Stephen Toulmin knowing ‘acting INVITATION TO PHILOSOPHY Macmillan Publishing Co., Inc. Collier Macmillan PublishersCopytieht © 1976, Stephen Tosinin Printed in the United Star of America All sighs reserved No pat ofthis book may be reproduced ‘etetumted In any form or by any meat leone ot ‘poten, copying ero ay Be ‘Demon orgs sn ta in'weting From the Publish NS Thied Avenue, New York, New York 10022 Colles Macmillan Canad, Li Lubrary of Congress Cataloging in Publation Data Toni, Sen Eden Bee co sae Printing’ 12363678 Years 6789012 Preface Let me say a word about the nature and intentions of this book. Anybody who merely adés one more to the existing supply oF introductory philosophy texts has a lot to answer for, and I dda not plead guilty to that charge. Rather, I have set out to produce a book that can serve as preparatory background read~ lng about philosophy, and about the philosopher's tasks and interests, of a kind that will serve not just “Intro Philosophy” students, but aleo students of the humanities, socal sciences, and behavioral sciences more generally ‘The occasion for writing the book is easly explained. In re ‘cent years, introductory texts in philosophy have tended to fall into two groups. There are those that treat the subject as being Inevitably a technical one from the word go: if the student is going to have to master the jargon and formal symbolism of professional philosophy anyway, the sooner he is pitched into fis technicalities the beter. Texte of this kind—and the best of them are very good—cin be of real use to students whose prior commitment to logie and philosophy is clear enough al- ready; but they too often drive away permanently those others who have a genuine feeling for the central issues of the philo- fophical enterprise, but no sense of the need for techniques or formalisms, ("What on earth have all these p’s and q's, dats and. wedges, to do with the problem of free will or the nature of human Knowledge?”) On the other hand, there are those texts that have taken care to he relevant: relevant, that is fo say, to issues that are already of interest to the nontechnical student for reasons prior to any explicit philosophical commitment. Books of this second kind also have a value, but they have defects of their own; the practical topics they deal with (abor- tion, civil disobedience, or whatever) are liable to be chosen for reasons of fashion or topicality, and the authors too often« asrace fal to make the necessary Links between these chosen topice and the longerdasting theoreti Isues debated by pliloso- ‘hers from Thales and Socrates up to Heidegger and Witigen- Stein, What on earth has all his trendy plea! stuf 0 €o ‘withthe problems of Plato, Kant, nd Marx?) There is no obvious reason, however, why these should be the only alternatives, The problems that have served as the starting points for philosoplcal dcusion throughout the hie tory ofthe subject and have drawn new generations of think ce and writers into the debate He ina, world of human ‘perience that is pio to all technlesitien. “The challenge thus {sto show the interested beginner why the effective dicuaion thee isues wil nthe ond become somewhat technical. In {heir main outlines, again, the chi styler of sack adopted by Plilocophers at diferent stages tn the Norcal development Of the tradition are not hard to describe in general terms. The {subject becomes laborious and diel o cover at an elementary level oly if we attempt to expound in deal the arguments that ‘we are led into when we follow there lines of atfack though {0 the end in one Beld of discussion or another. If we wish only fo convey a general sense of the manner in which philoso Thy has developed ae philosophers have brought new lines of [tack to beat on theif problems from one epoch to another, {his should not be imporsble fo achive. “The present book therefore taker as ie springboard a et of Assis that ll readers—however nonphlosophcal—wil be able to feel on thelr ules” These lrues turn out to have crucially Philosophical septs that serve atthe focus for the rest of the Exposition, Speccally, I have chosen to concenrate on the problem of relating two Kinde of explanation of human conduct *oaplantion in terme ofan agent “reason fr" his act and cxplanation in terme of the “causes of” hie acing as he does ‘ie gence of cure fair ayn enn twenteth-century sociology or psychology, eg. from the de- tates about Max Weber and BF Skinner. The thee central portions of the book accordingly deal with thee siles of Plilosophical attack onthe quertion, “What sit io have reneone Preface vil for our beliefs or actions?” and I have uced the different an- ‘were given to this question as « way of defining three tradi- tions in philosophy-a formal or logical tradition originating in Plato, 2 common-sense or empirical tradition most familiar from the eighteenth century, and a eritical or transcendental tradition inaugurated by Kant "My hope is that the resulting argument will be reasonably self-contained. Stl, there is no doubt that the best use can be ‘made ofthis book ifthe reader alzo has access to other books in the history of ideas, of science, and of philosophy and if he given ancillary readings from the key philosophers in each tradition. It would have been easy to clutter up the present text with detailed references, footnotes, illustrative material, and the rest, but the price seemed too great to be worth payin, ‘The fist priority was to produce a compact book which any- body could read—and keep on zeading—with the minimem ‘of built-in distractions and interruptions. How far have I succeeded in keeping, myself, and my own philosophical views, out ofthe argument? How far have I even fried to-do s0? That isnot easy to say. Certainly, in one respect at least, I have striven to be well behaved. For reasons that Ihave to do with the nature of our own times—partcularly with the tendency of all academic subjets to become isolated and specialized, philosophy among them—I have tended in my own professional career to adopt & deliberately antiformalist stance Free, by contrast, | have put the formalist tradition in the center of the picture, as having played (at least historically) » dominant role in the development of philosophical modes of largument {rom classical Athens on. At the same time, [ have ‘not disguised my opinion thatthe philosophical skepticism which has played almost as large apart in that historical development is merely the inevitable mieror image of formalism—the price philosophers have paid for taking #0 seriously the exaggerated aims originally made on behalf of geometry and formal logic {We can surely be grateful enough to Plato forall that he di, eg. in hs roles as the great-great-great-godfather of theoretical Physics and political science, not to fel traitorous in concedingthat he failed—as everyone before the eighteenth century 4.0, {alled—to recognise the limitations of geometry, both as a sci= fence of nature and as an exemplar of human knowledge!) In other respects, I have let my own fedings and preferences ‘become apparent wherever they have helped the argument slong. Since they are largely undisguised, the reader will be able 10 Aiscount them for himself, where they are not congenial to hie ‘own. The map the book provides is intended as a first rough ‘map of the territory of philosophy, and it will justify itself if It helps the reader to find hie way around the subject. The coloring of the map, however, is my ovin. ‘The conception of this book is owed largely to discussions swith Ken Scott of Macmillan, In the eazly planning stages, I was greatly helped by discussions with Nancy Baker. I must however take fall responsibility for the text, as it stands. It needs improving in all sorts of ways (as T know very well), but the time has come to let it go into the world and make its own way. If it helps others to find their way within the strange world of techalcal philosophical debate, and yet to retain the sense of being in touch with the problems that drew them to- ward philosophy in the frst place, it will have succeeded in doing all that I hoped it might Chicago Stephen Toulmin Bere exes 10. 11 2 Contents part I Introduction: What Are We to Make of Ourselves? Wonder, Puzzlement, and Perplexity “"Keow Thyself” Fatalism and Its Paradoxes ‘The Divided Image of Human Nature pet ‘The Philosopher as Geometer ‘The Claims of Logic Philosophy and the City of Truth Reasons and Inferences ‘The Rosds to Skepticim pt TO ‘The Philosopher as Anthropologist “The Claims of Common Sense Philosophy and the Human Standpoint Reasons and Consensus The Roads to Relativism a 2» “ Ea 53 2 a ut a3 12 20013. 16. pat IV The Philosopher as Critic The Chims of Function Rational Procedures and The Fontional Signieance The Preconditions of Good Reasoning The Thnee Faces of Philosophy pat V Conclusion: On Knowing Our Own Minds Philosophy and Human Nature Individual and Collective Self-Understanding Philosophy and Action 209 220 ar 29 Introduction: What Are We to Make of Ourselves ?1 Wonder, Puzzlement, and Perplexity When I was a child the family used to go for holidays to a cottage in the country, some fifty miles northeast ‘of London, In the normal way, I used to sleep in a tiny attic room, the only room on the top flor, with a bare dormer win- dow ooking out across the fields in front of the house. Yet on fone occasion (as T recall) T found myself using the front bed- oom, which was normally my elder sister's room and had a fine pair of heavy deep-red curtine across the window. Sitting up in the unaccustomed bed, 1 was intrigued by something about those richly colored curtains. What color were they? What vas their color, exactly? 1 closed each eye in turn and found to my perplexity that they looked slightly different when 1 was using my left eye alone—a fraction richer, deeper, more satu- rated—than they did when I was using my sight eye alone. ‘As I moved my head around, the changes were even more ‘matked. The play of the candlelight on the falling folds of the bination. The discussion of political affairs became common property, just at the same stage that language detached itself from the mouths and ears of individval speech users and be- ‘ame embodied in permanent, independent wetten form. So, the secularization and democratization of thought provided the new activity of “philosophy” with ite subject matter—namely, the rational principles of ethics, polis, and cosmology—while atthe same time the shift from speech to writing-—from "it is said that..." to “ie is writen that . .. "was helping to shape its methods. Once committed to writing and given an independent existence as records, “facts” ceased to be merely authoritaive sayings and acquired a new reality and permanence of their ovn. Similarly “arguments” ceased to be merely oral disputations between rival speakers, and. became independent objects of contemplation and rational critclem, namely, 26- quences of recorded propositions laying claim to some inner connection and coherence, (Recall the continued ambiguity, feven today, in the uses of the term, argument. Contrast, for Instance, “I was sitting in a bar when T got drawn into this "There are peoples in Aca today whose callectve History for the lst sic handed years I similarly, eroded inthe oral ations ofthe tal “remembrances: tad we fed Socrates himel reportedly complaining ‘Sat the apond of wring ma poling people's memories ‘The Claims of Logie oy argument," where the argument is a human interchange, with Pythagoras’ theorem depends on the following argument,” where the argument is a connected sequence of propositions) “Taken together, these changes created a new, and potentially universal forum of discussion; much of our traditional thought about rationality is associated with the idea of this universal Torum of Reason. On the one hand, politcal debate was taken jut from behind the closed doors of the Imperial Palace and brought down into the agora, or marketplace” On the other hand, written records and arguments were open to scrutiny and criticism by all alike, employing standards of judgment applica- ble by all humans equally. As a result, nether human nor divine affaire could remain any longer above criticism, The euisting order of the natural world and the current pattern of socal stracture alike manifested intelligible Iaws or principles In each case, human beings at lst appeared free to construct general arguments, or "theories", which would show what pria- fSples were manifested in the actual operations of nature, oF ‘what standards were relevant to a rational criticism of society. ‘Those arguments also had a new “objectivity”, They could be divorced from the plausible or authoritative tongues of paticu- lar speakers, set out on paper forall to consider, and judged for vihat they were worth, in thet own right. Fnaly, this same Kind of “rational criticism” could be applied, not just to the workings of the world and of societies, but also, reflxivey, #0 the, lives of individuals. In this way philosophers became launched, almost from the start, on their characteristic task of self-examination, and a road was opened which led, not just to physics and political theory, but also to ethics, to pure mathe- Inatics, and eventually to formal logic. ‘So the invention of philosophy in classical Greece was the Antellectual aspect of a broader historical change by which hu- rman beings began to accept a criiesl responsibilty for them- 2 the Athenian egos tike the Roman forum, was he central are ofthe city mtn which al cites could meet and discs together on the same lev than became symbol of equalty Before the Seas of lv aca ‘this and atonal.a ‘met rontosonnae AS cromeree seives—for the conduct of their persnal live, for the justice of their political arrangements, and for the stories of ereation in ‘terme of which they tried to make the natueal order intelligible If we examine this historical context more dlosely, we can also see why the newly invented enterprise of philosophy started off in the particular directions it did—why the first faural approach fo its problems was “objective”, and its initial program logicomathematical—or, more exactly, “geomet- ial”. For the formal procedures of Greek philosophy were jevised in order to mest some quite specific needs, and they Grew their inspiration from the simullancous discovery of 2 ‘new method for presenting and analyzing. arguments through ‘which the acience of geometry became the prime example of 4 "rational” body of Knowledge. As a result the philosophers Of classical Athens—notably, Socrates and Plato—put into cit- ‘ulation certain “formal” ideals that have alfeced all subsequent hhaman theorizing. ‘By adopting formal geometry as a general model for other areas of knowledge and experience, they could apparently satisfy three diferent sets of intellectual needs at once. As well as showing how one might replace the traditional mythologies by rmore “rational” accounts, Both of the cosmos (natural science) and ofthe pols (social theory, this method also pointed a way Of escape from certain paradoxes about knovledge and language by which philosophers were—not surprsingly—aflcted from the very beginning of their enterprise. Let us consider these three sorts of needs in turn, Politically speaking, then, during the hundred and fifty of 0 years before 450 5. (as we noted) the city of Athens was working its way toward a novel, democratic form of govern- ment by which all thore people who qualified as “citizens’— embracing a favored fraction of the male inhabitants of the The Claims of Logie 6 cityvere ble to take past in public debates and ecistons on tho pola rue fac the iy, ‘The forms of tat debate vr determined early on ough ie conaitonainovatons TF Slon and his sucssor, bat how Sols onto wa fo fncon eect nal nperoee anoher mater ‘Te paca quest wi! Could fe Athens lea hale the ls of douting ted lel jy, egy, and in pasty? Once they st sole the boditonal comment Tiny authored King pharsch or front he fe eect tight oily have bent remove the at foc of yay and tot rumeder to purtanipy puso op cod oe plays, wth te thea ano ent aveysin ee backgound Cet the new democrat ottton be erately operated then out of common yay tthe City Talo tven tom inore genera eal of ana ot rani? Betind Ent prt quston there ao lay theoretical question rely, the gosto “What i requted in oer fr pole, Id st hestons to become genuinely ratnalappeting ‘ot jst ome intrest bu fo Sener! pines binding Il suman being anf posble of pera niveau. ity” That gueston nas ooe of te esting pons forthe Systematic deepen’ of the posh aden a In TRgerted by Ta aches Seats "iar sor of man wes Sates? Much fou evidence about incomes fom Pte Dison whch the est of oth tren are ran fgeler and pened in» highly conscious Iter fom, Av afl shard e soe jt where "Scat rls ard Po bap So far te can tl om ou tance, Sores Hime wav fet sod Yocnst + socal tod mor eta man who suns te pola nd eal thet SF's time, andcalnged Reflow cltsene to scout for thelr oprtons in terms eapble sf meswising papain the seer nla tandade. In Ke youth cro had ES'stonlated by te more comoloily minded orl Phfospers of foin and aly, Pythagorean the Feats fora tne, be mas tnigued by tei guertons Soar te ervene and eee natal phenomena” (Phe postot ‘rr muitosorieen as ceoneerex Aristophanes in his ant-intellectual comedy, The Clow, e3 tatured him as unduly interested in astronomy) After a while, however, Socrates despaired of achieving. any rational agree~ rent on the physical constitution of the Universe because all the objects of processes visible in the Heavens were out of hhuman reach, 90 that their nature could only be a matter for ‘nresoluble speculation. From that time on, he concentrated en= tirely on ethics, politics, and similar matters of direct human concern. Subjects like astrophysics, geology, and physiology right be enticing, but in Socrates’ view they were of 10 i mediate relevance to the problems of human life. For himself, he beieved—with Michel de Montaigne and Aletander Pope much Inter—that “the proper stidy of mankind is Man”. As { result, philosophy must be concerned above all with the prin- ples of human reasoning, socal life, and personal conduct. ‘What does it mean, then, to take a political decision out of loyalty, not to the commands of a supreme ruler, but rather to ideale of humanity or rationality? In Socrates’ view this re- {quired the citizen to act always in ways that could be justified ih terms of general principles, that is, to be able to relate par ticular decisions or policies to an entirely general view of human life and the purposes of society. Ideally the specific goals of particular decisions could then be justised by relating them to the general goals of ethics and politics. Specific justifications were to be deduced from the general principles of a valid po- litical philosophy in the form of particular instances. The supposition wae that the “deductive link” between general principles and particular cases in politics can be presented in 4s strict and unquestionable a form as that between, say, the {general belief (or principle) that All men are mortal and the particular belief (or instance) that Socrates—being a man—is mortal. In practice, of course, practical life in classical Athens was carried on in a less theoretical oF intellectual manner than Soc- rates demanded, and his insistence on subjecting political actions and beliefs to a philosophical critique struck some of his fellow ‘tizens az “radial troublemaking”. As he himself saw the task ‘The Claims of Logie 6 of scl citciam, ete groups and scees often delayed the same confess aS terial aspen Sly and lee people wee compelled fo fice, in an xpi nd ciel manaer,the quetion of what they tally waned Gr what they relly believed, the whole polis debate would ‘enn srdd and confused. So, for instance, interred partes Tul be sble to drum up suport for desons that would be ‘ognied a indfenln Coker sod more sonal context. “the proce of plea seit war no, of cour sy or pails It canbe ats agonising fr people to reappraie thet callecve gas and polices win sncmbles and ote fonts tcf thm to fae with onerty the cofeion snd Sloppy-mindednes of thee ndviual lives ‘So: demande for ‘slexamination can be que a tvestening (0 oupr ao indivi and ae Socrates found to his cost they were capable Of provoking” an equally lent revtance: (The story of is Scctsaton, condemnation, and exciton on charge of "3 ‘pling te mds of the young” ha become pt of the Western triton through the aunt given in Pa'sdlages ‘The Apology and Phe) sil, the efor to uncover the general princi, ot presuppositions, undeying our bel and pol Ets mast be Denti in the long run. Thee was no oter ay te break the pall of exaished concepts or sumption, and $2 oblige people to consider the whole range of pies aval thie in te cual stone, (Tie "Sonate metho” tno ine evant today. To ive jm one topel example: Think ‘Tout the damaging fect on pli! deate in ut own tie of ievinguncricaed the serupons alt into woch concepts ts spss national produ and "economic growth") Above all for Sores and hie lowers rational acsson of pola sues meant an mpartldscaion. At hl hands {i he topes of anal eidem-argoment, pence, loge connection, questions of tah and vat, and 30 on—took cna fe med evens of ther owm, and Became mater for public eras independently of the puter invidals a eating oneview cr aneher. Again this sep sean not—and INI fates taken ts on thing to reogize the dei66 ‘He PimLosorneR AS crowerER ability of separating our personal egor from our arguments in theory, but many people still find this as hard as ever to do in practic. Still, from the ime of Socrates on, “truth” and "valid ty” had to be distinguished from whatever a glib or persuasive speaker could make convincing, quite as much as “justice” had to be separated from whatever men of power could impose upon their fellow citizens. Questions like “Are there in fact good ressons for this policy?” and “Does that conclusion really follow from what was agreed before?” must be debated and settled ina forum within which everyone had equal rights Either suctvand-such a conclusion really did follow from suck- land-such a principle, in point of fact, or elee it did not; either Satisfactory arguments could in fact be advanced 9 support suchvand-such a belief, or else they did not; .. . Rhetoric land power were beside the point. A genaine philosopher had tolook and see what was infact the case, ‘Thus, there came into circulation a new picture of “logical” entities and relations—teuths, implications, contradictions, and the rest—as cccupying not the physical world of tastes and feelings, preferences and opinions, but a "Third World”® of independent, sel-subsisting rational creations. These logical en- tities and relations stood by themselves, apart from their cre- ators, quite as much as statues, paintings, and other works of art. As such they could be viewed and criticized for themselves, ln their own terms, regardless of the intentions and wishes of thei proponents. Only by exposing them to public cltiism in the impartial forum of a common Human Reacon—Socrates believed-could we have any real confidence in their basis and strength, “Today this notion of a rational forum for impartial criticism seems all too cbvious and familiar; yet its invention some 2500 ‘years ago was a notable and hard-won victory. Nor was it an Irreversible victory, about which we can afford to be com- placent. Nowadays a5 much a5 ever, doctsinaire prejudice and 2 This phrase sa modern one, ntoduced by Kal Popper bat the ane desing Heats an old ne The Claims of Logie o individual caprice ave ther advocates. In many areas of morals, politics, and art, thee ae stil people who would gladly reverse Socrates” achievement by misleading appeals to “decent feel ings”, “good taste’, “personal iberty”, or "the national interest”. [AS a'matier of theory, the need for impartial, objective pro- cedures of ethical, aesthetic, and political discussion may be ‘generally admitted, but the actual practice of rational criticism Stull remains contentions. If Socrates were alive at the present time, he would undoubtedly become involved in public debates bout education or obscenity or capital punishment or the wel- fare system; and if he did so he would probably have as hard atime with our contemporaries as he did in his own lifetime when he challenged the intellectual habits of his fellow citizens in classical Athens During this same period the first important steps were being taken toward the development of "scientific" picture of the natural world, In southern Italy, the followers of Pythagoras were making discoveries in geometry; in Ionia, on the western coast of Asia Minor, Exdoxss wat elaborating. 2 mathematical theory of harmonies and proportions; in Sicily, Empedocles was working oat a frst account of the different forms of matter; tnd all these new idegs were entering into the intelectual dis- ‘cussion at Athens. Indeed, the subject we ourselves know as “geometry"—regarded ae an exact branch of mathematics, rather than os the practical art of "land-messurement”, or sur- ‘veying—first took on 2 coherent shape and method in the newly founded Academy of Athens, under the direction of Plato Ihimself, There, Thesetetus presented the Best sealy powerful and original modern geometrical theorem when he proved by a ‘igorous argument that there are only five convex “regular solide”* Meanwhile, « group of Plato's other associates, led “Regula sl" ar thyee-imensonl bodies, rach a the cbe, having cpilatrl (oa! sided) fguror ae fces The cbe has six squares 2810 [lan the dolecbadon har 12 polagone aes faceny while elsesby Eadovas and Heracles, launched a systematic attack on the problem that was to dominate stronomeal theory Hight up 1 1. 1620~the problem of combining simple Grxle moverents ih sch a way ost acount forthe ebrerved motions of the sm, the oct, and the planets ‘oowing the intl ower of geometry, Plato could not stand soo rom mathemati and sete theory, as Socrates fad done, Quite the revere. In his eer, the new mathemati Bp of game stony, sd Parone trong sports inept, not st atl pao By but to phlowophy in genrl. The discoveries of men ike rms, Exons, and Theseus had shown how much we ‘ight hope to do, within asrly “rational” since, to derive Stkingly unexpected tele in’ a rigorous logical manner, from evidently wall seumptons, Nor was the achievement Gin laos view) merely a formal and abtract one, conned 'o pose mathunatics ‘The athetatal theories of threlimen- Sal geometry were the key to « physical understanding of planelatyastonomy, while ‘Theses’ theory of the replat {alle mht very poly open the door sso Yo explaining the Giferencs between diferent matral substances Farthenmoe, the fora srngth and iterconectednese of geometry made it's model or cumple indicating what form our sandade of tonal citciom should take more generally. For example, the procedures by which Socrates had cntiied the pllcal st imptins of hie conterporariee might likewise be tested a= ‘xeres informal analyse. There, ain any Bed of argent the task was fo: make explct the formal elaions between feneal principles nd patialar applications and thot Tinks oaid then be a unbreakable a the connexons between the “Nioms”, or ntl asumptions of formal geometry and the mes form the facts of the teeaelron (fs), the octahedvon (6 face and the ecesshedson GO Tacs), Ae Theat showed one can eronstate by vary single segement that ony these Bve forms ate pose The Claims of Logic « theorems” deduced from them In this way, the science of geometry came to play a central part inthe philosophical ener- prise of Plato and the early Academy. Notice carefully that the Greek enterprise of organizing geo- metrical knowledge into. a “formal system” in which funds- mental assumptions and derived theorems were linked by deductive arguments—the kind of system that served Plato 36 «1 model of “rational knowledge”, oF epistome—vwas an entzely novel one. Elsewhere, as in Babylonia and Egypt, geometrical knowledge had consisted only in collections of practical rules fof thumb for use in land-measuring ("geo-metry"). As such, it represented not "theoretical understanding”, or episteme but rather fechne, or "practical know-how”. The Greeks alone dis- covered how the formulas of geometry could be linked together in a single, formal or logical system. As a result, they were able to connect back the varied properties of different geomet- rial figures, step by step, to more general theorems and axioms: and the “validity” of those theorems was confirmed by the absolute rigor of the resulting deductive arguments. Soon, in= deed, philosophers came to regard the pragmatic utlity of the practical, cule-of-thumb formulas used by earlier “geometers” (Ge, surveyors) asa pale shadow of their new theoretical con- ception of “formal validity”. Rough empirical estimates of geo- metrical relations (@,, the fraction 22/7 for pi, the ratio of the circumference of a ctcle to its diameter) gave way for theo- Fetical purposes to exact calelations within a formal system ‘of axioms and theorems, according to which no straightforward fraction could ever serve to give an exact value for pi © The fl presentation of casicl geometry, in exlilyexomaic form, vas gven only afer Pate death, in Cac Elements atthe nda ‘heorems and proot called together by Euclid In many eases dated ack Tie antiy canbe exposed numerically only by an enening series of deine beginning a4t100 To estimate value 2277 = 31008 fensined« perelly neque pronation for moet practi purer, bring covet to pst i 200"‘Thus, Plato's philosophical ideal of “rational knowledge” wae born. This ideal had a clear enough application to geometry itself. There, all the valid theorems were “entailed by” the Initial axioms of the system; that is, they were linked back to those axioms by formally rigorous deductions, so that accepting earlier axioms or theorems committed one logically o accepting also all the theorems that could be validly deduced from them, (Correspondingly, any doubt that arose about a later, derived theorem reflected back doubt, also, either onto the rigor of the arguments by which it was proved, or onto the axioms them= selves) In this way, the formal science of geometry became an intellectual construction whose inteenal articulation was abso- lute sound and secure, [As for the basic axioms that were the ultimate foundations ‘of that construction: these appeared to comprise a series of truisms, and so 10 be “selF-evident”, or completely transparent to the intellect. Who, for instance, could reasonably doubt that adding equal magnitides to other equal magnitudes would give ‘qual totals? Or, that any two staight nonperale lines drawn In the same unbounded plane would intersect at one-and-only- fone point? Thus, at one and the same time, the new science of Formal geometry apparently possessed two veparate and crucial requirements for absolute certainty and validity. It theorems were all strictly deduced from the inti axioms ofthe system, and those axioms themselves were unchallengeable. (Only after ‘a. 1700 was the status of the second truism, about the inter- section of lines in an unbounded plane—the co-called “axiom of paralels"—successflly called in question) As a result, the mathematicians of classical Greece seemed to have succeeded in erecting an absolutely sound intellectual structure on abso- Tutely fem foundations, Given this brand-new example, itis no wonder that Plto saw in geometry the model for any ra tional, self-respecting body of knowledge. And itis no wonder that for students at the Academy a grasp of geometry re- Pedy became a preregubite for al other branches of know. From this new, formal point of view, the professional phi- | The Claims of Logie na losopher had two concerns. In his role as a critic his task was to check the validity of the arguments that people use to sup- port their various belies, In this respect, the hope was that Eeguments in physics and poltce, say, coeld eventually meet the same rigorous standards as those in geometrical theory” ‘This meant replacing the practical idea of utlity as a measure of valdity—in physics and polities, a¢ in surveying—by the ‘heoretcal notion of "rigorous provabllity”; and showing how the substantive conclusions of those subjects could be formally related back to sel-evident fandamental principles. Meanwhile, fn his other role as an ethical, political, o scientific teacher, the philosopher's task was to expound the general principles from ‘which all genuinely valid Knowledge is to be deduced. In this respect the hope was that eventually the perceptive philosopher right somehow—by reflection, analysis, and/or insight —recog- nize the general trths which could provide politics and astron- omy, physiology and aesthetics, with foundations as firm and tanchallengeable as those of geometry itself ‘The general art of eritczing deductive arguments in all fields by the same demanding standards as those of geometry was accordingly made the concern of a further newly invented for- imal science, which was given the name of “logic”. Whatever the content of any arguments, their validity was from now on to be criticized in the same general terms, in terme of their form. All rigorous deductive arguments (t was Supposed) would prove to be valid or invalid, acceptable or unacceptable, for the fame Kinds of formal reasons. Thus, by a masterly stroke of Abstraction, Plato and bis colleagues transformed the whole problem of “good reasons” and “rationality”. The task of es [lishing that beliefs were “well founded’'—securely connected back to firm foundations—was thus divided into two separate, tnd supposedly simple, tasks. Fest, it was necessary to Identify hs dalogue the Republi, Pato quite exlly lds sins beech the forte geomet of tees Sensis an ‘Sina sry story haem a spn on an ‘ample foul be poieal thd vocal heern ‘HE MuMLOsOPHER AS ceOMETER the self-evident principles at the base of any field of inquiry, and then to check the formal validity of the arguments linking the beliefs in question to the basic principles of the field Third, we may lok at the formal program inplsophy as a way of excaping from cetln phloephica!paadonet ope ‘icaly, paradoxes about rationality and knowlege, bel tod languagenand see what ataconé the geomettl model of “lla Knowledge” possessed fom this tundpoint ton, The 4uestons that gave nse to the most fesing perpenties and Baradons stl occupy the center of our ow intelectual stage foday: namely, "How can there be ouch a thing aban nll dingtage at al” The dificultes over this question sere fst made clear by Crates, a pupil of Heracln of Ephesus. They naturaly ex presed thosves In the frm of paradoxes forthe ee feneral reasons that we denied in Chapters 3 and 4. We saw {here how eay its to be mised into the pradoncal postures oft, as areal of misconciving te rations Geo, say, explaining and jstifying recognising eects and geng sre, guessing at cases and giving reson, 50 ase, St ‘ery outset of philosophy, the ideas of Knowedge and language fave ise to pllosopical probleme, The rations bewec be lcving and Knowing sexing and understanding, hong sensory experiences and talling about the word of which Wwe have txperonce are suble nd comple, Te casy to msconese jus what have elatone ace and, having dons, tobe deven Into pradowical conclusions Latur se how this happened for the inmdiatepredeerorsof Scat and Pleo: Ht Here clits then Cetyae Hime. eralts had emphasized that ll our empirical bles sbout the everyday word et ultimately onthe tesimony of the var cus senses, We ble what we do about the word because of A the things that we can sehen fel, sell, and ste. So he had posed the queston, “What sot of validity can such ‘The Claims of Logie 7 elles tte Wee ele Satter beeen a sree tera ic taeig dt pcm ee ay tak erage fo ere ae, weds arya ea wie te Sra se meat racers ie en a icy si ea tte oma ah feo eae FF contingencies, 30, Heraclitus formulated “smack uted Tedw W from sensory experience alone, is a jomble of dis for pace, ask yours, "Wt dws the word prow’ mean nth conned S's nto thn coves a apply ich sro™ "HE riaosonmER AS cooMerER So fat, nothing specially paradoxical has emerged from Hera- ality’ argument. The person who underlined its perplexing consequences for the theory of language was Cratylus, Suppos, ing that we take Heraclitus’ argument entirely to heart, he re torted, we shall have to conchide that intelligible language is impossible. For language is part of the world of the senses, and four beliefs about language are just az dependent upon the senses—notably, on sight and hearing—as all our other beliefs about the world, How can we find out what (say) the word “earle” means? Evidently, we can do s0 only by listening 10 man using the English language and watching what he does with it, So our linguistc knowledge, our knowledge of what words mean, lies within the bounds of the same contingency that Heraclitus had very properly Isid down for all sensory knowledge, Suppose we observe a particular man using the word “scarlet” to indicate some specific shade of red, All that this observation strictly entitles us to say is, “Our senses tell ue that the word ‘scarlet indicates (to kim, here and now) this exact shade of red.” Once again, this statement, taken by itself, can be used to prove nothing further about, for example, what the same word means to anybody else, at any other time or place, Still less does it justify us in leaping to the general, necessary con- slusion, “Scarlet is a shade of red.” That conclusion would loverstep the legitimate bounds of experience, implying. that scarlet is not merely-contingently—a specific shade of red to 4 particular person at « particular time and place, but must be the same thing. permanently, everywhere—even necesarly— for anybody at any time. On the other hand—as Cratylus noted wrily—filing general and stable “meanings” language is apparently unintelligible How can two people communicate with each other linguistically at all unless, at the very least, their words mean the same 10 ‘both partes throughout their conversation? If all that anyone can stricly say is, for example, “At 10 a, on November 30, 1973, in Chicago, Hines, 'Scatle” indicates this shade of red ‘The Claims of Logie 7s for mei” that wil scarcely provide the stable basis we nest frefore Wwe can we the sume Word in commenting ith other fe, at other times and places Pius mas one of the fst martes to philosophical con- sistency. Having reached this conlision, Artal tell os, be fi his best to lve in accordance with and from then om he declined to sey anything but mesely wagged his ner! Even fo, was he completly consistent? Shosld he not rather, have fen p al atempts to communicate with his flow men? On Aso argument signe and gestres were no beter ff than words: he‘had no way of Knowing thet hs Ainge-wagging Mould mew the same, both to ime now, apd. to someone Jeet ec torom er een ina mie So the” of Heraitus apparently destroyed the possblity, not of purely Stouary knowledge alone, nor merely of verbal Tangzage, But SF all symbol spats whstever [Notice that Cats lands inthe same kindof perplenity as the fatal They claim to ve us good reasons fo believe that me can never believe anything for good reaons whereas Crys uses ineligible language to explain why intligible [Rnguage is imposible. Ths the argument to cscedt langeage ingly uneret isl Yet where in Cratos’ sequence of Steps could one find fal? The whole argument apparesiy Fogg together well enough, Faced with this pase, Pato on Chaded that nothing else could be done except to stand the entire argument on is head “The original sequence of steps had gone a5 follows [Al of our knowledge of the world comes by way of the five senses; So, all of our knowledge ofthe world is contingent; So, we can make no necessary or permanent Ssvertions about anything in the world—even bout words and their meanings:So, language is “in fex"—from place to place and moment to moment—lke ‘everything else; So, we cannot use language intelligibly. (On the other hand, nothing was more obvious and undeniable than the fact that we douse language intelligibly. Although all of Cratylus’ individual steps were apparently valid, his con- clusion was just as obviously false. That left only one alterna- tive: to challenge the stating point of the argument, Reversing the sequence of steps: We do use language intligibly; So, language cannot be entirely “in Bux”; So, we can make some 7 necessary or permanent assertions about the meanings of words; So, not al of our knowledge of the world |5 contingent; So, not all of our knowledge of the world comes by way of the five senses alone. ‘Wenay pu thet exrne sent slong oe anther the a an hagas abe ool ne ie pil vay inp Cocrdng to Put) he oe Pane oe Fir sor sf bool oat i sb hc Wt i is alematve suc of knot? Pas ie imine ansner wart Ineleeaal leigh had what eomsle Be hve af Sncey ae ae te sry Once apn Fao hed sn mele saver Gaonetea Inioge Har tung cope it We se ps crake dlp 0 les degen nat dy pte ‘The Claims of Logie 7 some local, contingent, and transitory plece of sensory informa: tion, For in that event, I could only prove that they added up to 160 degrees for, say, one particular triangle at Chicago on [Novemiber 30, 1973. Yet, evidently, the tuths of geometry are ‘quite general and are not concerned with particular times and places—with some triangle in Chicago, rather than in Cairo fr in Peking—with ap, 1973, rather than A.D. 44 of 2500 2. ‘Aboct the well-established theorems of formal geometry, there- {ote the Heraclitean ieeue of "cootingency’” does net arise. Well- ‘established theorems derive their validity less from local sensory ‘observations than from their places in the formal framework ‘of geometrical theory. We prove that the three angles of any plane triangle must add up to 180 degrees by rigorous ded: fion, vithout having to observe or measure any actual triangle, in any carefully drawn diagram, however perfect; while sensory observation of such an actual diagram could, by itself, never fstablish what the sum of those angles must be, only what it in fact is ‘What held good for the truths of geometry held good, in Plato's view, for the truths about language also. We can con- firm the statement that scale is a shade of red without having to find actual scarlet objects, observe them with ear, and check the truth again and agsin--as we should be obliged to do if this were really & contingent, sensory truth. Once we have properly grasped the concepts (er “ideas”) of scarlet and red, the intercelatedness of the two ideas is quite evidently a nec- fisary and general matler, not a contingent or here-and-now afin. There is n0 rick of our subsequently discovering, some Sneaky counter-example: some actual object that contrives to be scarlet without being red. (What would such a thing be?) ‘Therefore, despite all Heraclitus’ warnings, we can confidently vith his lst gmat, we cam put on cord one of the cental—and continously parting -sinsions of pilsophy namely the distinction Beemer Neary and “contingent lth, egy beeen "The angles of ‘Tirmnple od up 280 degree and “The sn i ow shining” Hew can “ny tithe be Reena? And how cam ne Row anything aboot sach selstery tthe? Abou thove guenions, the debate continues7 {HE PimLosorneR AS croMETER sere « neesary tath—"Anything that & cet ust In this way, the geometrical ideal of episteme (or “systematic ional knowiedge”) became central to Plato's account of knowl- tg and language, as well as politics and geometry. The “abso- Iutely true knowledge” that we have of geometry is not to be achieved by accumulating sensory observations of actual tr- angles, circles, and the rest, but comes from reflecting. on the properties of ‘certain intellectual ideale—the perfect triangle, ‘itcle, etc which are independent of actual material objects. Az for the more or less “triangular” and “citcular” objects and figures that we can handle, draw, manufacture and/or observe with our senses—these ae, at best, approximations to the under- lying ideas... . The same intellectual insight is evidently required—on Piato's view-—if we are to share a common in- teligble language. If the words of that language are to have enduring, meanings for all of us, we must grasp these meanings {intellectually as we do the ideas of pure mathematics... At this point, we have arrived not just 3t Plato's geometrical ideal of knowledge but atthe threshold to his most famous doctrine, the so-alled "Theory of Ideas” For the ments of he Acad, the the femal po gram characterise of Plat's phlosophy-te program dat Frese gromery as the idea of ated! knowledge ood ke ‘giving reasons” with “proving’™~hadatacons of atleast, three eiferent hinds, To say this is not fo imply that those attractions were “culture bound” hat sted tothe ape Sic context of the Academy, of of lasial Athens. On the Conta, once the new problems of phlsophy had been raed in that parte cote, they fenaied te be dealt wil and the mers of diferent lines of attack om them were largely independent of tral ifn The momen we ate to move away fom purely mytholgil sceouns of he cosmos ted the pli and ive ot Hes but ature aed svity some imoe univer basi, we shall have to take the propos of ‘The Claims of Logie 7 Socrates and Plato seriusy, Silly, the moment we sec to five any lear account ofthe relationship between (4 kro Sige sense experince, and langage, we shall be contonted by Cottylus’ challenge and Plato's response. By throwing, uP ost ands nd flaring fo pretheoretical oc mythological sane, tre car petaps turn ou Back on thse philosophical problems Th dele: Otperwise, they remain tobe tackled head on Tn geist of htory, then, the formal program inated by Socrates and Pato ho ocuped a ental position in pilsophy Srsnce So it sno srpae fo ind similar eas bing argued for once again inthe seventeenth centry a, at atime When the imtlettal foundations of modern sclnce ese being lid wn. Many scholars would argee, indeed, that Plato was in SEen ane ofthe gottathers of moder physi. For the basi {aterm of physi theory. as system of anioms and deductions Toi natched up ageinn experimental observaton—the patie followed, for tstance in Tnvac Newton’ classic Mathematical Principles of Natura Philosophy of 1687—had been advected Teme ty years eal by René Descartes for reatons very uur to Pat's own, and, however far Newton disocated Himself from Descarte cents scenic notion he substan Tal followed Deserter’ methods of theory construction, ‘Both Socrates and Montaigne (we remasked) ended by being sepa abot the posoliy of saional phys or cosmo Sep Intstcenth-century France and Tay, as in hithcentary Be ena here had been a mulitade of rival theories about the ‘tural world, yet no apparent way of deciding Bebveen then Sek oy mae han sonion seed, wa 0 Shun phlosophy aside In favor of more “humane” topics. ‘though no. sthentic knowledge could be hoped for about the coatation ofthe Heaven, othe general Iaws of ate, SAS he remained to be investigated im aman hfe, closer at Tund Thie at any fate, was the proposal of such “humanists” pe Mantagae whe reiterated for his own contemporaries the TANOTEI Siete fad given come to thousand years erie. Yai Daten, im the net generation, was as unhappy about UNE Slepucal’ posture a Plo had boom at the ear stagecd THE MinLosoPsER AS ciomerER (Our ideas about the world were not pure Babel, as the human Jats suggested. In geometry, at any rate, we had a subject that «ould be expounded in a formal, systematic manner, acceptable to all men alike whatever their other intellectual commitments. So, like Plato before him, Descartes saw geometty as an in strument of intellectual construction—a universal template, or model, against which we might test out other would-be bodies oF knowledge Granted, the results of physics could not intially hope for the absolate formal certainty of mathematics bu, rather would be hypothetical. For instance, a geometrical theory about the shapes of the atoms of diferent material substances (euch as that which Plato had expounded in the Tirmaens) was at best “likely story”—a “decipherment” ofthe symbolism of Nature having at most, in Descartes’ phrate, “moral certainty”; or, as ve ourselves would put it, a theory having a certain degree of verisimiltude, or probability. All the same, we might in the long run hope to confer the same “rational” status onthe Funda- ‘mental concepts and principles of physic that had been achieved for geometrical figures and selationships by Pythagoras and Plato, Theaetetus and Euclid, The initial axioms and postulates of geometry, as Descartes saw them, captured those “clear and distinct” ideas about spatial magnitudes and relations that God has given all men in common. Accordingly, they provided a ‘uniquely firm ground on which to construct rational hypotheses about the physical world, while the formal rigor characteristic ‘of all strict mathematical arguments gave them also the kind of strength needed by any well-constructed theory. Certainly, there was no clear alternative route for arsiving at a picture of Nature whose concepts and. principles could command general assent, The Internal coherence of geometry ‘vas unquestioned, and it was built on seemingly rock-hard in tellectual ground. The task vas, presumably, to. work from there outward and to develop physical theories of mechanics, heat, matter, and the rest that would serve as straightforward extensions of this geometrical core. (Something of this ideal has ‘remained operative in all subsequent physical theory. Think, for The Claims of Logie a inane aout he way in which Newton, Manvel Es and the other great schitecte of physi theory have ep ‘fore their eyes the vision of a comprehensive and unified mahal nto pope the seventeenth: and eghtenth-centary phlosophes, issue there were some sharp difereses oP opinion sbet the bse avons, potltes, or data from which the rest of our inowiedge as tbe constructed, proved, or deduced. Once fre go a slap ov tio into ther actual debate, we shall a5 2 Tell nd snes bacoming ery compen, wil rival schools oF emp", “atonal sod coon, aathing the same frometal model of Knowledge against diferent aspects of [Xtnun experience in quite diferent ways. Yet the popes Common leyalty tothe formal program isa estas important Sth sebequent dilerencer Spinoza inhi Ei etbiz is Ths vn He aN Treva of em. {for granted thot we can give a good reason (or rational ost- fecal) for our belt, adequate by phlosophical standards ny if we fate those Blfs back by formally rigorous dedc- Se atguments to "nn ground” of whatever Lind they then 19S much then forthe Hoc context within which the fomal program for pilorphy fest showed ts atactons. We ‘Tall shor tun and see ow it works ost, a applied fo some Simple philosophical probleme. But, fst, a brie digression te inate to deny one curios and characterise asemption anderying this Formal or “objective” approach t piosophy.6 Philosophy and the City of Truth sn elaborate and pervasive metaphor ‘up around this init " to philosphy, se aunts intl “pamela hy henge redlang agey ee ENP Seed agh ou Ceay Hoag and pace Te aes phor come tm the wold of woke er et lang ontrecson dpi gal fated a thecontcton aft Berea Cy atone ae be ‘arowy bee ve man-made ec, Ike oe ney hee parts of a building. Some of these constructions ake steongh built, and so intellectually safe. Others are more ramshackle, and we inhabit them at our own risk. The success of ot ™ putes is measured by the strength of the sesuling constr, tone, and the nal ambition of the human reson fifo ae 2 complete ad pomment ales ne ees ee sin "he Sy eh fo hs pot of es lial fey 7 eile te demands Ost bland mgunens sant bat ol Smtr sd ache fo frmtgaaed Paneer ooo Spuments may be we allow He renga ene Philosophy and the City of Truth 3 questionable assumptions we are building on sand. Correspond- ingly if our structure of arguments is not rigorously connected together, the building may collapse around us however frm may be the ground on which it stands In any field of inguiry where the results can be set out in formal manner, lke those of geometry, there are clear enough procedures for checking the security of our belies. The basic {xioms and postulates of such a formal system provide the necessary “fren ground”; the deductive steps linking axioms land theorems provide the “rigid structure”; between them, they give any validly deduced theorem all the intellectual guarantees Wve could apparently ask. In thie way, the philosophers initial acceptance of 2 geometrical model for theory and knowledge twas ragarded as setting 2 pattern for all intellectual criticism; and the ohole terminology of logical discussion has, as a result been shot through with architectural analogies, Reliable beliefs are “well founded”, so that they put us on “solid ground”. Data are “hard” or “soft”, so that they will support more or less "jeight”. Trustworthy theories are built on “firm foundations", and the “rigidity” of their “structures” enables them to give corresponding “support” to their conclusions. Other arguments, by contrast, are “shaky” or “il constructed”. (Notice tha their failings, or sources of “weakness”, can thus be described in structural terms quite a8 much as their virtues.) "The hold this architectural metaphor has on our ways of thought Is a tribute to the metaphysical charm of its asociated imagery. Individual items of secure belie, or knowledge, resem- ble particular rooms or bulldings-—in spectacular cases, individ- ual skyscrapers, Well-organized theories resemble complexes of buildings whose diferent rooms or structures lend each other support, s0 that we ean move with ease from one to another. Meanwhile, the ultimate totality of collective human knowledge + wheter cory valle in such system can be proved to be vail te anor sate Since the work of COS, mae twenteticentary mathe tuticans have trations boat thi, weno f floats on the horizon as an ideal, lke some Eternal City, made up of none but well-attested truths. (Is this City of Truth a reality, or is ita mirage? Who can tell?) ‘As for human rationality: from the present point of view, this is the capacity to arrive at secure, strong, well-founded knowledge. It works with two complementary instruments, ex perience and logic. Experience determines the site on which the House of Knowledge is to be built, and procures bricks and ‘mortar, lumber and roofing materials. Logic checks the design of the skeleton or “framework” (another building, metaphor!) from which the House gains its strength. And the specific "rea- sons” we give in defense of particular belies demonstrate that wwe know our way around within the House; that s, we show that a particular belief is “well founded”——we prove it—by pointing out just where in the entie construction itis located, and how it obtains the necessary firm support. The influence of this imagery is already clear in the classical Grock philosophers’ discussions of knowledge, belief, and truth, What isi, for instance, to have “reasons” for your be- liefs? Or, to ask a stronger question, What is it to “know” something? From the present point of view, all such notions as well-founded belief, knowledge, and rationality reer, either directly or indirectly, to the formal arguments by which we ‘would justify our claims to knovledge or rational belief, In a sentence: “To know something i to believe it, because we can support that belief with a conclusive argument.” ‘his account dates back to Plato himself. We find him dis. ‘cussing at length a proposed definition to the effect that knowl- cege is the same as “true belief” combined with a logos, ot “verbal reckoning”. Claims to knowledge or well-founded be- lief are to be established (using a somewhat diferent metaphor) by a kind of “intellectual accountancy”; and our claims stand ofall on the merits of the resulting intellectual balance sheets, Having. good reasons for our belies, correspondingly, means being able to prove the soundness of those beliefs, that is, t0 show that they are well based because the supporting. argu ‘ments and grounds are sufliciently valid and strong. From the Philosophy and the City of Truth 8s standpoint ofthis ancient tration, “glvng reasons” isthe same 4 proving, “having good reasons” i the same ae being able 10 rove, andthe “rationality” of our thoughts and belles is mea ured by the sttngth and solidity of the format proofs and founde on which they rst. Still, despite their familiarity, these architectural metaphors and images have nothing compulsory about them. On the con- teary, ther plausibility and charm depend on our making * Stl cman: nae, at pene fe a seneeregarded ar the outcome or product of the activites of angulng, Justifying, and/or giving reasons—can be considered {gute separately from those activities themselves, that is, from “arguments”, in th alternative, human sense. As we remarked cave, the colloguil term “argument” embodies both of these Senses. There is the Furman activity sense in which arguments can be cool of heated, swift or protracted, and can end in en- thusiastic agreement or tears; and there Is also the intellectual outcome sense, in which we can begin identifying and con- sidering any “argument” only after It has been made explicit, that is, only after itis presented as a sequence of propositions, preferably in writing, whose validity or fallaciousness, support ‘of lack of support, can be directly investigated. (For a hundred ‘reasons, of course, human arguments may end in agreement ot Aisagreement, regardless of the intellectual merits of the formal arguments implicit in them Valid, well-founded reasoning may Fale move eae: wel-praed Falls may carry the day) ‘The preliminary question then arises: How far is this assump- tion legitimate? How far can “arguments” (considered as tellectual constructions or creations) be legitimately isolated and analyzed independently of the “arguments” (considered as hur ‘man transactions of interchanges) that provide their context or ein Eien the formal lp or genet! apron to philosophy—particulatly tothe problem of good reasons— will be open to us, only fo the extent that this assumption i legitimateoa ‘Ht PinLOsOPHER AS crOMETER Granted, the history of serous philosophisig. began wih just this abstraction. As acest, We products of ou rtonal Inge crne tobe teeta as "mental works of at" jets cing in te own ight, sanding on thrown fet, needing 1 be fuged in thir own terms formal argument>—qulte ‘ide from all qertione about human sguer Granted, so, 2 great many subsequent philosophers have adopted the same formal approach They have deme al questions about argu: ing and persuading, investigating ond apecting at “emp” mutters for study by peyehologie of eoclogots. They have Sra ied on the Ted Word af inlet en tes'—date and. xguments, inferences and. principles whore tri they hve fbr ss intend’ fal he chance ad tlonycrates of Ban history. ll hese, in explosng the plilosopiel consequences of the loicogeometialspprosch, End it problems, we shoud keep in nd (even ony at he xc of our minds) both the Ntorcal dramstancer of it ie todocton snd the asumpone that restet ite application "Ar we go alng, we shal find that other approaches ar, at the vey lst posle Instead of rocking formal sguments from active asin, for instance, we might alternatively bepin by viewing "anging” ae 4 foncon, o performance, wich olan ih tan gel pup mi A tre might then develop terminology for des i's eterno “age” tal ater than in Biacural ones. other words, we might brgin, not by sky shout fe foundation and sen spe, bo rae: baking iow one can argue for ble! with complete effectiveness, of prevents propery convincing cate. To do tis would of couse Trench out dscston in avery diferent direction fom before— toward rhefoerather than formal ogi-and indeed, in dis ‘wane of shelve and dba, the funtion etcam of fryuments i entirely Fela. Thur the cascl tradition Sn filonophy started (we might say) by making a threefold die Erction, “acguments” were abstacted from "arguing" “logic” vias diferentated from “thetore™ nd “Formal rchre” a spate from "sbi fancon™ Philosophy and the City of Truth @ We can in fact go further: for, as we shall see in due cours, alternative approaches to philosophy are not only possible but necessary. Although the metaphysical charm of the architectural, Or structural, style may be unquestionable itis style that can Sls pive rice to metaphysical difcalties. For what right have Wwe to assume that there can be any “knowledge” (inthis archi- tectural sense) at al? Why should human beings imagine that they have the power to construct an Eternal City of well attested truths; or even that such a “ety” is really conceivable? Considered at a human activity, we know about “arguing” From firsthand experience; considered as independently existing {intellectual entities, divorced from that activity, “arguments” tre problematic, hypothetical, open to skeptical questioning. Perhaps this initial, threefold abstraction can be justfed—at any rate, on appropriate conditions—pechaps it cannot, As we tplore the consequences of the structural approgch here, we Ist be careful not to overlook a deeper question: “To what extent, on what conditions, and in what types of case (any) can the results of our experience, thought, and argument really be frdered in permanent strictures of knowiedge, {> that our beliefs can be thought of as anchored back to secure foundations, and ‘ur reasons as providing us with safe footing fon firm ground? In point of history (we shall find) itis no accident that the formal approach to philosophy has repeatedly led philosophers ott the end of the day-—into skepticism; that is, the doctrine that no reasons, in practice, can ever quite meet the strictest Standards, This fact is simply a sga that a formal analysis can tell at best, only part of the story about knowledge, belief, tnd “teasons”; while other parts of the story can be made intligible only by returning to the attack, subsequently, from ther directions also. One thing at least is clear enough. SomeFavored fields of experience, together with their associated argu ‘ments, lend themselves to structural analysis much better than others. If the City of Truth is to be built at all, may thus turn out to accommodate only these favored, “well-fting” Jhuman activities and enterprises: while other, more “il-iting” ‘ones elude satisfactory description in architectural term. ‘Any resulting system of philosophy will, naturally enough, tend to focus attention on the wellftting Belds of experience at the expense of the others. The more the products of our in- tellectual activities appear “objective” and “certain”, by struc- tural standards, the more seriously philosophers will tend to take those activities. And certainly, if we do think of “human Knowledge” as an enduring architectural construction—whether asa single vast ziggurat or skyscraper, or a8 a growing accu ‘mulation of separate bulldings—that image will impoce one specific division of labor between science and philosophy and place corresponding limits on the philosopher's work and con ‘cerns; for the intellectual value of human reasoning and belief ‘wil in that event, spring entirely from the contribstion they make to the permanent Edifice of Knowledge. Beliefs that are collective, permanent, and general will be accepted as “solid” fr “objective”. By contract, views and attitudes that are per- sonal, transitory, and particular will appear “merely subjective” and this will tend to be so, even in cases (eg. aesthetic prefer- fences) over which we are accustomed to giving highly articulate ‘True, abjective knowledge—public, permanent, and general Iknovledge—will then be embodied in clear, well-founded and certain, or “scientific”, priniples—principles that are applicable to the relevant Kinds of experiences and phenomena by any rational thinker, at all places and al times, The different sciences may be more or less compartmentaized, concerning themselves (co to say) with different “apartments” or “rites” in the entice “edifice” of Human Knowledge; but they will all be the out- come of a common and universal Human Reason, engaged in adding extra wings to the existing structure, By contrast, the ‘business of philosophy will be to consider, not any separate Philosophy and the City of Truth “9 ‘omer of the resulting structure, but rather how all ts diferent parts are connected together—what gives them their inherent Erength, and hove they are anchored back to their foundations. ‘On this view, then, the philosopher has tiple task, His responsibilities are, fist, to spell out the basic assumptions cur- tent in different areas of thought and belief; next, to make explicit the structures of argument from waich diferent kinds (of belief crave their strength; and finally, to criticize the result- fing azgumente by universal, public standards ‘What exactly are those public and universal standards? The answer to that question shows clearly the central thrust of the architectural metaphor, and of the formal approach to philoso- phy with which ts associated: We “know” something (in the fll and strict sense of the term) i-and-only-if we have a twel-founded belief in it; our belief in itis well- founded if-and-only-if we can produce good reasons in its support; and our reasons are really "good" (by the ssictet philosophical standards) ifand-only i we can produce a “conclusive”, ‘or formally-oulid argument, linking that be- lef back to an unchallenged (ond perferably unchallengeabe) starting pont.7 Reasons and Inferences men EE ee a perenne ha) Ska Speman ae wiiateinaetetianmner Baan Yara ieee at srg. ue ante Ue tht We tes ons bis a STOP. SN weeny pe te Le Sifs Some Eat te bates et Fbebes pene osken batten Tig echoed coca pole Garena atsS Reasons and Inferences a tics or twitches than hunger does is stil a problem, not for philosophical analysis, but for psychological theory.) The point Bt iseue will become truly “philosophical” only when we move to 2 completely general level—when we ask, for example, how it is that any overt, public signs or utterances can serve as indicators of people's inner, pereonal states of mind a all ‘Our first general style of ‘philosophical answer (a5 we have sen) links the notion of “good reasons” to that of “valid well-founded inferences”, According to this thesis, while ali Kinds of belie rest on correspondingly different sorts of reasons, the relations between ous beliefs and those supportive reasons will have certain quite general features, of a formal kind; and itis these formal features that make the reasons philosophically acceptable or unacceptable, that s, that make them by philo- Sophical standards “good” reasons for the belief in question. Specifically, reasons willbe “good”, on this view, provided that they can lead ue back, by a sequence of formally valid deduce tive steps, from the belle! in question to. an unquestionable tasting point. “This philovophical requirement i, of courte, a theoretical ideal rather than a practical reality. Out colloguial ways of talking About beliefe—whether psychological, astronomical, or aesthetic (say), whether in everyday life or in professional debate—are tnderstandably governed as much by pragmatic uty and teconomy as by logic] exactitude or analytical literal-mindedness. So, those colloguial modes of speech often leave unstated the ‘more general features that are important to philosophers; and before we ean pase any sure judgment, from the standpoint of formal philosophy, on the adequacy of our reasons for bel ing what we do, these missing features will have to be restored. ‘We must begin (that i) by “reconstructing”, in explicit form, the chains of assumptions and inferences implicitly involved in the particular examples being considered: we must identify lany generally acknowledged traths that are available to serve fsa final baste ("solid ground”) for the beliefs in question; tnd we must spell out openly any unspoken assumptions (“con- Cealed struts and beams") that aze required in order to fl outthose parts of the argument that would normally, thet is, colloquially, go without saying From the time of Plato to the present day a substantial past of European philosophy has been concemed with one of an- other of those two tasks: either with reconstructing. the “grounds” and “arguments” implicitly required to “support” ‘our different kinds of beliefs or with considering in what re- spects, and on what conditions, the ultimate foundations of those belies can be recognized as “solid”, and the associated inferences as “strong”. ‘A comprehensive textbook of philosophy written from this standpoint would take all the major flds of knowledge and cexperience—scientific, artistic, and religious individual, collec- tive, and interpersonal—and it would show historically how, in each of these main felds, philosophers of different schools hhave tackled the twin problems of solid foundations” and “valid forms of inference’. The outcome could be a valuable survey of the development and current state of the philosophical ‘enterprise, but it would be lengthy and laborious, and it would g0 far beyond our present tertory” Our ovn task is a more limited one. Rather than survey the whole area of formal phi losophy itself, we have chosen to map only the approaches to that area In this way our hope is, first, {0 indicate how the subsequent philosophical debates got started; second, what fields we can in general expect the formal approach to encom- pass most satisfactorily; and finally, at what points the obstacles toa purely formal analysis are probably insuperable. From the formal point of view, then, our “reasons” for be- Iieving what we do may be good oz bad in two independent ways: "tt would embrace, for istanc, sch vale tapes as Rend Dect on the avoms of phi Joba Locke on scondary gules, David ase on Induction John Vsdom on other minds, Benedetto Croce nd RC. Calling, wood om the principles of at Schopetaser on the wil, Maurice Mela Fonty on ou understending of spatial rlatons, and Ledwig Witgestesn onthe secigiton of eymbele Reasons and Inferences 93 either the grounds they provide to suport our beliefs may be fn sha othe nfreces linking hone ound othe bees tury be song/vea or bo, If we copier some sample Telefe fom diferent fede of experience and sk what an Sone fo give tov belts the sequred "strstr andre rings e shall be able fo indicate what diferent Kinds of Staarephal dele aren cach ase—whther ti source Bone round, o inthe connec, in Boh, We hall Took at four groupe of exes tar Problems ont “To bepn withthe least problematic examples: There ae cases sn which ip rensonably lear both what the grounds for ost Tulcea (hat we have fo go on) and how these grounds ae Connected fo the conlasons thw we get from the ground (0 the bole). About these cases we cm as: How solid are the grounds? And how firmly do the {inferences connect them tothe belifs in question? iow far, in particular, can our grounds in sch ‘cases achieve the Platonic ideal of unchallenge- ability, or our inferences achieve the ideal of formal, quas-geometrial validity? Next, we may lok at some example in which the character of the arguments/inferences/connections is clear enough, but in fohich the exact nature ofthe fandamental grounds is a matter of bate. About these cases we must ask, in addition: ‘What kind of secure foundations do these beliefs realy possess? Is it really possible to ground them firmly in our experience, orin other less problematic beliefs, and 4 90, how? case 3 In other cases again, i maybe renonaby lear what grounds wwe have for ou bli, yet may be hard to explaln peceely how those grounds lend proper support to our belief. Abost {hs third group of cases, we fae the question Given the grounds we have, what arguments oF inferences, if any, can we wee to connect ‘them with the corresponding beliefs? And what kind of strength or validity can be claimed for those connections? case 4 Finally, there are cases in which both the grounds and the connections are problematic; that is, situations in which—on 4 practical level—we claim “good reasons” for believing what wwe do, yet in which—on a philosophical level_we have no clear way elther of explaining our precise grounds or of dem- ‘onstrating how those grounds that we have validly support the beliefs concerned. About this last group of cases, we must ask the more basic question: Can there, after all, be legtimate beliefs for ‘which we have quite adequate pragmatic reasons, even though we cannot specify either philo- sophically acceptable grounds or formally valid Inferences in their support? And, if so, what docs this conclusion imply about the formal, ‘quasl-geometrical approach to philosophy? Reasons and Inferences 9s Examples case 1 (4) Predictive Arguments. Not surprisingly, the least dificult tases are to be found in fields of experience where formal, Imathematical procedures have been most consciously and de- Iiberately applied. Consider, for instance, our “reasons” for thinking that a total eclipse of the moon willbe visible in New York City on such-and-such an evening, between such-and-such times; of for believing that it will be high tide at Rockland, Maine, at 3:38 rat, this afternoon. Beliefs of this kind, con: cerned with recurrent events lke tides and eclipses, are arived St by applying appropriate forecasting techniques or formulas. ‘These formulas and techniques have been developed by accu- milating, accurate records of previous occurrences, analyzing the different periodic processes or componeats involved in those past recurrences, and. projecting forward (extrapolating) the Fesulting “time series” into the Fature “This is an essentially arithmetical procedure, which does not involve any scientific understanding. of the mechanisms in- wolved! A¥ the same time, it is a strictly formal procedure: that is to say, provided we have adequate records of previous tclipses or tides, the task of predicting further ones over a ‘eotonable period of time into the Future is simple matter of Calculation. Using the eatler records as “grounds”, or “data”, Snd the established arithmetical formulas as “principles of in- ference”, we can artive by way of “conclusion” ata single, un- Ambiguous prediction. The resulting calculation will lad us to Ae a mater of Iisory, the at of predicting eclipses and stoilar lela osttenes ws teady yell developed, long efore people had SeyrTSbatin de of what sorts of things stars and planets were, fr less TELSSENMy tat fees were soponsble for thet motions, Tia see SEND ty wore vecodad snd clalted Tog before the 1480, cae mee lMlsuceey theory wor advanced fo expan how these96 ‘mir rumtosortie as cronertR conclude that this afternoon's high tide will be at 3:28 rat, not 2:05 rat. or 6:50 rac; of, alternatively, that the eclipse will take place tonight rather than tomorrow or the day after, CCofrespondingly, if the eclipse or high tide does not take place as predicted, we are entitled to ask what went vrong with bur calculation, Were our data (grounds) insufclently full or precise? Was our formula (principle of inference) incorect, i completely applicable, or insufficiently detailed? Or what else Jed us astray? Ae with any other formally valid argument, we cannot accept that the conclusion isin fact false without con- ceding that either the inital data or the procedure of inference employed was in some respect at fault. ‘We can present the point in structural terms. The “founds- tione” on which such a tdal or astronomical prediction finally rests are provided by the accumulated records or observations of prior high and low tides, lana eclipses, and so forth; while the “framework” of formal argument rigidly linking back the prediction to those foundation is provided by general formulas ‘embodying the recurrent cycles of patterns found in thoee rec- rds, For practical purpose, the accumulated observations can nor- anally be accepted ae unchallenged, even if not unchallengesble. TE some of them have in fact been recorded or transmitted in- correcily, that may not seriously matter. Like a few soft spots in otherwise firm ground, these doubtful records need not have rach weight put on them, and provided there are comparatively {ew of them, the foundations of our subsequent inferences will not be seriously weakened. By contrast, the status of the gen- tral formulas, of regularities, om which our calculations. are bated it very far from unchallengeable. Rather than making any claim to being final or absolutely certain—as the theorems Of geometry did for so long—they are always subject to refine iment and revision We rely on such formulas for the future because and only because we have found them to fit the records hitherto—that [s the Force ofthe teem “extrapolating”. Yet we should certainly Reasons and Inferences 7” estate to him anything Uke the same neces for them that tre tight in the geome cave. Although we might set cut tS prove say tha the thee angles of 2 tangle mast add up to Teo degrees, without appealing to actual mesvuremens oF Cheervatons, we wou hatdly sxpet Yo profoce a smarly “pure” argument igpring all he acta empire records— proving 6) tn slr pes mat confor fo + tne ear cycle Yat from such syle seeming in any way sl iden the fact that they Fc inthe Stronomia 6 al corde at all guts further explanation from sstonomy or We by represent thee eltonsipsdlagammaticly (ee Figute 1). Provided the records are generally acute, and the fotmulae have proved sufient fo forecast al changes 3¢- Figure 1* {HE mimosornnR AS croMeTER curately on former occsions, we have evident “easons”—in fac all the “ease” the naar ofthe case equtes fo suppore that this next forecast (38 rat. tomorrow) willbe eelly accurate, At the same ime, the forecast remains « forease, tnd the supposition of its acuracy remains forthe tine being 4 supposition. Although the conclsion arrived at anombgee ous ("High de will be at 3:38 ra") it remain an etapa Similar considerations apply equally to many of the formal arguments discussed. by tadtonallogians under the te ‘vlogs’. Compare Figure 1, for instance, with Figure 2 In this case, too, the generalization on which the coelesion fomaty ea rel ao he bt fal eleva previnn experience, whereas the corrnt agement reappies,(etapo™ Ins) io as Inclade some Further futue eve Whe is CONCLUSION: —o} Sse era Tie, ‘ne ye abet Gam 00" \N Figue 2 Reasons and Inferences ° syllogism is also, on the surface, as formally valid an argument {5 any other, it conclusion too remains a supposition—s sup- position for which we have all the “reasons” the nature of the case demands, but a supposition nonetheless. Tt may be asked at this point: Could we not do better than this? Could we not, somehow or other, replace the purely ob- servational (or empirical) formulas and generalizations relied fon in such exces by more firmly based principles, reflecting a {genuine insight into processes and mechanisms necessarily in- volved in life and death eclipses, and/or tidal rises and falls? [And if we did that, might we not put ourselves in a postion to say, not merely that tomorzow afternoon's high tide will in fact be at 3:38 vat, oF that Socrates will in fact die, but that the events in question mst happen as predicted? Indeed, is it not precisely the business of a valid scientific theory to provide just this kind of additional understanding and certainty? (2) Scintife Theories. From the very begining of tee tn tus, thee are among the cena questions that philosophers Tove aked shout team and ote of rata stone, And Cerny, so far ab the words fo, we do often find sles prepared fo go beyond saying fo Sond wil i fact Eojpen in sudeand-ouh «way to the strongersounding con clei, "Sond so mst happen that way.” Hor ineances Galle feo didnot only chim to have dacovered, observational hat Soules aceerating niormly from rest alge tend fo ave four times at fr after two seconds as they did afler one Fates, he set out t prove-—aso mater of theory—that oy rif aceeating body rast do jst that His theory of tation (Ge Kinematis) vet thu a stil formal analyse of the eltons beeen the concep, "velocity" and "acl: tw “ltance ade, oma pat oth Ea’ analysis of the rlaons betwen the concept, “ne, "point, “angle”, “Sad and the ree Subsequent ofcourse allo was able to apply tha formal analyse bepetnetly, a way that threw gt on cera sexed and"ongstanding, emputeal problems in piysics andf 100 "HE FiMLosoPnER AG CEOMETER astronomy. Thus, for example, he showed that heavy bodies falling freely neat the surface of the earth do in fact accelerate uniformly; that i do in fact move in the ways that any “uni- formly. accelerating” body must conform to on. theoretical grounds. Given Galileo's theorems, this implied that a free- falling body must travel four times as far in two seconde From rest ab it did in one; yet it implied this, not categorically and ‘unconditionally, but only on the assumption that—and to the extent that—freeflling bodies do in fact accelerate at 2 uni- form rate. (Roughly speaking, the term “Treefalling” is an empirical designation of bodies; the term "uniformly accelerat- Jing a theoretical designation. But this distinction is not a ard- and-fast one) The present case calls, accordingly, for a slightly more com- plex diagrammatic representation, as shown in Figure 3. How far might we present in similar ways other branches of scientific theory and thei applications? That question i stil under discussion among philorophers of scence. To make just ‘one historical comment Some of the most eminent eightoenth- century mathematicians and philosophers—notably, Euler and Kant—attempted to demonstrate that the whole of Newton's theoretical physics was also a formal, “necessary” science, like Galileo's Kinematics and Euclid’s geometry. What Euclid had done theoretically for “point” and “line”, etc, and Galileo for “velocity” and “acceleration”, Newton id (on this view) for “mass”, "force, and the rest; that is, he cecognized how they ‘must be defined and drew the appropriate mathematical con- clusions. Indeed at certain points in his writings Kant even seems tobe supposing that the inverse-squace law of gravitation can itself be regarded as yet one more theorem in a totally general and analytical system of cosmic dynamics. Just a8 any authentic trlangle must have angles totaling 180 degrees (he argued), so any truly intelligible picture of the universe must conform to the inverse-square law. Nowadays most physicists fe He RITE ‘enouNos: A would consider this comparizon more misleeding than Kepfuly # g stil Kane was right to draw attention both to the purely 8 x formal character of arguments within Newtonian dyanamics 3g § BB 5 Figure 3< sole onl pen of epg How an Ss“ Beso confident thatthe renting formal theorem apply 1 ‘ctl material bodies, such a the sun and the planes tet ws make appropriste subsists in our Galilean dae sam so as to reapply io tie Newtonian example. In place the question, “How do we know that freetaling beds Sceleratearformly2” we can consider the question, “How do sre now thatthe plante~Mercary, Vent Jupiter, ele-—are ‘moving, predominant, under he infvenceof the suns inverse= Square sracive fore?” Both questions wil ten be abot what £2 fact the cate, rather than aboot what mus be the cases and in both she crcl fame ist establish that certain ea Pilly designated object in the Newtonian case the Fain ‘sully dentable plants) answer ale to a paticuat theo” Fetal designation (eg, a8 bodies acted on by a single, cen- telly directed inversesgoare fre) See igure & So all three men—Galeo and Newton, quite as auch as Eacid~separated the general formal analysis (whether of gem etry oF kinematics or Gynamics) from is empires aplistions fo actual physical examples. Tt ws this in act that gave all thes theories the appearance of having sound Patni form Forthermore, the thee analyses wee interconnected, For ox ample, Gale's kinematics fook Eucld's geometry for granted and built wpon it by adding fresh, dynamical concepts, Yet How far, aba res, could ther Calleo or Newton bring physics up tothe level of “true intl grasp” demanded by Pit’ del of episteme? Fora hie they seemed to have done jos that. Kant for one, was very anxiovs to demonstrate how it was tht an empirical science, suchas physi, might achieve that Kind of certainty. Bt this happy sate of fas di not endure Dynanice nnd Kinematic ‘ght be a8 good, in ths respect, as Eucli’s geometry, but even geomety Y nounes: A seein ara t Wwe shall sex) tuned out tobe less “unique” and “certain” than Plato had supposed. Ths, even Galeo and Newton gave us something les than epsteme, that is, absolutly certain under. # z 4 standing, based on intelectoal intuition, of what “aceleration” 35 a 3 and “force” must necessarily be. a a & Figwe 4108 ‘ru0 mintosorne as oronerx To sum wp: Two preliminary pont should be made shout thee ist examples. On the obe hand, the adequacy of any formals for making, for example alot astonomzal predic Won nt in ea philsopil mater, The rat f fore Cauing dal movement ot ftronomical phenomena fas i fn procedures snd standards of jugmenty the relevant fre mlat have been rcned and improved. inthe Hight of exper ce bythe professional ety ivalved* On the other hand, the acguments by which these formas yi spctc forecast tre im ptt of loge, formal deductions. This ing the se, Shee "validity" not normally Jn doubt, any remining ernie about ther conclusions affect ater, the formas ‘or oher generalizations tvaled In the sume may i the u- Neca th of the proposition tat all men are moral cold be guaranteed absolutely, the conson that Socater—being 'mamcis mortal would be established. with “geometial” Cersiny. As things stand, we femain short ofthe geomet dual even in thi ase Uecatoe we lack any sboltesearance of that univer eth, Our confidence thatthe generalization wl continue to holdin fete Instancer ike on confidence Inthe cuvette or estonamielformalaelnvlve, in the tue of the case, a resiual supposition. Tost beause any sich argument est deductive in form, slave the same “event” al Ser formally vad dpuments; that iif the conlason proves fale, someing Imus be wrong ether with the “grounde” or withthe “aking fenerliation® that connects the grands Yo the predict Conclusion. It nevertheless fall short of the Phonic ideal of Knowledge, beeure the event linking, generalization ae sever af. cidenly cores I for this reason at palcophers sometimes speak of arguments of this Kind as “hype Gctve” argument: Ene are "dedctive” becuse they can Be set out int putely formal manner, and “hypothe” be 1 an, of couse, bea phllosopher’s basins to consider, in gate gene teal terms ast ow such experiance gives the revling format eve “pragmatic” val ‘Reasons and Inferences 105 ‘cause the crucial generalizations (eg, All men are mortal) are Bill, to however slight a degree, subject to revision and correc tion, rather than self-evidently true, By contrac, the theorems of geometry-the chief exemplars of true knowledge in Plato's feyes—might be labeled ae “eategoricodeductive"; they were ‘supposedly connected by formally valid arguments back to slf- fevident axioms and principles, and thelr truth accordingly had fan absolute guarantee case 2 [As our second group of examples, we may consider cases in which dificulties arise over the grounds for our beliefs, This ‘an happen in two distinct ways. (2) In one class of situations, ‘we think, believe, and/or claim to know things for which we could give reasons cast in formally valid arguments, though we might find it hard, nonetheless, to demonstrate that those “grounds” are relevant and sulfclent to establish the corre- sponding conclusions past question. (2) In the other class of situations, by contrast, we think, believe, and/or claim to know things for which we would scarcely know hove to produce grounds” at all Moral Arguments. Moral and ethical discussions can provide many examples of the fist type. We may say with confidence, for instance. "im really ought to give up gambling.” Yet when wwe are asked why we believe this, we may not find it easy to make the grounds for our belief wholly clear and convincing. If the point at issue were legal rather than moral, matters might bbe much easier. Suppose that Jim had gone into bankruptcy 25 2 result of his gambling and was under a court order to desist; In that case, we could establish, formally and conclusively Jim ought to give up gambling-in point of law—since it would bea clear breach ofthe court order for him to continue. Ethical questions differ from questions of law, however, in jast this respect, We possess nelther such clear and decisive nor such sgreed-upon ways of demonstrating what is ethically good/bad- ve mene ce PE or morally right/wrong as we do of demonstrating what is within/contray to the law. About the form of our moral arguments, there is nothing like the same difficulty. Suppose that the particular considera™ tion that most strongly leads us to disapprove of Jims gambling is the fact that he spends on bets money he really needs for ood for his childten. Once this is made clear, it Is not hard to construct an argument connecting that ground to the re-