being plugged into an ail­ both from such heuristic uses and inductive ing musician as his kidney dialysis machine analogy. It is usually directed to locating are well known. or relocating a thing in our world-view. 10 Beardsley's thesis that the argument Beardsley's common formal characteriza­ from analogy is a fallacy is less paradox­ tion of all kinds of analogy arguments leads ical with such analogies, for the difference him to misconceive the phenomenological in type of things analogized means that it goal, as we can see in his analysis of an is naive to infer a factual conclusion about argument making an analogy between por­ one type merely from its analogy with the nography and an infectious disease. (TS4 other. His view that the argument from 112-15) analogy is a fallacy parallels his assertion that a metaphor is always literally false (PL Here the writer is not merely presenting an 101). The former is, of course, formally in­ analogy between pornography and disease: valid, but only the most unsophisticated that each is harmful, that neither has positive value, that one can be 'infected' by either, reasoner would credit an argument based that both 'infections' can be spread, etc. He on analogy as formally valid, just as one is presenting an analogy to support a con­ who thought a metaphor literally true would clusion about pornography .... In the present miss its point. case the property p is perhaps somewhat S. F. Barker calls this kind of analogy unusual: It is the property of being "noninductive," for reasoning based on it something that ought to be eliminated. does not issue in a conclusion verifiable in Beardsley's formalist approach makes principle by subsequent experience. 9 Its role him interpret the argument as a straightfor­ is phenomenological, as Beardsley su~­ ward argument reaching a conclusion whose gested in his first treatment of analogy III validity depends on the plausibility of the a chapter on figurative language (PL Ch. hidden generalization: "Whatever does 2), where he observed that such analogies harm but no good should be eliminated." "put an object in a new light." (TS4 112). With this assumption made explicit, the Beardsley approved the nonargumen­ argument becomes a deductive argument in tative uses of such analogy, noting its which the analogy of pornography with an capacity to explain abstractions in terms of infectious disease makes "no logical con­ concretely perceptible things, as the tribution to supporting the conclusion." dependence of the individual upon the com­ mon good appears in the analogy between Pornography does harm but no good. body parts and divisions of the state. He elo­ Whatever does harm but no good should quently sketches the phenomenology of their be eliminated. use in science, illustrating or clarifying a Pornography should be eliminated. general principle, and even suggesting a In trying to capture the argument's testable hypothesis: "logic," Beardsley has overlooked that the It helps us to understand the nature and fimc­ analogy serves a phenomenological func­ tion of the blood stream to think of it as 'the tion, directed to making us see pornography 192 Evelyn M. Barker in a way that will affect our moral attitude analogy often lies in its depressing some toward it: not as an innocent and harmless features of a thing, while simultaneously private pleasure, but as an insidious threat highlighting others, to make it fit smoothly to self, spreading to others. The "somewhat into a specific world-view, eliminating unusual property" disguises this practical trouble-making bits. Thus the fetus, in the phenomenological aim in the predicate of Thomson analogy, is no longer an innocent the argument's conclusion, in Beardsley's dependent being biologically linked to the awkward reduction. The "hidden woman, but an unwelcome exploitative generalization" is a vapid ineffective alien. It is sometimes a whole world view, truism, that does not replace the provocative rather than a "hidden generalization, " that analogy of the original argument either empowers the analogy. logically or phenomenologically. A counteranalogy is sometimes more ef­ Beardsley's condemnation of analogy fective in dealing with the phenomenolog­ arguments incites him to offer three helpful ical use of analogy. In a recent legal case strategies for revealing its fallaciousness: involving the fate of frozen embryos belong­ 1) Weaken the analogy by questioning ing to a couple engaged in divorce pro­ ceedings, competing analogies with joint the assumed resemblances and pointing out differences between the two types of property and human lives dominated the reasoning: A doctor supporting the implan­ things. 2) Expose the invalidity of the' 'hidden tation of the embryos in the biological generalization" of the reduced argument mother exploits the analogy with the' 'tiny human": "Putting tiny human beings in a that justifies the inference. 3) Extend the analogy "to the point very cold space, deprived of liberty, depriv­ ed even of time, they are suspended as it where it boomerangs," leading to an ab­ were in a concentration camp. It is not as surd conclusion (PL 110-112). hospitable a place as the secret temple of But these methods often fail to blunt the a woman's womb. "12 phenomenological point of the analogy. For example, Thomson's analogy provides a powerful defense of abortion because it por­ 4. Reprise: Is There a Fallacy of trays pregnancy as an involuntary offensive Denaturalized Epistemology? intrusion in a woman's life, a violation of her own dignity that makes abortion look Despite his negative estimate of the justified in a moral scheme stressing in­ argument, Beardsley took analogical dividual rights and voluntary personal rela­ reasoning seriously, more seriously than a tionships. II It is easy to attack Thomson's host of textbook writers in this acme of in­ analogy as Beardsley instructs, by pointing formal logic. He advanced our understan­ out that the natural function of a woman's ding both of its powers and limitations. He womb is to nourish the fetus, and that it does appreciates the fruitfulness of analogies in so without impairing a woman's life or scientific thinking, because they suggest dignity, while no human body is designed working hypotheses capable of empirical to perform kidney dialysis for another adult proof. He denigrates appeals to analogy in human. polemical writing, where verification is not Beardsley attributes the strength of an in order, advising the reasoner to attack the analogy to the degree of isomorphism be­ weakness of the premises or ridicule the tween the things analogized, as a map conclusion by reducing it to an absurdity. represents an area's geographic features (PL He is silent, however, on the presence 106). If so, Beardsley's three strategies of analogies in reflective legal and would always work. But the power of an philosophical reasoning where they are in- Beardsley on Analogy 193

troduced within an extended course of Variability, he rejects it from the canon of argument, as in Thomson's paper. This gap "good argument." Consequently, when a reveals the deficiency of Beardsley's theory "critical thinker" meets an argument by of argument: It looks to logic and empirical analogy, he charitably looks for the' 'hid­ science for the two models of "good" den generalization, " or rigorously follows deductive and inductive argument the rules to expose the fallacy. respectively. Like many of us, Beardsley's aim was Logic and empirical science deal with to make principles of inference discovered subject matter abstractable from its setting by philosophers accessible to nonlogicians, in the concrete circumstances of human life. so they can do "critical thinking." "Critical When the subject of the discourse is not so thinking" can always produce the "rule of "clean," one often deploys a number of dif­ inference" that legitimates each step of a ferent thinking strategies to bear on the mat­ piece of reasoning, after one has schematiz­ ter at hand, in a way that resists formaliza­ ed it to reveal the form, deductive or in­ tion. Arguments using analogy comprise ductive. Only such arguments are "convinc­ one such strategy. In foregoing sections I ing," in opposition to the hodge-podge of point out how analogical reasoning func­ reasoning strategies characterizing the tions successfully in a number of "rhetorical point of view," which has the situations-discovering historical precedents power but not the right to "persuade." for prudent action, predicting individual (TS4 1-6) outcomes, evoking morally significant Is this "critical thinking" model of good features of a thing, classifying unusual or reasoning applicable to all thinking we novel phenomena. rightly admire? Does it express an elitist Beardsley's theory of argument leads conception of "good argument" which him to concentrate on the form of an argu­ favors abstract reasoning, overlooking ment in search of some valid principle of reasoning that does not move in a linear inference which negotiates the step from lockstep or fit the twin Procrustean beds of premises to conclusion. Since arguments us­ deductive and inductive argument? When ing analogy are diverse, Beardsley's we persuade our students that only "critical uniform analysis and common abstract for­ thinking" is good reasoning, and immunize mulation, by confounding them, makes his them with a list of fallacies to ward off condemnation of the argument type credi­ arguments that merely "persuade," are we ble. Since its "principle of inference" committing the fallacy of "denaturalized violates the abstract Principle of Indefinite epistemology"?

Notes

I In this paper I discuss Monroe C. Beardsley's equilibrium. Now that two-power world is treatment of analogies in Practical Logic, 1950. disintegrating-more in the East than in the New York: Prentice Hall; Thinking Straight. West.'· Article printed in The Baltimore Sun. 2nd edition. 1956 and 4th edition. 1975. Monday. August 28, 1989. 7 A. Englewood Cliffs: Prentice Hall. I refer to the first as PL; the other two as TS2 and TS4. 3 For example, I distinguish arguments of induc­ tive analogy, non inductive analogy. using an 2 ••... our generation faces a problem similar to analogy to support a moral judgment or defend that which engaged statesmen 70 years ago­ a classification, in Everyday Reasoning. how to construct a stable international order. Englewood Cliffs: Prentice-Hall, 1981. Most of the postwar period has been characterized by a relatively stable European 4 By "tends to justify" I mean the premises about 194 Evelyn M. Barker

observed similarities always or mostly confer p.282. some plausibility to the supposition of an addi­ tional similarity. Much depends on whether the 8 David Kelley, The Art of Reasoning with Sym­ further resemblance is independently known to bolic Logic, New York: W. W. Norton & be causally or structurally related to the observ­ Company, 1990, ch. 16. ed similarities. 9 S. F. Barker, Elements, pp. 225-228.

5 Note a contemporary formulation in Irving M. 10 Howard Kahane stresses the importance of Copi, Introduction to Logic, 2nd edition, New "background beliefs" and "world-views" in York: The MacMillan Company, 1956 (also in more recent editions of Logic and Contem­ later editions). porary Rhetoric, Belmont: Wadsworth Publishing Company. 6 Beardsley makes inductive generalization and explanatory hypothesis into mutually exclusive 11 1. J. Thomson, "A Defense of Abortion", types of inductive reasoning: But a generaliza­ Philosophy and Public Affairs I (1971). tion, as well as a singular judgment, may be regarded as an explanatory hypothesis for a set 12 Dr. Jerome Lejeune, quoted in various news of data; and each may function predictively as reports of The Baltimore Sun, August 1989. a conclusion from like instances.

7 S. F. Barker points out Beardsley-type inter­ EVEL YN M. BARKER pretations of analogy arguments as implying a PHILOSOPHY DEPARTMENT generalization make it commit a non sequitur THE UNIVERSITY OF MARYLAND or circular reasoning in Elements of Logic, 5th BALTIMORE COUNTY CAMPUS edition, New York: McGraw-Hill, 1989, n. 46, CATONSVILLE, MD 21228 o