The ancient commentators grouped together several of Aristotle's treatises under the title Organon ("Instrument") and regarded them as comprising his logical works: 1. Categories . On Interpretation !. Prior Analytics ". Posterior Analytics #. Topics $. On Sophistical Refutations In fact% the title Organon reflects a much later controvers& a'out whether logic is a part of philosoph& (as the (toics maintained) or merel& a tool used '& philosoph& (as the later )eripatetics thought)* calling the logical works "The Instrument" is a wa& of taking sides on this point. Aristotle himself never uses this term% nor does he give much indication that these particular treatises form some kind of group% though there are fre+uent cross,references 'etween the Topics and the Analytics. -n the other hand% Aristotle treats the Prior and Posterior Analytics as one work% and On Sophistical Refutations is a final section% or an appendi.% to the Topics). To these works should 'e added the Rhetoric% which e.plicitl& declares its reliance on the Topics. 3. The Subject of Logic: "Syllogisms" All Aristotle's logic revolves around one notion: the deductio (sullogismos). A thorough e.planation of what a deduction is% and what the& are composed of% will necessaril& lead us through the whole of his theor&. /hat% then% is a deduction0 Aristotle sa&s: A deduction is speech (logos) in which% certain things having 'een supposed% something different from those supposed results of necessit& 'ecause of their 'eing so. (Prior Analytics I.% "'11,2) 3ach of the "things supposed" is a !remise (protasis) of the argument% and what "results of necessit&" is the coclusio (sumperasma). The core of this definition is the notion of "resulting of necessit&" (ex ananks sumbainein). This corresponds to a modern notion of logical conse+uence: 4 results of necessit& from 5 and 6 if it would 'e impossi'le for 4 to 'e false when 5 and 6 are true. /e could therefore take this to 'e a general definition of "valid argument". 3." #ductio ad $eductio 7eductions are one of two species of argument recogni8ed '& Aristotle. The other species is iductio (epagg). 9e has far less to sa& a'out this than deduction% doing little more than characteri8e it as "argument from the particular to the universal". 9owever% induction (or something ver& much like it) pla&s a crucial role in the theor& of scientific knowledge in the Posterior Analytics: it is induction% or at an& rate a cognitive process that moves from particulars to their generali8ations% that is the 'asis of knowledge of the indemonstra'le first principles of sciences. 3.2 Aristotelia $eductios ad %oder &alid Argumets 7espite its wide generalit&% Aristotle's definition of deduction is not a precise match for a modern definition of validit&. (ome of the differences ma& have important conse+uences: 1. Aristotle e.plicitl& sa&s that what results of necessit& must 'e different from what is supposed. This would rule out arguments in which the conclusion is identical to one of the premises. :odern notions of validit& regard such arguments as valid% though triviall& so. . The plural "certain things having 'een supposed" was taken '& some ancient commentators to rule out arguments with onl& one premise. !. The force of the +ualification "'ecause of their 'eing so" has sometimes 'een seen as ruling out arguments in which the conclusion is not ;relevant< to the premises% e.g.% arguments in which the premises are inconsistent% arguments with conclusions that would follow from an& premises whatsoever% or arguments with superfluous premises. -f these three possi'le restrictions% the most interesting would 'e the third. This could 'e (and has 'een) interpreted as committing Aristotle to something like a relevance logic. In fact% there are passages that appear to confirm this. 9owever% this is too comple. a matter to discuss here. 9owever the definition is interpreted% it is clear that Aristotle does not mean to restrict it onl& to a su'set of the valid arguments. This is wh& I have translated sullogismos with ;deduction< rather than its 3nglish cognate. In modern usage% ;s&llogism< means an argument of a ver& specific form. :oreover% modern usage distinguishes 'etween valid s&llogisms (the conclusions of which follow from their premises) and invalid s&llogisms (the conclusions of which do not follow from their premises). The second of these is inconsistent with Aristotle's use: since he defines a sullogismos as an argument in which the conclusion results of necessit& from the premises% "invalid sullogismos" is a contradiction in terms. The first is also at least highl& misleading% since Aristotle does not appear to think that the sullogismoi are simpl& an interesting su'set of the valid arguments. :oreover (see 'elow)% Aristotle e.pends great efforts to argue that ever& valid argument% in a 'road sense% can 'e "reduced" to an argument% or series of arguments% in something like one of the forms traditionall& called a s&llogism. If we translate sullogismos as "s&llogism% "% this 'ecomes the trivial claim "3ver& s&llogism is a s&llogism"% '. (remises: The Structures of Assertios (&llogisms are structures of sentences each of which can meaningfull& 'e called true or false: assertios (apophanseis)% in Aristotle's terminolog&. According to Aristotle% ever& such sentence must have the same structure: it must contain a subject (hupokeimenon) and a !redicate and must either affirm or den& the predicate of the su'=ect. Thus% ever& assertion is either the affirmatio kataphasis or the deial (apophasis) of a single predicate of a single su'=ect. In On Interpretation% Aristotle argues that a single assertion must alwa&s either affirm or den& a single predicate of a single su'=ect. Thus% he does not recogni8e sentential compounds% such as con=unctions and dis=unctions% as single assertions. This appears to 'e a deli'erate choice on his part: he argues% for instance% that a con=unction is simpl& a collection of assertions% with no more intrinsic unit& than the se+uence of sentences in a length& account (e.g. the entire Ilia% to take Aristotle's own e.ample). (ince he also treats denials as one of the two 'asic species of assertion% he does not view negations as sentential compounds. 9is treatment of conditional sentences and dis=unctions is more difficult to appraise% 'ut it is at an& rate clear that Aristotle made no efforts to develop a sentential logic. (ome of the conse+uences of this for his theor& of demonstration are important. '." Terms (u'=ects and predicates of assertions are terms. A term (horos) can 'e either individual% e.g. Socrates% Plato or universal% e.g. human% horse% animal% !hite. (u'=ects ma& 'e either individual or universal% 'ut predicates can onl& 'e universals: Socrates is human% Plato is not a horse% horses are animals% humans are not horses. The word ui)ersal (katholou) appears to 'e an Aristotelian coinage. >iterall&% it means "of a whole"* its opposite is therefore "of a particular" (kath" hekaston). ?niversal terms are those which can properl& serve as predicates% while particular terms are those which cannot. This distinction is not simpl& a matter of grammatical function. /e can readil& enough construct a sentence with "(ocrates" as its grammatical predicate: "The person sitting down is (ocrates". Aristotle% however% does not consider this a genuine predication. 9e calls it instead a merel& accidetal or icidetal (kata sumbebkos) predication. (uch sentences are% for him% dependent for their truth values on other genuine predications (in this case% "(ocrates is sitting down"). @onse+uentl&% predication for Aristotle is as much a matter of metaph&sics as a matter of grammar. The reason that the term Socrates is an individual term and not a universal is that the entit& which it designates is an individual% not a universal. /hat makes !hite and human universal terms is that the& designate universals. Aurther discussion of these issues can 'e found in the entr& on Aristotle's metaph&sics. '.2 Affirmatios* $eials* ad +otradictios Aristotle takes some pains in On Interpretation to argue that to ever& affirmation there corresponds e.actl& one denial such that that denial denies e.actl& what that affirmation affirms. The pair consisting of an affirmation and its corresponding denial is a cotradictio (antiphasis). In general% Aristotle holds% e.actl& one mem'er of an& contradiction is true and one false: the& cannot 'oth 'e true% and the& cannot 'oth 'e false. 9owever% he appears to make an e.ception for propositions a'out future events% though interpreters have de'ated e.tensivel& what this e.ception might 'e (see further discussion 'elow). The principle that contradictories cannot 'oth 'e true has fundamental importance in Aristotle's metaph&sics (see further discussion 'elow). '.3 All* Some* ad ,oe -ne ma=or difference 'etween Aristotle's understanding of predication and modern (i.e.% post,Aregean) logic is that Aristotle treats individual predications and general predications as similar in logical form: he gives the same anal&sis to "(ocrates is an animal" and "9umans are animals". 9owever% he notes that when the su'=ect is a universal% predication takes on two forms: it can 'e either ui)ersal or !articular. These e.pressions are parallel to those with which Aristotle distinguishes universal and particular terms% and Aristotle is aware of that% e.plicitl& distinguishing 'etween a term 'eing a universal and a term 'eing universall& predicated of another. /hatever is affirmed or denied of a universal su'=ect ma& 'e affirmed or denied of it it ui)ersally (katholou or "of all"% kata pantos)% i !art (kata meros% en merei)% or idefiitely (aihoristos). Affirmatios $eials -i)ersal ) affirmed of all of ( 3ver& ( is )% All ( is (are) ) ) denied of all of ( Bo ( is ) (articular ) affirmed of some of ( (ome ( is (are) ) ) denied of some of ( (ome ( is not )% Bot ever& ( is ) #defiite ) affirmed of ( ( is ) ) denied of ( ( is not ) '.3." The "S.uare of /!!ositio" In On Interpretation% Aristotle spells out the relationships of contradiction for sentences with universal su'=ects as follows: Affirmatio $eial -i)ersal 3ver& A is C Bo A is C -i)ersal (ome A is C Bot ever& A is C (imple as it appears% this ta'le raises important difficulties of interpretation (for a thorough discussion% see the entr& on the s+uare of opposition). In the Prior Analytics% Aristotle adopts a somewhat artificial wa& of e.pressing predications: instead of sa&ing "4 is predicated of 5" he sa&s "4 'elongs (huparchei) to 5". This should reall& 'e regarded as a technical e.pression. The ver' huparchein usuall& means either "'egin" or "e.ist% 'e present"% and Aristotle's usage appears to 'e a development of this latter use. '.3.2 Some +o)eiet Abbre)iatios Aor clarit& and 'revit&% I will use the following semi,traditional a''reviations for Aristotelian categorical sentences (note that the predicate term comes first and the su'=ect term secon): Abbre)iatio Setece Aa' a 'elongs to all ' (3ver& ' is a) 3a' a 'elongs to no ' (Bo ' is a) Ia' a 'elongs to some ' ((ome ' is a) -a' a does not 'elong to all ' ((ome ' is not a) 0. The Syllogistic Aristotle's most famous achievement as logician is his theor& of inference% traditionall& called the syllogistic (though not '& Aristotle). That theor& is in fact the theor& of inferences of a ver& specific sort: inferences with two premises% each of which is a categorical sentence% having e.actl& one term in common% and having as conclusion a categorical sentence the terms of which are =ust those two terms not shared '& the premises. Aristotle calls the term shared '& the premises the middle term (meson) and each of the other two terms in the premises an e1treme (akron). The middle term must 'e either su'=ect or predicate of each premise% and this can occur in three wa&s: the middle term can 'e the su'=ect of one premise and the predicate of the other% the predicate of 'oth premises% or the su'=ect of 'oth premises. Aristotle refers to these term arrangements as figures (schmata): 0." The 2igures 2irst 2igure Secod 2igure Third 2igure (redicate Subject (redicate Subject (redicate Subject (remise a ' a ' a c (remise ' c a c ' c +oclusio a c ' c a ' Aristotle calls the term which is the predicate of the conclusion the major term and the term which is the su'=ect of the conclusion the mior term. The premise containing the ma=or term is the major !remise% and the premise containing the minor term is the mior !remise. Aristotle's procedure is then a s&stematic investigation of the possi'le com'inations of premises in each of the three figures. Aor each com'ination% he seeks either to demonstrate that some conclusion necessaril& follows or to demonstrate that no conclusion follows. The results he states are e.actl& correct. 0.2 %ethods of (roof: +o)ersio ad 3eductio Aristotle shows each valid form to 'e valid '& showing how to construct a deduction of its conclusion from its premises. These deductions% in turn% can take one of two forms: direct or !robati)e (eiktikos) deductions and deductions through the im!ossible (ia to aunaton). A direct deduction is a series of steps leading from the premises to the conclusion% each of which is either a co)ersio of a previous step or an inference from two previous steps rel&ing on a first,figure deduction. @onversion% in turn% is inferring from a proposition another which has the su'=ect and predicate interchanged. (pecificall&% Aristotle argues that three such conversions are sound: 3a' D 3'a Ia' D I'a Aa' D I'a 9e undertakes to =ustif& these in An# Pr# I.. Arom a modern standpoint% the third is sometimes regarded with suspicion. ?sing it we can get Some monsters are chimeras from the apparentl& true All chimeras are monsters* 'ut the former is often construed as impl&ing in turn There is something !hich is a monster an a chimera% and thus that there are monsters and there are chimeras. In fact% this simpl& points up something a'out Aristotle's s&stem: Aristotle in effect supposes that all terms in s&llogisms are non,empt&. (Aor further discussion of this point% see the entr& on the s+uare of opposition). As an e.ample of the procedure% we ma& take Aristotle's proof of Camestres. 9e sa&s: If : 'elongs to ever& B 'ut to no 4% then neither will B 'elong to an& 4. Aor if : 'elongs to no 4% then neither does 4 'elong to an& :* 'ut : 'elonged to ever& B* therefore% 4 will 'elong to no B (for the first figure has come a'out). And since the privative converts% neither will B 'elong to an& 4. (An# Pr# I.#% EaF,1) Arom this te.t% we can e.tract an e.act formal proof% as follows: Ste! 4ustificatio Aristotle's Te1t 1. :aB If $ belongs to e%ery & . :e4 but to no '( To prove: then neither !ill & belong to any '# Be4 !. :e4 (% premise) )or if $ belongs to no '( ". 4e: (!% conversion of e) then neither oes ' belong to any $* #. :aB (1% premise) but $ belonge to e%ery &* $. 4eB ("% #% Celarent) therefore( ' !ill belong to no & +for the first figure has come about,# E. Be4 ($% conversion of e) An since the pri%ati%e con%erts( neither !ill & belong to any '# 0.3 %ethods of $is!roof: +outere1am!les ad Terms Aristotle proves invalidit& '& constructing countere.amples. This is ver& much in the spirit of modern logical theor&: all that it takes to show that a certain form is invalid is a single instance of that form with true premises and a false conclusion. 9owever% Aristotle states his results not '& sa&ing that certain premise,conclusion com'inations are invalid 'ut '& sa&ing that certain premise pairs do not "s&llogi8e": that is% that% given the pair in +uestion% e.amples can 'e constructed in which premises of that form are true and a conclusion of an& of the four possi'le forms is false. /hen possi'le% he does this '& a clever and economical method: he gives two triplets of terms% one of which makes the premises true and a universal affirmative "conclusion" true% and the other of which makes the premises true and a universal negative "conclusion" true. The first is a countere.ample for an argument with either an 3 or an - conclusion% and the second is a countere.ample for an argument with either an A or an I conclusion. 0.' The $eductios i the 2igures 5"%oods"6 In Prior Analytics I.",$% Aristotle shows that the premise com'inations given in the following ta'le &ield deductions and that all other premise com'inations fail to &ield a deduction. In the terminolog& traditional since the middle ages% each of these com'inations is known as a mood (from >atin mous% "wa&"% which in turn is a translation of Greek tropos). Aristotle% however% does not use this e.pression and instead refers to "the arguments in the figures". In this ta'le% " " separates premises from conclusion* it ma& 'e read "therefore". The second column lists the medieval mnemonic name associated with the inference (these are still widel& used% and each is actuall& a mnemonic for Aristotle's proof of the mood in +uestion). The third column 'riefl& summari8es Aristotle's procedure for demonstrating the deduction. Table of the $eductios i the 2igures 2orm %emoic (roof Aa'% A'c Aac -arbara )erfect 3a'% A'c 3ac Celarent )erfect Aa'% I'c Iac .arii )erfect* also '& impossi'ilit&% from Camestres 3a'% I'c -ac )erio )erfect* also '& impossi'ilit&% from Cesare (3@-B7 AIG?H3 3a'% Aac 3'c Cesare (3a'% Aac)D(3'a% Aac) @el 3'c Aa'% 3ac 3'c Camestres (Aa'% 3ac)D(Aa'% 3ca)I(3ca% Aa') @el 3c'D3'c 3a'% Iac -'c )estino (3a'% Iac)D(3'a% Iac) Aer -'c Aa'% -ac -'c -aroco (Aa'% -ac JA'c) Car (Aac% -ac) Imp -'c T9IH7 AIG?H3 Aac% A'c Ia' .arapti (Aac% A'c)D(Aac% Ic') 7ar Ia' 3ac% A'c -a' )elapton (3ac% A'c)D(3ac% Ic') Aer -a' Iac% A'c Ia' .isamis (Iac% A'c)D(Ica% A'c)I(A'c% Ica) 7ar I'aDIa' Aac% I'c Ia' .atisi (Aac% I'c)D(Aac% Ic') 7ar Ia' -ac% A'c -a' -ocaro (-ac% JAa'% A'c) Car (Aac% -ac) Imp -a' 3ac% I'c -a' )erison (3ac% I'c)D(3ac% Ic') Aer -a' 0.0 %etatheoretical 3esults 9aving esta'lished which deductions in the figures are possi'le% Aristotle draws a num'er of metatheoretical conclusions% including: 1. Bo deduction has two negative premises . Bo deduction has two particular premises !. A deduction with an affirmative conclusion must have two affirmative premises ". A deduction with a negative conclusion must have one negative premise. #. A deduction with a universal conclusion must have two universal premises 9e also proves the following metatheorem: All deductions can 'e reduced to the two universal deductions in the first figure. 9is proof of this is elegant. Airst% he shows that the two particular deductions of the first figure can 'e reduced% '& proof through impossi'ilit&% to the universal deductions in the second figure: (.arii)K(Aa'% I'c% J3ac) @amestres (3'c% I'c) Imp Iac ()erio)K(3a'% I'c% JAac) @esare (3'c% I'c) Imp -ac 9e then o'serves that since he has alread& shown how to reduce all the particular deductions in the other figures e.cept Caroco and Cocardo to .arii and )erio% these deductions can thus 'e reduced to -arbara and Celarent. This proof is strikingl& similar 'oth in structure and in su'=ect to modern proofs of the redundanc& of a.ioms in a s&stem. :an& more metatheoretical results% some of them +uite sophisticated% are proved in Prior Analytics I."# and in Prior Analytics II. As noted 'elow% some of Aristotle's metatheoretical results are appealed to in the epistemological arguments of the Posterior Analytics. 0.7 Syllogisms 8ith %odalities Aristotle follows his treatment of "arguments in the figures" with a much longer% and much more pro'lematic% discussion of what happens to these figured arguments when we add the +ualifications "necessaril&" and "possi'l&" to their premises in various wa&s. In contrast to the s&llogistic itself (or% as commentators like to call it% the assertoric s&llogistic)% this moal s&llogistic appears to 'e much less satisfactor& and is certainl& far more difficult to interpret. 9ere% I onl& outline Aristotle's treatment of this su'=ect and note some of the principal points of interpretive controvers&. 0.7." The $efiitios of the %odalities :odern modal logic treats necessit& and possi'ilit& as interdefina'le: "necessaril& )" is e+uivalent to "not possi'l& not )"% and "possi'l& )" to "not necessaril& not )". Aristotle gives these same e+uivalences in On Interpretation. 9owever% in Prior Analytics% he makes a distinction 'etween two notions of possi'ilit&. -n the first% which he takes as his preferred notion% "possi'l& )" is e+uivalent to "not necessaril& ) and not necessaril& not )". 9e then acknowledges an alternative definition of possi'ilit& according to the modern e+uivalence% 'ut this pla&s onl& a secondar& role in his s&stem. 0.7.2 Aristotle's 9eeral A!!roach Aristotle 'uilds his treatment of modal s&llogisms on his account of non,modal (assertoric) s&llogisms: he works his wa& through the s&llogisms he has alread& proved and considers the conse+uences of adding a modal +ualification to one or 'oth premises. :ost often% then% the +uestions he e.plores have the form: "9ere is an assertoric s&llogism* if I add these modal +ualifications to the premises% then what modall& +ualified form of the conclusion (if an&) follows0". A premise can have one of three modalities: it can 'e necessar&% possi'le% or assertoric. Aristotle works through the com'inations of these in order: Two necessar& premises -ne necessar& and one assertoric premise Two possi'le premises -ne assertoric and one possi'le premise -ne necessar& and one possi'le premise Though he generall& considers onl& premise com'inations which s&llogi8e in their assertoric forms% he does sometimes e.tend this* similarl&% he sometimes considers conclusions in addition to those which would follow from purel& assertoric premises. (ince this is his procedure% it is convenient to descri'e modal s&llogisms in terms of the corresponding non,modal s&llogism plus a triplet of letters indicating the modalities of premises and conclusion: B I "necessar&"% ) I "possi'le"% A I "assertoric". Thus% "Car'ara BAB" would mean "The form -arbara with necessar& ma=or premise% assertoric minor premise% and necessar& conclusion". I use the letters "B" and ")" as prefi.es for premises as well* a premise with no prefi. is assertoric. Thus% -arbara BAB would 'e BAa'% A'c BAac. 0.7.3 %odal +o)ersios As in the case of assertoric s&llogisms% Aristotle makes use of conversion rules to prove validit&. The conversion rules for necessar& premises are e.actl& analogous to those for assertoric premises: B3a'DB3'a BIa'DBI'a BAa'DBI'a )ossi'le premises 'ehave differentl&% however. (ince he defines "possi'le" as "neither necessar& nor impossi'le"% it turns out that x is possibly ) entails% and is entailed '&% x is possibly not ). Aristotle generali8es this to the case of categorical sentences as follows: )Aa'D)3a' )3a'D)Aa' )Ia'D)-a' )-a'D)Ia' In addition% Aristotle uses the intermodal principle BDA: that is% a necessar& premise entails the corresponding assertoric one. 9owever% 'ecause of his definition of possi'ilit&% the principle AD) does not generall& hold: if it did% then BD) would hold% 'ut on his definition "necessaril& )" and "possi'l& )" are actuall& inconsistent ("possi'l& )" entails "possi'l& not )"). This leads to a further complication. The denial of "possi'l& )" for Aristotle is "either necessaril& ) or necessaril& not )". The denial of "necessaril& )" is still more difficult to e.press in terms of a com'ination of modalities: "either possi'l& ) (and thus possi'l& not )) or necessaril& not )" This is important 'ecause of Aristotle's proof procedures% which include proof through impossi'ilit&. If we give a proof through impossi'ilit& in which we assume a necessar& premise% then the conclusion we ultimatel& esta'lish is simpl& the denial of that necessar& premise% not a "possi'le" conclusion in Aristotle's sense. (uch propositions do occur in his s&stem% 'ut onl& in e.actl& this wa&% i.e.% as conclusions esta'lished '& proof through impossi'lit& from necessar& assumptions. (omewhat confusingl&% Aristotle calls such propositions "possi'le" 'ut immediatel& adds " not in the sense defined": in this sense% "possi'l& -a'" is simpl& the denial of "necessaril& Aa'". (uch propositions appear onl& as premises% never as conclusions. 0.7.' Syllogisms 8ith ,ecessary (remises Aristotle holds that an assertoric s&llogism remains valid if "necessaril&" is added to its premises and its conclusion: the modal pattern BBB is alwa&s valid. 9e does not treat this as a trivial conse+uence 'ut instead offers proofs* in all 'ut two cases% these are parallel to those offered for the assertoric case. The e.ceptions are -aroco and -ocaro% which he proved in the assertoric case through impossi'ilit&: attempting to use that method here would re+uire him to take the denial of a necessar& - proposition as h&pothesis% raising the complication noted a'ove% and he must resort to a different form of proof instead. 0.7.0 ,A:A, +ombiatios: The (roblem of the "T8o ;arbaras" ad /ther $ifficulties (ince a necessar& premise entails an assertoric premise% ever& AB or BA com'ination of premises will entail the corresponding AA pair% and thus the corresponding A conclusion. Thus% ABA and BAA s&llogisms are alwa&s valid. 9owever% Aristotle holds that some% 'ut not all% ABB and BAB com'inations are valid. (pecificall&% he accepts -arbara BAB 'ut re=ects -arbara ABB. Almost from Aristotle's own time% interpreters have found his reasons for this distinction o'scure% or unpersuasive% or 'oth. Theophrastus% for instance% adopted the simpler rule that the modalit& of the conclusion of a s&llogism was alwa&s the "weakest" modalit& found in either premise% where B is stronger than A and A is stronger than ) (and where ) pro'a'l& has to 'e defined as "not necessaril& not"). -ther difficulties follow from the pro'lem of the "Two Car'aras"% as it is often called% and it has often 'een maintained that the modal s&llogistic is inconsistent. This su'=ect +uickl& 'ecomes too comple. for summari8ing in this 'rief article. Aor further discussion% see Cecker% :c@all% )atterson% van Hi=en% (triker% Bortmann% Thom% and Thomason. 7. $emostratios ad $emostrati)e Scieces A demostratio (apoeixis) is "a deduction that produces knowledge". Aristotle's Posterior Analytics contains his account of demonstrations and their role in knowledge. Arom a modern perspective% we might think that this su'=ect moves outside of logic to epistemolog&. Arom Aristotle's perspective% however% the connection of the theor& of sullogismoi with the theor& of knowledge is especiall& close. 7." Aristotelia Scieces The su'=ect of the Posterior Analytics is epistm. This is one of several Greek words that can reasona'l& 'e translated "knowledge"% 'ut Aristotle is concerned onl& with knowledge of a certain t&pe (as will 'e e.plained 'elow). There is a long tradition of translating epistm in this technical sense as sciece% and I shall follow that tradition here. 9owever% readers should not 'e misled '& the use of that word. In particular% Aristotle's theor& of science cannot 'e considered a counterpart to modern philosoph& of science% at least not without su'stantial +ualifications. /e have scientific knowledge% according to Aristotle% when we know: the cause wh& the thing is% that it is the cause of this% and that this cannot 'e otherwise. (Posterior Analytics I.) This implies two strong conditions on what can 'e the o'=ect of scientific knowledge: 1. -nl& what is necessaril& the case can 'e known scientificall& . (cientific knowledge is knowledge of causes 9e then proceeds to consider what science so defined will consist in% 'eginning with the o'servation that at an& rate one form of science consists in the possession of a demostratio (apoeixis)% which he defines as a "scientific deduction": '& "scientific" (epistmonikon)% I mean that in virtue of possessing it% we have knowledge. The remainder of Posterior Analytics I is largel& concerned with two tasks: spelling out the nature of demonstration and demonstrative science and answering an important challenge to its ver& possi'ilit&. Aristotle first tells us that a demonstration is a deduction in which the premises are: 1. true . !rimary (prota) !. immediate (amesa% "without a middle") ". better ko8 or more familiar (gnrimtera) than the conclusion #. !rior to the conclusion $. causes (aitia) of the conclusion The interpretation of all these conditions e.cept the first has 'een the su'=ect of much controvers&. Aristotle clearl& thinks that science is knowledge of causes and that in a demonstration% knowledge of the premises is what 'rings a'out knowledge of the conclusion. The fourth condition shows that the knower of a demonstration must 'e in some 'etter epistemic condition towards them% and so modern interpreters often suppose that Aristotle has defined a kind of epistemic =ustification here. 9owever% as noted a'ove% Aristotle is defining a special variet& of knowledge. @omparisons with discussions of =ustification in modern epistemolog& ma& therefore 'e misleading. The same can 'e said of the terms "primar&"% "immediate" and "'etter known". :odern interpreters sometimes take "immediate" to mean "self,evident"* Aristotle does sa& that an immediate proposition is one "to which no other is prior"% 'ut (as I suggest in the ne.t section) the notion of priorit& involved is likel& a notion of logical priorit& that it is hard to detach from Aristotle's own logical theories. "Cetter known" has sometimes 'een interpreted simpl& as "previousl& known to the knower of the demonstration" (i.e.% alread& known in advance of the demonstration). 9owever% Aristotle e.plicitl& distinguishes 'etween what is "'etter known for us" with what is "'etter known in itself" or "in nature" and sa&s that he means the latter in his definition. In fact% he sa&s that the process of ac+uiring scientific knowledge is a process of changing what is 'etter known "for us"% until we arrive at that condition in which what is 'etter known in itself is also 'etter known for us. 7.2 The 3egress (roblem In Posterior Analytics I.% Aristotle considers two challenges to the possi'ilit& of science. -ne part& (du''ed the "agnostics" '& Lonathan Carnes) 'egan with the following two premises: 1. /hatever is scientificall& known must 'e demonstrated. . The premises of a demonstration must 'e scientificall& known. The& then argued that demonstration is impossi'le with the following dilemma: 1. If the premises of a demonstration are scientificall& known% then the& must 'e demonstrated. . The premises from which each premise are demonstrated must 'e scientificall& known. !. 3ither this process continues forever% creating an infinite regress of premises% or it comes to a stop at some point. ". If it continues forever% then there are no first premises from which the su'se+uent ones are demonstrated% and so nothing is demonstrated. #. -n the other hand% if it comes to a stop at some point% then the premises at which it comes to a stop are undemonstrated and therefore not scientificall& known* conse+uentl&% neither are an& of the others deduced from them. $. Therefore% nothing can 'e demonstrated. A second group accepted the agnostics' view that scientific knowledge comes onl& from demonstration 'ut re=ected their conclusion '& re=ecting the dilemma. Instead% the& maintained: 7emonstration "in a circle" is possi'le% so that it is possi'le for all premises also to 'e conclusions and therefore demonstrated. Aristotle does not give us much information a'out how circular demonstration was supposed to work% 'ut the most plausi'le interpretation would 'e supposing that at least for some set of fundamental principles% each principle could 'e deduced from the others. ((ome modern interpreters have compared this position to a coherence theor& of knowledge.) 9owever their position worked% the circular demonstrators claimed to have a third alternative avoiding the agnostics' dilemma% since circular demonstration gives us a regress that is 'oth unending (in the sense that we never reach premises at which it comes to a stop) and finite ('ecause it works its wa& round the finite circle of premises). 7.3 Aristotle's Solutio: "#t <)etually +omes to a Sto!" Aristotle re=ects circular demonstration as an incoherent notion on the grounds that the premises of an& demonstration must 'e prior (in an appropriate sense) to the conclusion% whereas a circular demonstration would make the same premises 'oth prior and posterior to one another (and indeed ever& premise prior and posterior to itself). 9e agrees with the agnostics' anal&sis of the regress pro'lem: the onl& plausi'le options are that it continues indefinitel& or that it "comes to a stop" at some point. 9owever% he thinks 'oth the agnostics and the circular demosntrators are wrong in maintaining that scientific knowledge is onl& possi'le '& demonstration from premises scientificall& known: instead% he claims% there is another form of knowledge possi'le for the first premises% and this provides the starting points for demonstrations. To solve this pro'lem% Aristotle needs to do something +uite specific. It will not 'e enough for him to esta'lish that we can have knowledge of some propositions without demonstrating them: unless it is in turn possi'le to deduce all the other propositions of a science from them% we shall not have solved the regress pro'lem. :oreover (and o'viousl&)% it is no solution to this pro'lem for Aristotle simpl& to assert that we have knowledge without demonstration of some appropriate starting points. 9e does indeed sa& that it is his position that we have such knowledge (An# Post# I.%)% 'ut he owes us an account of wh& that should 'e so. 7.' =o8ledge of 2irst (rici!les: Nous Aristotle's account of knowledge of the indemonstra'le first premises of sciences is found in Posterior Analytics II.1F% long regarded as a difficult te.t to interpret. Criefl&% what he sa&s there is that it is another cognitive state% nous (translated variousl& as "insight"% "intuition"% "intelligence")% which knows them. There is wide disagreement among commentators a'out the interpretation of his account of how this state is reached* I will offer one possi'le interpretation. Airst% Aristotle identifies his pro'lem as e.plaining how the principles can "'ecome familiar to us"% using the same term "familiar" (gnrimos) that he used in presenting the regress pro'lem. /hat he is presenting% then% is not a method of discover& 'ut a process of 'ecoming wise. (econd% he sa&s that in order for knowledge of immediate premises to 'e possi'le% we must have a kind of knowledge of them without having learned it% 'ut this knowledge must not 'e as "precise" as the knowledge that a possessor of science must have. The kind of knowledge in +uestion turns out to 'e a capacit& or power (unamis) which Aristotle compares to the capacit& for sense,perception: since our senses are innate% i.e.% develop naturall&% it is in a wa& correct to sa& that we know what e.g. all the colors look like 'efore we have seen them: we have the capacit& to see them '& nature% and when we first see a color we e.ercise this capacit& without having to learn how to do so first. >ikewise% Aristotle holds% our minds have '& nature the capacit& to recogni8e the starting points of the sciences. In the case of sensation% the capacit& for perception in the sense organ is actuali8ed '& the operation on it of the percepti'le o'=ect. (imilarl&% Aristotle holds that coming to know first premises is a matter of a potentialit& in the mind 'eing actuali8ed '& e.perience of its proper o'=ects: "The soul is of such a nature as to 'e capa'le of undergoing this". (o% although we cannot come to know the first premises without the necessar& e.perience% =ust as we cannot see colors without the presence of colored o'=ects% our minds are alread& so constituted as to 'e a'le to recogni8e the right o'=ects% =ust as our e&es are alread& so constituted as to 'e a'le to perceive the colors that e.ist. It is considera'l& less clear what these o'=ects are and how it is that e.perience actuali8es the relevant potentialities in the soul. Aristotle descri'es a series of stages of cognition. Airst is what is common to all animals: perception of what is present. Be.t is memor&% which he regards as a retention of a sensation: onl& some animals have this capacit&. 3ven fewer have the ne.t capacit&% the capacit& to form a single e.perience (empeiria) from man& repetitions of the same memor&. Ainall&% man& e.periences repeated give rise to knowledge of a single universal (katholou). This last capacit& is present onl& in humans. (ee (ection E of the entr& on Aristotle's ps&cholog& for more on his views a'out mind. >. $efiitios The defiitio (horos% horismos) was an important matter for )lato and for the 3arl& Academ&. @oncern with answering the +uestion "/hat is so,and,so0" are at the center of the ma=orit& of )lato's dialogues% some of which (most ela'oratel& the Sophist) propound methods for finding definitions. 3.ternal sources (sometimes the satirical remarks of comedians) also reflect this Academic concern with definitions. Aristotle himself traces the +uest for definitions 'ack to (ocrates. >." $efiitios ad <sseces Aor Aristotle% a definition is "an account which signifies what it is to 'e for something" (logos ho to ti n einai smainei). The phrase "what it is to 'e" and its variants are crucial: giving a definition is sa&ing% of some e.istent thing% what it is% not simpl& specif&ing the meaning of a word (Aristotle does recogni8e definitions of the latter sort% 'ut he has little interest in them). The notion of "what it is to 'e" for a thing is so pervasive in Aristotle that it 'ecomes formulaic: what a definition e.presses is "the what,it,is,to,'e" (to ti n einai). Homan translators% ve.ed '& this odd Greek phrase% devised a word for it% essentia% from which our "essence" descends. (o% an Aristotelian definition is an account of the essence of something. >.2 S!ecies* 9eus* ad $ifferetia (ince a definition defines an essence% onl& what has an essence can 'e defined. /hat has an essence% then0 That is one of the central +uestions of Aristotle's metaph&sics* once again% we must leave the details to another article. In general% however% it is not individuals 'ut rather s!ecies (eios: the word is one of those )lato uses for "Aorm") that have essences. A species is defined '& giving its geus (genos) and its differetia (iaphora): the genus is the kind under which the species falls% and the differentia tells what characteri8es the species within that genus. As an e.ample% human might 'e defined as animal (the genus) ha%ing the capacity to reason (the differentia). <ssetial (redicatio ad the (redicables ?nderl&ing Aristotle's concept of a definition is the concept of essetial !redicatio (katgoreisthai en ti ti esti% predication in the what it is). In an& true affirmative predication% the predicate either does or does not "sa& what the su'=ect is"% i.e.% the predicate either is or is not an accepta'le answer to the +uestion "/hat is it0" asked of the su'=ect. Cucephalus is a horse% and a horse is an animal* so% "Cucephalus is a horse" and "Cucephalus is an animal" are essential predications. 9owever% "Cucephalus is 'rown"% though true% does not state what Cucephalus is 'ut onl& sa&s something a'out him. (ince a thing's definition sa&s what it is% definitions are essentiall& predicated. 9owever% not ever&thing essentiall& predicated is a definition. (ince Cucephalus is a horse% and horses are a kind of mammal% and mammals are a kind of animal% "horse" "mammal" and "animal" are all essential predicates of Cucephalus. :oreover% since what a horse is is a kind of mammal% "mammal" is an essential predicate of horse. /hen predicate 4 is an essential predicate of 5 'ut also of other things% then 4 is a geus (genos) of 5. A definition of 4 must not onl& 'e essentiall& predicated of it 'ut must also 'e predicated onl& of it: to use a term from Aristotle's Topics% a definition and what it defines must "counterpredicate" (antikatgoreisthai) with one another. 4 counterpredicates with 5 if 4 applies to what 5 applies to and conversel&. Though 4's definition must counterpredicate with 4% not ever&thing that counterpredicates with 4 is its definition. "@apa'le of laughing"% for e.ample% counterpredicates with "human" 'ut fails to 'e its definition. (uch a predicate (non,essential 'ut counterpredicating) is a !eculiar !ro!erty or !ro!rium (iion). Ainall&% if 4 is predicated of 5 'ut is neither essential nor counterpredicates% then 4 is an accidet (sumbebkos) of 5. Aristotle sometimes treats genus% peculiar propert&% definition% and accident as including all possi'le predications (e.g. Topics I). >ater commentators listed these four and the differentia as the five !redicables% and as such the& were of great importance to late ancient and to medieval philosoph& (e.g.% )orph&r&). >.3 The +ategories The notion of essential predication is connected to what are traditionall& called the categories (katgoriai). In a word% Aristotle is famous for having held a "doctrine of categories". Lust what that doctrine was% and indeed =ust what a categor& is% are considera'l& more ve.ing +uestions. The& also +uickl& take us outside his logic and into his metaph&sics. 9ere% I will tr& to give a ver& general overview% 'eginning with the somewhat simpler +uestion "/hat categories are there0" /e can answer this +uestion '& listing the categories. 9ere are two passages containing such lists: /e should distinguish the kinds of predication (ta gen tn katgorin) in which the four predications mentioned are found. These are ten in num'er: what,it,is% +uantit&% +ualit&% relative% where% when% 'eing,in,a,position% having% doing% undergoing. An accident% a genus% a peculiar propert& and a definition will alwa&s 'e in one of these categories. (Topics I.F% 12!'2,#) -f things said without an& com'ination% each signifies either su'stance or +uantit& or +ualit& or a relative or where or when or 'eing,in,a,position or having or doing or undergoing. To give a rough idea% e.amples of su'stance are man% horse* of +uantit&: four, foot% five,foot* of +ualit&: white% literate* of a relative: dou'le% half% larger* of where: in the >&ceum% in the market,place* of when: &esterda&% last &ear* of 'eing,in,a,position: is,l&ing% is,sitting* of having: has,shoes,on% has,armor,on* of doing: cutting% 'urning* of undergoing: 'eing,cut% 'eing,'urned. (Categories "% 1'#,a"% tr. Ackrill% slightl& modified) These two passages give ten,item lists% identical e.cept for their first mem'ers. /hat are the& lists of0 9ere are three wa&s the& might 'e interpreted: The word "categor&" (katgoria) means "predication". Aristotle holds that predications and predicates can 'e grouped into several largest "kinds of predication" (gen tn katgorin). 9e refers to this classification fre+uentl&% often calling the "kinds of predication" simpl& "the predications"% and this ('& wa& of >atin) leads to our word "categor&". Airst% the categories ma& 'e kins of preicate: predicates (or% more precisel&% predicate e.pressions) can 'e divided into ten separate classes% with each e.pression 'elonging to =ust one class. This comports well with the root meaning of the word katgoria ("predication"). -n this interpretation% the categories arise out of considering the most general t&pes of +uestion that can 'e asked a'out something: "/hat is it0"* "0o! much is it0"* "/hat sort is it0"* "/here is it0"* "/hat is it oing0" Answers appropriate to one of these +uestions are nonsensical in response to another ("/hen is it0" "A horse"). Thus% the categories ma& rule out certain kinds of +uestion as ill,formed or confused. This pla&s an important role in Aristotle's metaph&sics. (econd% the categories ma& 'e seen as classifications of preications% that is% kinds of relation that ma& hold 'etween the predicate and the su'=ect of a predication. To sa& of (ocrates that he is human is to sa& what he is% whereas to sa& that he is literate is not to sa& what he is 'ut rather to give a +ualit& that he has. Aor Aristotle% the relation of predicate to su'=ect in these two sentences is +uite different (in this respect he differs 'oth from )lato and from modern logicians). The categories ma& 'e interpreted as ten different wa&s in which a predicate ma& 'e related to its su'=ect. This last division has importance for Aristotle's logic as well as his metaph&sics. Third% the categories ma& 'e seen as kins of entity% as highest genera or kinds of thing that are. A given thing can 'e classified under a series of progressivel& wider genera: (ocrates is a human% a mammal% an animal% a living 'eing. The categories are the highest such genera. 3ach falls under no other genus% and each is completel& separate from the others. This distinction is of critical importance to Aristotle's metaph&sics. /hich of these interpretations fits 'est with the two passages a'ove0 The answer appears to 'e different in the two cases. This is most evident if we take note of point in which the& differ: the Categories lists substace (ousia) in first place% while the Topics list 8hat?it?is (ti esti). A su'stance% for Aristotle% is a t&pe of entit&% suggesting that the Categories list is a list of t&pes of entit&. -n the other hand% the e.pression "what,it,is" suggests most strongl& a t&pe of predication. Indeed% the Topics confirms this '& telling us that we can "sa& what it is" of an entity falling under an& of the categories: an e.pression signif&ing what,it,is will sometimes signif& a su'stance% sometimes a +uantit&% sometimes a +ualit&% and sometimes one of the other categories. As Aristotle e.plains% if I sa& that (ocrates is a man% then I have said what (ocrates is and signified a su'stance* if I sa& that white is a color% then I have said what white is and signified a +ualit&* if I sa& that some length is a foot long% then I have said what it is and signified a +uantit&* and so on for the other categories. /hat,it,is% then% here designates a kind of predication% not a kind of entit&. This might lead us to conclude that the categories in the Topics are onl& to 'e interpreted as kinds of predicate or predication% those in the Categories as kinds of 'eing. 3ven so% we would still want to ask what the relationship is 'etween these two nearl&,identical lists of terms% given these distinct interpretations. 9owever% the situation is much more complicated. Airst% there are do8ens of other passages in which the categories appear. Bowhere else do we find a list of ten% 'ut we do find shorter lists containing eight% or si.% or five% or four of them (with su'stanceMwhat,it,is% +ualit&% +uantit&% and relative the most common). Aristotle descri'es what these lists are lists of in different wa&s: the& tell us "how 'eing is divided"% or "how man& wa&s 'eing is said"% or "the figures of predication" (ta schNmata tNs katNgorias). The designation of the first categor& also varies: we find not onl& "su'stance" and "what it is" 'ut also the e.pressions "this" or "the this" (toe ti% to toe% to ti). These latter e.pressions are closel& associated with% 'ut not s&non&mous with% su'stance. 9e even com'ines the latter with "what,it,is" ($etaphysics 6 1% 121a12: "O one sense signifies what it is and the this% one signifies +ualit& O"). :oreover% su'stances are for Aristotle fundamental for predication as well as metaph&sicall& fundamental. 9e tells us that ever&thing that e.ists e.ists 'ecause su'stances e.ist: if there were no su'stances% there would not 'e an&thing else. 9e also conceives of predication as reflecting a metaph&sical relationship (or perhaps more than one% depending on the t&pe of predication). The sentence "(ocrates is pale" gets its truth from a state of affairs consisting of a su'stance ((ocrates) and a +ualit& (whiteness) which is in that su'stance. At this point we have gone far outside the realm of Aristotle's logic into his metaph&sics% the fundamental +uestion of which% according to Aristotle% is "/hat is a su'stance0". (Aor further discussion of this topic% see the entr& on Aristotle's metaph&sics% and in particular% (ection on the categories.) (ee Arede 1F11% 3'ert 1F1# for additional discussion of Aristotle's lists of categories. Aor convenience of reference% I include a ta'le of the categories% along with Aristotle's e.amples and the traditional names often used for them. Aor reasons e.plained a'ove% I have treated the first item in the list +uite differentl&% since an e.ample of a su'stance and an e.ample of a what,it,is are necessaril& (as one might put it) in different categories. Traditioal ame Literally 9reek <1am!les ((u'stance) su'stance "this" what,it,is ousia toe ti ti esti man% horse (ocrates "(ocrates is a man" Puantit& 9ow much poson four,foot% five,foot Pualit& /hat sort poion white% literate Helation related to what pros ti dou'le% half% greater >ocation /here pou in the >&ceum% in the marketplace Time when pote &esterda&% last &ear )osition 'eing situated keisthai lies% sits 9a'it having% possession echein is shod% is armed Action doing poiein cuts% 'urns )assion undergoing paschein is cut% is 'urned >.' The %ethod of $i)isio In the Sophist% )lato introduces a procedure of "7ivision" as a method for discovering definitions. To find a definition of 4% first locate the largest kind of thing under which 4 falls* then% divide that kind into two parts% and decide which of the two 4 falls into. Hepeat this method with the part until 4 has 'een full& located. This method is part of Aristotle's )latonic legac&. 9is attitude towards it% however% is comple.. 9e adopts a view of the proper structure of definitions that is closel& allied to it: a correct definition of 4 should give the geus (genos: kind or famil&) of 4% which tells what kind of thing 4 is% and the differetia (iaphora: difference) which uni+uel& identifies 4 within that genus. (omething defined in this wa& is a s!ecies (eios: the term is one of )lato's terms for "Aorm")% and the differentia is thus the "difference that makes a species" (eiopoios iaphora% "specific difference"). In Posterior Analytics II.1!% he gives his own account of the use of 7ivision in finding definitions. 9owever% Aristotle is strongl& critical of the )latonic view of 7ivision as a method for establishing definitions. In Prior Analytics I.!1% he contrasts 7ivision with the s&llogistic method he has =ust presented% arguing that 7ivision cannot actuall& prove an&thing 'ut rather assumes the ver& thing it is supposed to 'e proving. 9e also charges that the partisans of 7ivision failed to understand what their own method was capa'le of proving. >.0 $efiitio ad $emostratio @losel& related to this is the discussion% in Posterior Analytics II.!,12% of the +uestion whether there can 'e 'oth definition and demonstration of the same thing. (ince the definitions Aristotle is interested in are statements of essences% knowing a definition is knowing% of some e.isting thing% what it is. @onse+uentl&% Aristotle's +uestion amounts to a +uestion whether defining and demonstrating can 'e alternative wa&s of ac+uiring the same knowledge. 9is repl& is comple.: 1. Bot ever&thing demonstra'le can 'e known '& finding definitions% since all definitions are universal and affirmative whereas some demonstra'le propositions are negative. . If a thing is demonstra'le% then to know it =ust is to possess its demonstration* therefore% it cannot 'e known =ust '& definition. !. Bevertheless% some definitions can 'e understood as demonstrations differentl& arranged. As an e.ample of case !% Aristotle considers the definition "Thunder is the e.tinction of fire in the clouds". 9e sees this as a compressed and rearranged form of this demonstration: (ound accompanies the e.tinguishing of fire. Aire is e.tinguished in the clouds. Therefore% a sound occurs in the clouds. /e can see the connection '& considering the answers to two +uestions: "/hat is thunder0" "The e.tinction of fire in the clouds" (definition). "/h& does it thunder0" "Cecause fire is e.tinguished in the clouds" (demonstration). As with his criticisms of 7ivision% Aristotle is arguing for the superiorit& of his own concept of science to the )latonic concept. Qnowledge is composed of demonstrations% even if it ma& also include definitions* the method of science is demonstrative% even if it ma& also include the process of defining. @. $ialectical Argumet ad the Art of $ialectic Aristotle often contrasts ialectical arguments with demonstrations. The difference% he tells us% is in the character of their premises% not in their logical structure: whether an argument is a sullogismos is onl& a matter of whether its conclusion results of necessit& from its premises. The premises of demonstrations must 'e true an primary% that is% not onl& true 'ut also prior to their conclusions in the wa& e.plained in the Posterior Analytics. The premises of dialectical deductions% '& contrast% must 'e acce!ted (enoxos). @." $ialectical (remises: The %eaig of Endoxos Hecent scholars have proposed different interpretations of the term enoxos. Aristotle often uses this ad=ective as a su'stantive: ta enoxa% "accepted things"% "accepted opinions". -n one understanding% descended from the work of G. 3. >. -wen and developed more full& '& Lonathan Carnes and especiall& Terence Irwin% the enoxa are a compilation of views held '& various people with some form or other of standing: "the views of fairl& reflective people after some reflection"% in Irwin's phrase. 7ialectic is then simpl& "a method of argument from RtheS common 'eliefs Rheld '& these peopleS". Aor Irwin% then% enoxa are "common 'eliefs". Lonathan Carnes% noting that enoxa are opinions with a certain standing% translates with "reputa'le". :& own view is that Aristotle's te.ts support a somewhat different understanding. 9e also tells us that dialectical premises differ from demonstrative ones in that the former are 1uestions% whereas the latter are assumptions or assertions: "the demonstrator does not ask% 'ut takes"% he sa&s. This fits most naturall& with a view of dialectic as argument directed at another person '& +uestion and answer and conse+uentl& taking as premises that other person's concessions. An&one arguing in this manner will% in order to 'e successful% have to ask for premises which the interlocutor is lia'le to accept% and the 'est wa& to 'e successful at that is to have an inventor& of accepta'le premises% i.e.% premises that are in fact accepta'le to people of different t&pes. In fact% we can discern in the Topics (and the Rhetoric% which Aristotle sa&s depends on the art e.plained in the Topics) an art of dialectic for use in such arguments. :& reconstruction of this art (which would not 'e accepted '& all scholars) is as follows. @.2 The T8o <lemets of the Art of $ialectic Given the a'ove picture of dialectical argument% the dialectical art will consist of two elements. -ne will 'e a method for discovering premises from which a given conclusion follows% while the other will 'e a method for determining which premises a given interlocutor will 'e likel& to concede. The first task is accomplished '& developing a s&stem for classif&ing premises according to their logical structure. /e might e.pect Aristotle to avail himself here of the s&llogistic% 'ut in fact he develops +uite another approach% one that seems less s&stematic and rests on various "common" terms. The second task is accomplished '& developing lists of the premises which are accepta'le to various t&pes of interlocutor. Then% once one knows what sort of person one is dealing with% one can choose premises accordingl&. Aristotle stresses that% as in all arts% the dialectician must stud&% not what is accepta'le to this or that specific person% 'ut what is accepta'le to this or that t&pe of person% =ust as the doctor studies what is healthful for different t&pes of person: "art is of the universal". @.2." The "Logical System" of the Topics The method presented in the Topics for classif&ing arguments relies on the presence in the conclusion of certain "common" terms (koina) T common in the sense that the& are not peculiar to an& su'=ect matter 'ut ma& pla& a role in arguments a'out an&thing whatever. /e find enumerations of arguments involving these terms in a similar order several times. T&picall&% the& include: I. -pposites (antikeimena% antitheseis) 1. @ontraries (enantia) . @ontradictories (apophaseis) !. )ossession and )rivation (hexis kai stersis) ". Helatives (pros ti) II. @ases (ptseis) III.":ore and >ess and >ikewise" The four t&pes of o!!osites are the 'est represented. 3ach designates a t&pe of term pair% i.e.% a wa& two terms can 'e opposed to one another. +otraries are polar opposites or opposed e.tremes such as hot and cold% dr& and wet% good and 'ad. A pair of cotradictories consists of a term and its negation: good% not good. A !ossessio (or condition) and !ri)atio are illustrated '& sight and 'lindness. 3elati)es are relative terms in the modern sense: a pair consists of a term and its correlative% e.g. large and small% parent and child. The argumentative patterns Aristotle associated with cases generall& involve inferring a sentence contaning adver'ial or declined forms from another sentence containing different forms of the same word stem: "if what is useful is good% then what is done usefull& is done well and the useful person is good". In 9ellenistic grammatical usage% ptsis meant "case" (e.g. nominative% dative% accusative)* Aristotle's use here is o'viousl& an earl& form of that. ?nder the heading more ad less ad like8ise% Aristotle groups a somewhat motle& assortment of argument patterns all involving% in some wa& or other% the terms "more"% "less"% and "likewise". 3.amples: "If whatever is A is C% then whatever is more (less) A is more (less) C"* "If A is more likel& C than @ is% and A is not C% then neither is @"* "If A is more likel& than C and C is the case% then A is the case". @.2.2 The Topoi At the heart of the Topics is a collection of what Aristotle calls topoi% "places" or "locations". ?nfortunatel&% though it is clear that he intends most of the Topics (Cooks II,UI) as a collection of these% he never e.plicitl& defines this term. Interpreters have conse+uentl& disagreed considera'l& a'out =ust what a topos is. 7iscussions ma& 'e found in Crunschwig 1F$E% (lomkowski 1FF$% )rimavesi 1FFE% and (mith 1FFE. @.3 The -ses of $ialectic ad $ialectical Argumet An art of dialectic will 'e useful wherever dialectical argument is useful. Aristotle mentions three such uses* each merits some comment. @.3." 9ymastic $ialectic Airst% there appears to have 'een a form of st&li8ed argumentative e.change practiced in the Academ& in Aristotle's time. The main evidence for this is simpl& Aristotle's Topics% especiall& Cook UIII% which makes fre+uent reference to rule,governed procedures% apparentl& taking it for granted that the audience will understand them. In these e.changes% one participant took the role of answerer% the other the role of +uestioner. The answerer 'egan '& asserting some proposition (a thesis: "position" or "acceptance"). The +uestioner then asked +uestions of the answerer in an attempt to secure concessions from which a contradiction could 'e deduced: that is% to refute (elenchein) the answerer's position. The +uestioner was limited to +uestions that could 'e answered '& &es or no* generall&% the answerer could onl& respond with &es or no% though in some cases answeres could o'=ect to the form of a +uestion. Answerers might undertake to answer in accordance with the views of a particular t&pe of person or a particular person (e.g. a famous philosopher)% or the& might answer according to their own 'eliefs. There appear to have 'een =udges or scorekeepers for the process. G&mnastic dialectical contests were sometimes% as the name suggests% for the sake of e.ercise in developing argumentative skill% 'ut the& ma& also have 'een pursued as a part of a process of in+uir&. @.3.2 $ialectic That (uts to the Test Aristotle also mentions an "art of making trial"% or a variet& of dialectical argument that "puts to the test" (the Greek word is the ad=ective peirastik% in the feminine: such e.pressions often designate arts or skills% e.g. rhtorik% "the art of rhetoric"). Its function is to e.amine the claims of those who sa& the& have some knowledge% and it can 'e practiced '& someone who does not possess the knowledge in +uestion. The e.amination is a matter of refutation% 'ased on the principle that whoever knows a su'=ect must have consistent 'eliefs a'out it: so% if &ou can show me that m& 'eliefs a'out something lead to a contradiction% then &ou have shown that I do not have knowledge a'out it. This is strongl& reminiscent of (ocrates' st&le of interrogation% from which it is almost certainl& descended. In fact% Aristotle often indicates that dialectical argument is '& nature refutative. @.3.3 $ialectic ad (hiloso!hy 7ialectical refutation cannot of itself esta'lish an& proposition (e.cept perhaps the proposition that some set of propositions is inconsistent). :ore to the point% though deducing a contradiction from m& 'eliefs ma& show that the& do not constitute knowledge% failure to deduce a contradiction from them is no proof that the& are true. Bot surprisingl&% then% Aristotle often insists that "dialectic does not prove an&thing" and that the dialectical art is not some sort of universal knowledge. In Topics I.% however% Aristotle sa&s that the art of dialectic is useful in connection with "the philosophical sciences". -ne reason he gives for this follows closel& on the refutative function: if we have su'=ected our opinions (and the opinions of our fellows% and of the wise) to a thorough refutative e.amination% we will 'e in a much 'etter position to =udge what is most likel& true and false. In fact% we find =ust such a procedure at the start of man& of Aristotle's treatises: an enumeration of the opinions current a'out the su'=ect together with a compilation of "pu88les" raised '& these opinions. Aristotle has a special term for this kind of review: a iaporia% a "pu88ling through". 9e adds a second use that is 'oth more difficult to understand and more intriguing. The Posterior Analytics argues that if an&thing can 'e proved% then not ever&thing that is known is known as a result of proof. /hat alternative means is there where'& the first principles of sciences are known0 Aristotle's own answer as found in Posterior Analytics II.1F is difficult to interpret% and recent philosophers have often found it unsatisf&ing since (as often construed) it appears to commit Aristotle to a form of apriorism or rationalism 'oth indefensi'le in itself and not consonant with his own insistence on the indispensa'ilit& of empirical in+uir& in natural science. Against this 'ackground% the following passage in Topics I. ma& have special importance: It is also useful in connection with the first things concerning each of the sciences. Aor it is impossi'le to sa& an&thing a'out the science under consideration on the 'asis of its own principles% since the principles are first of all% and we must work our wa& through a'out these '& means of what is generall& accepted a'out each. Cut this is peculiar% or most proper% to dialectic: for since it is e.aminative with respect to the principles of all the sciences% it has a wa& to proceed. A num'er of interpreters ('eginning with -wen 1F$1) have 'uilt on this passage and others to find dialectic at the heart of Aristotle's philosophical method. Aurther discussion of this issue would take us far 'e&ond the su'=ect of this article (the fullest development is in Irwin 1F11* see also Buss'aum 1F1$ and Colton 1FF2* for criticism% 9aml&n 1FF2% (mith 1FFE). A. $ialectic ad 3hetoric Aristotle sa&s that rhetoric% i.e.% the stud& of persuasive speech% is a "counterpart" (antistrophos) of dialectic and that the rhetorical art is a kind of "outgrowth" (paraphues ti) of dialectic and the stud& of character t&pes. The correspondence with dialectical method is straightforward: rhetorical speeches% like dialectical arguments% seek to persuade others to accept certain conclusions on the 'asis of premises the& alread& accept. Therefore% the same measures useful in dialectical conte.ts will% mutatis mutandis% 'e useful here: knowing what premises an audience of a given t&pe is likel& to 'elieve% and knowing how to find premises from which the desired conclusion follows. The Rhetoric does fit this general description: Aristotle includes 'oth discussions of t&pes of person or audience (with generali8ations a'out what each t&pe tends to 'elieve) and a summar& version (in II.!) of the argument patterns discussed in the Topics. Aor further discussion of his rhetoric see Aristotle's rhetoric. "B. So!histical Argumets 7emonstrations and dialectical arguments are 'oth forms of valid argument% for Aristotle. 9owever% he also studies what he calls cotetious (eristikos) or so!histical arguments: these he defines as arguments which onl& apparentl& esta'lish their conclusions. In fact% Aristotle defines these as apparent ('ut not genuine) ialectical sullogismoi. The& ma& have this appearance in either of two wa&s: 1. Arguments in which the conclusion onl& appears to follow of necessit& from the premises (apparent% 'ut not genuine% sullogismoi). . Genuine sullogismois the premises of which are merel& apparentl&% 'ut not genuinel&% accepta'le. Arguments of the first t&pe in modern terms% appear to 'e valid 'ut are reall& invalid. Arguments of the second t&pe are at first more perple.ing: given that accepta'ilit& is a matter of what people 'elieve% it might seem that whatever appears to 'e enoxos must actuall& 'e enoxos. 9owever% Aristotle pro'a'l& has in mind arguments with premises that ma& at first glance seem to 'e accepta'le 'ut which% upon a moment's reflection% we immediatel& reali8e we don not actuall& accept. @onsider this e.ample from Aristotle's time: /hatever &ou have not lost% &ou still have. 5ou have not lost horns. Therefore% &ou still have horns This is transparentl& 'ad% 'ut the pro'lem is not that it is invalid: the pro'lem is rather that the first premise% though superficiall& plausi'le% is false. In fact% an&one with a little a'ilit& to follow an argument will reali8e that at once upon seeing this ver& argument. Aristotle's stud& of sophistical arguments is contained in On Sophistical Refutations% which is actuall& a sort of appendi. to the Topics. To a remarka'le e.tent% contemporar& discussions of fallacies reproduce Aristotle's own classifications. (ee 7orion 1FF# for further discussion. "". ,o?+otradictio ad %eta!hysics Two fre+uent themes of Aristotle's account of science are (1) that the first principles of sciences are not demonstra'le and () that there is no single universal science including all other sciences as its parts. "All things are not in a single genus"% he sa&s% "and even if the& were% all 'eings could not fall under the same principles" (On Sophistical Refutations 11). Thus% it is e.actl& the universal applica'ilit& of dialectic that leads him to den& it the status of a science. In $etaphysics IU (V)% however% Aristotle takes what appears to 'e a different view. Airst% he argues that there is% in a wa&% a science that takes 'eing as its genus (his name for it is "first philosoph&"). (econd% he argues that the principles of this science will 'e% in a wa&% the first principles of all (though he does not claim that the principles of other sciences can 'e demonstrated from them). Third% he identifies one of its first principles as the "most secure" of all principles: the principle of non, contradiction. As he states it% It is impossi'le for the same thing to 'elong and not 'elong simultaneousl& to the same thing in the same respect ($et# ) This is the most secure of all principles% Aristotle tells us% 'ecause "it is impossi'le to 'e in error a'out it". (ince it is a first principle% it cannot 'e demonstrated* those who think otherwise are "uneducated in anal&tics". 9owever% Aristotle then proceeds to give what he calls a "refutative demonstration" (apoeixai elenktiks) of this principle. Aurther discussion of this principle and Aristotle's arguments concerning it 'elong to a treatment of his metaph&sics (see Aristotle: :etaph&sics). 9owever% it should 'e noted that: (1) these arguments draw on Aristotle's views a'out logic to a greater e.tent than an& treatise outside the logical works themselves* () in the logical works% the principle of non,contradiction is one of Aristotle's favorite illustrations of the "common principles" (koinai archai) that underlie the art of dialectic. (ee Aristotle's :etaph&sics% 7anc& 1FE#% @ode 1F1$ for further discussion. "2. Time ad ,ecessity: The Sea?;attle The passage in Aristotle's logical works which has received perhaps the most intense discussion in recent decades is On Interpretation F% where Aristotle discusses the +uestion whether ever& proposition a'out the future must 'e either true or false. Though something of a side issue in its conte.t% the passage raises a pro'lem of great importance to Aristotle's near contemporaries (and perhaps contemporaries). A cotradictio (antiphasis) is a pair of propositions one of which asserts what the other denies. A ma=or goal of On Interpretation is to discuss the thesis that% of ever& such contradiction% one mem'er must 'e true and the other false. In the course of his discussion% Aristotle allows for some e.ceptions. -ne case is what he calls idefiite propositions such as "A man is walking": nothing prevents 'oth this proposition and "A man is not walking" 'eing simultaneousl& true. This e.ception can 'e e.plained on relativel& simple grounds. A different e.ception arises for more comple. reasons. @onsider these two propositions: 1. There will 'e a sea,'attle tomorrow . There will not 'e a sea,'attle tomorrow It seems that e.actl& one of these must 'e true and the other false. Cut if (1) is no! true% then there must 'e a sea,'attle tomorrow% and there cannot fail to 'e a sea,'attle tomorrow. The result% according to this pu88le% is that nothing is possi'le e.cept what actuall& happens: there are no unactuali8ed possi'ilities. (uch a conclusion is% as Aristotle is +uick to note% a pro'lem 'oth for his own metaph&sical views a'out potentialities and for the commonsense notion that some things are up to us. 9e therefore proposes another e.ception to the general thesis concerning contradictor& pairs. This much would pro'a'l& 'e accepted '& most interpreters. /hat the restriction is% however% and =ust what motivates it are matters of wide disagreement. It has 'een proposed% for instance% that Aristotle adopted% or at least flirted with% a three,valued logic for future propositions% or that he countenanced truth,value gaps% or that his solution includes still more a'struse reasoning. The literature is much too comple. to summari8e: see Anscom'e% 9intikka% 7. Arede% /hitaker% /aterlow. 9istoricall&% at least% it is likel& that Aristotle is responding to an argument originating in the :egarian (chool. 9e ascri'es the view that onl& that which happens is possi'le to the :egarians in $etaphysics I4 (W). The pu88le with which he is concerned strongl& recalls the ":aster Argument" of 7iodorus @ronus% especiall& in certain further details. Aor instance% Aristotle imagines the statement a'out tomorrow's sea 'attle having 'een uttered ten thousand &ears ago. If it was true% then its truth was a fact a'out the past* if the past is now unchangea'le% then so is the truth value of that past utterance. This recalls the :aster Argument's premise that "what is past is necessar&". 7iodorus @ronus was active a little after Aristotle% and he was a :egarian (see 7orion 1FF# for criticism of 7avid (edle&'s attempt to re=ect this). It seems to me reasona'le to conclude that Aristotle's target here is some :egarian argument% perhaps an earlier version of the :aster. "3. 9lossary of Aristotelia Termiology Accept: tithenai (in a dialectical argument) Accepted: enoxos (also ;reputa'le< ;common 'elief<) Accident: sumbebkos (see inciental) Accidental: kata sumbebkos Affirmation: kataphasis Affirmative: kataphatikos Assertion: apophansis (sentence with a truth value% declarative sentence) Assumption: hupothesis Celong: huparchein @ategor&: katgoria (see the discussion in (ection E.!). @ontradict: antiphanai @ontradiction: antiphasis (in the sense "contradictor& pair of propositions" and also in the sense "denial of a proposition") @ontrar&: enantion 7eduction: sullogismos 7efinition: horos% horismos 7emonstration: apoeixis 7enial (of a proposition): apophasis 7ialectic: ialektik (the art of dialectic) 7ifferentia: iaphora* specific difference% eiopoios iaphora 7irect: eiktikos (of proofs* opposed to "through the impossi'le") 3ssence: to ti esti% to ti n einai 3ssential: en ti ti esti (of predications) 3.treme: akron (of the ma=or and minor terms of a deduction) Aigure: schma Aorm: eios (see also (pecies) Genus: genos Immediate: amesos ("without a middle") Impossi'le: aunaton* "through the impossi'le" (ia tou aunatou)% of some proofs. Incidental: see Accidental Induction: epagg :iddle% middle term (of a deduction): meson Begation (of a term): apophasis -'=ection: enstasis )articular: en merei% epi meros (of a proposition)* kath2hekaston (of individuals) )eculiar% )eculiar )ropert&: iios% iion )ossi'le: unaton% enechomenon* enechesthai (ver': "'e possi'le") )redicate: katgorein (ver')* kategoroumenon ("what is predicated") )redication: katgoria (act or instance of predicating% t&pe of predication) )rimar&: prton )rinciple: arch (starting point of a demonstration) Pualit&: poion Heduce% Heduction: anagein% anagg Hefute: elenchein* refutation% elenchos (cience: epistm (pecies: eios (pecific: eiopoios (of a differentia that "makes a species"% eiopoios iaphora) (u'=ect: hupokeimenon (u'stance: ousia Term: horos ?niversal: katholou ('oth of propositions and of individuals)