Aristotle on Dialectic and First Principles

This is an open access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion 4.0 In e na ional (CC BY 4.0), which pe mi s ep oduc ion, adap a ion, and dis ibu ion p o ided he o iginal au ho and sou ce a e c edi ed. Filoso ia Unisinos Unisinos Jou nal o Philosophy 18(1):57-62, jan/ap 2017 Unisinos – doi: 10.4013/ su.2017.181.07 PHILOSOPHY SOUTH ABSTRACT O e ime, se e al au ho s ha e a gued ha , o A is o le, dialec ic is a me hod o disco e y o es ablishmen o he p inciples o science in gene al. In his a icle, I will p esen ou a gumen s agains his iew and p opose a de la iona y iew o he ole o dialec ic in his ega d. Acco d ing o his iew, such a ole consis s only in he de ence o common p inci- ples agains po en ial e is ic a acks. Keywo ds: Science, dialec ic, i s p inciples, common p inciples, p ope p inciples. 1 Cen o de Filoso ia da Uni e - sidade de Lisboa. Faculdade de Le as da Uni e sidade de Lisboa, Alameda da Uni e sidade, 1600- 214, Lisboa, Po ugal. E-mail: [email p o ec ed] 2 See, o example: Weil (1975); Owen (1961); I win (1977); Nussbaum (1986); F ede (1987). I should be highligh ed ha only he i s p inciples o demons a i e science (ἐπιστήµη ἀποδεικτικός) a e he e aken in o accoun , wi h he excep ion, he e o e, o hose o non-demons a i e sciences (as, o ins ance, he p ac ical and p oduc i e sciences). 3 See, among many occu ences o his lemma: APo. A2, 71b19–72a8; Top. Ζ4, 141a23–142a21; EN Ζ3, 1139b31–35. A is o le on Dialec ic and Fi s P inciples An ónio Ped o Mesqui a1 I O e ime, se e al au ho s ha e de ended ha , o A is o le, dialec ic is a me hod o disco - e y o es ablishmen o he p inciples o science in gene al.2 In my iew, ou easons speak agains his hesis. The i s is ha while, o A is o le, each science, besides he common p inciples i sha es wi h o he sciences, has ce ain p inciples ha a e p ope o i , he ne e in e p e s dialec ic as a me hod o disco e ing, es ablishing o e en jus i ying p ope p inciples o pa icula sciences. Ra he , dia- lec ic is simply a me hod o discussing (and some imes upholding) common p inciples. The second is ha i dialec ic we e a gene al me hod o disco e y o es ablishmen o he p inciples o all sciences, i would ha e o disco e o posi no only hei common p inciples, bu also hose ha a e p ope o each science. And since he ask o disco e ing o posi ing he p ope p inciples o each science belongs o he science o which he p inciples a e p ope , dialec ic would become he science o which all p inciples a e p ope and he e o e a “common” o “uni e sal” science. Howe e , he ac is ha A is o le is adaman ha he e canno be a common science and ha , speci ically, dialec ic is no such a common o uni e sal science. The hi d eason has o do wi h he e y special sense unde which, acco ding o A is o le, a uni e sal cha ac e can be a ibu ed o dialec ic. This sense is he ollowing: dialec ic, albei no a uni e sal science, has none heless a uni e sal scope inso a as i add esses ma e s abou which e - e ybody knows and on which e e yone may gi e an opinion. Hence, only when i add esses issues o which e e yone has some knowledge and can speak – ha is o say, only when i add esses such issues a a le el ha e e yone can access and in such a way ha e e yone can gi e hei opinion – can dialec ic ha e a uni e sal cha ac e . Now, p inciples a e by hei e y na u e “mo e knowable in hemsel es han hey a e o us”;3 hus, i is no by disco e ing o es ablishing he p inciples o science, which ew people know abou and owa d which no inexpe opinions coun , ha dialec ic is acknowledged by A is o le as a comp ehensi e and ans e sal, all-encopassing (and, in his sense, uni e sal) me hod. An ónio Ped o Mesqui a Filoso i a Unisinos – Unisinos Jou nal o Philosophy – 18(1): 57-62, jan/ap 2017 58 The ou h and las eason o ejec he iew ha dia- lec ic is a me hod o disco e y o es ablishmen o he i s p inciples o science in gene al is he ollowing: A is o le e- pea edly men ions he p ocess by which (p ope ) p inciples o science a e posi ed and such p ocess is no dialec ic, bu a he expe ience. I is he e o e qui e clea ha o him he disco e y o es ablishing o p inciples is no o be included unde he scope o dialec ic’s asks. In wha ollows, I will expose hese ou a gumen s a g ea e leng h and, in a inal, mo e posi i e sec ion, disclose and de end my own iew on he ole o dialec ic in his ma - e , namely ha his ole consis s in he de ence o common p inciples agains po en ial e is ic a acks. II As is well known, acco ding o A is o le, each science has, besides he common p inciples i sha es wi h o he sci- ences, o wi h a wide g oup o o he sciences (τὰ κοινά, τὰ ἀξιώµατα), ce ain p inciples ha a e p ope o i (τὰ ἴδια) (c . APo. A10.76a37–b11; APo. A32.88b27–29; SE 11.172a36– b1; bu also see AP . A30.46a17–22, APo. A32.88a30–b6, e De an. A1.402a18–22). Now, A is o le ne e in e p e s dialec ic as a me hod o disco e y, pos ula ion o e en jus i ica ion o p ope p inciples o pa icula sciences, bu a he simply as a me hod o discussion o common p inciples. And e en i i is ue ha dialec ic may also be applied, by ex ension, o p ope p in- ciples o sciences, i can only be so unde e y es ic ed e ms. In e ec , dialec ic is epea edly desc ibed by A is o le as a special discipline which, since i has no p ope genus as i s ob- jec (c . APo. A11, 77a26–35; SE 11, 172a11–15; 11, 172a27– 30; Rh. A1, 1355b8–9; A 2, 1356a30–34), bu a he discuss- es p oblems in any domain (c . Top. A1, 100a18–24; SE 11, 172a27–30; 34, 183a37–b8; c . SE 11, 172a36–b1; Me aph. Γ2, 1004b17–26; EE A8, 1217b16–19; Rh. A1, 1354a1–3) – no based on ue p emises ei he , as i occu s wi h pa icula sci- ences, bu on “ epu able” o “accep ed opinions” (τὰ ἔνδοξα) (see Top. A1, 100a18–24, and SE 34, 183a37–b8; c . AP . A27, 43b1–17; APo. A33, 88b30–89a4; B19, 100b5–17; Me aph. Γ2, 1004b17–26) – has he special abili y o c ossing o e wi h all sciences h ough he p inciples common o all o hem (c . APo. A 11, 77a26–35; SE 9, 170a20–b11; 11, 171b6–7; 11, 172a27–30; 11, 172a36–b1). I is his abili y ha g an s i he possibili y (and some imes he mission) o de ending such common p inciples,4 as well as – and he e a e he speci ic e ms ha p eside o e i s ex ension o p ope p inciples o sciences, which A is o le ecognises only once5 – o ca ying ou a p e- limina y examina ion o discussion o he p inciples p ope o each science, based on he “accep ed” o “ epu able” opinions abou he ma e s co e ed by hese sciences. These a e A is o le’s own wo ds in his ega d:6 Fo i we eason om he p inciples app o- p ia e o he science in ques ion, i is impos- sible o make any s a emen abou hese (since hese p inciples a e he i s o hem all) and i is by means o accep ed opin- ions abou each ha is necessa y o discuss hem. Bu his ask is p ope , o a any a e mos app op ia e, o dialec ic: o since i s abili y o examine applies o he p inciples o all s udies, i has a way o p oceed (Top. A2, 101a37-b4). The ole a ibu ed o dialec ic in his ex clea ly con- o ms o he desc ip ion abo e. I consis s in lis ing and ana- lyzing he ἔνδοξα which ha e implica ions on each science and in discussing hei p inciples in he ligh o hose ἔνδοξα, no mo e gene al conclusion being jus i iably d awn om i .7 Mo eo e , his is wha A is o le himsel does in se e al o his scien i ic ea ises, when he p ecedes he scien i ic dis- cussion o he objec wi h a lis o ecei ed o gene ally accep - ed opinions8 and wi h he p esen a ion o he p oblems ha he analyzed adi ion has le pending (τὰ ἀπορούµενα) (as o example in Ph. A8, Γ4, ∆1, ∆6, ∆10, Ζ9, Θ2; De an. A1, 402a10–403b19; Me aph. B1–6; EN Ε11–12, Η3). In no way do hese necessa y bu p elimina y momen s eplace he scien i ic discussion which is consequen ly de el- oped, as would be he case i he me hodology o science, and especially he pos ula ion o i s p inciples, we e o an exclu- si ely dialec ic na u e. Close obse a ion allows us o se ou he six ollowing poin s: (i) The in e en ion o dialec ic on i s p inciples is always con ined o common p inciples, wi h he well-de ined excep ion e e ed o abo e. 4 C . SE 9, 170a20–b11 (on he con a y, he de ence o p ope p inciples is exp essly a ibu ed o pa icula sciences). 5 Top. A2, 101a37–b4 (quo ed below, in he ex ). O all e e ences o dialec ic in associa ion wi h he p inciples (APo. A11, 77a26–35; Top. A2, 101a36–b4; SE 9, 170a20–b11; 11, 171b6–7; 11, 172a27–30; 11, 172a36–b1), his is indeed he only passage ha conce ns p ope p inciples. 6 All ansla ions o he Topics by Smi h (1994), wi h some adap a ions in he key concep s. 7 Along he same lines, see Smi h: “All ha he p esen passage says is ha such c i ical examina ions a e ‘use ul’ in ‘discussing’ scien- i ic s a ing-poin s, and ha alls a sho o claiming ha dialec ic ei he es ablishes o disco e s hose s a ing-poin s. Finally, gi en A is o le’s epea ed claims in he Topics and On Sophis ical Re u a ions ha dialec ical a gumen canno es ablish any hing, i would be mo e han a li le su p ising o him o hold ha i can es ablish wha scien i ic demons a ions canno ” (Smi h, 1994, p. 54). 8 Fo example, in Physics (A2–4, 9, and as a ule a he onse o he discussion on each new opic), in De anima (A2–5), in Me aphyics (A3–10) o in Nicomachean E hics (A3–4, A8–9, Η3, Η12, Κ2). This is a p ocedu e which A is o le himsel heo ises: see Cael. A10, 279a5–12, as well as Me aph. α1, 993a30–b19, and B1, 995a24–b4. A is o le on Dialec ic and Fi s P inciples Filoso i a Unisinos – Unisinos Jou nal o Philosophy – 18(1): 57-62, jan/ap 2017 59 (ii) The ela ionship be ween dialec ic and common p inciples is no one o disco e y, pos ula ion o jus i ica ion. (iii) Common p inciples a e no hough o as he p e- e ed objec o dialec ic, bu only as ha by which dialec ic ela es o all sciences. (i ) Common p inciples as such only become he objec o dialec ic when hey need o be de ended agains an e is ic a ack. ( ) The e o e, in no way can dialec ic be ega ded as a gene al science o he ἀξιώµατα, bu only as a me hod o de ending common p inciples (mo eo e , o easons ela ed o he e y na u e o common p inciples hemsel es, as will be discussed la e ). ( i) The uni e sali y o dialec ic, as a discipline capable o discussing p oblems in any gi en ield wha soe - e , is no o be a ibu ed o any special ela ionship wi h common p inciples (since such a ela ionship, as p e iously men ioned, is con ined o i s de ence in he e en o an e is ic a ack), bu a he o he speci ic na u e o dialec ic as a discipline wi h no ma e ial objec . The e is no e idence, hus, o sus ain he hesis ha di- alec ic cons i u es a gene al science o p inciples, o ha i is he only o a o ed me hod o access i s p inciples o science. Ra he , he a ailable da a poin in he opposi e di ec ion. III Howe e , no only does A is o le no say ha dialec ic is, in any sense o he exp ession, a science o p inciples o a me hod o accessing p inciples, bu his heo y o science would also clea ly p e en him om doing so – a ac ha A is o le himsel explici ly obse es. In ac , i dialec ic we e he gene al me hod o disco - e y, pos ula ion o jus i ica ion o he p inciples o all sci- ences, i would be so wi h ega d o no only he common p inciples o all sciences, bu also hose p ope o each one o hem. And since his ask belongs o he science o which he p inciples a e p ope (c . AP . A30, 46a17–22), dialec ic would ex hypo hesi become he science o which all p inci- ples a e p ope and he e o e would become a common o uni e sal science. None heless, he ac is ha A is o le is e y clea ha he e canno be a common science, as each science s udies a single genus and each genus alls unde ce ain p inciples, which a e he p ope p inciples o ha science (c . APo. A7; A9, 75b37–76a15; A12, 77a36–b15; A28, 87a38–b4; A32; SE 9, 170a20–b11; 11, 172a11–15; GA B8, 747b27–748a15; see also AP . A30, 46a17–22; De an. A1, 402a18–22; Me aph. B2, 997a18–25; K3, 1060b31– 36; K7, 1063b36–1064a4). Fo dialec ic o be a common science, i would be necessa y o all gene a o lead o an o e all genus, whose p inciples would be he p ope p in- ciples o he common science and he common p inciples o all pa icula sciences, which is no he case.9 O dialec ic would ha e o be he supe o dina e science o all pa icula sciences, a condi ion ha does no ob ain ei he , as he e canno be, o he eason gi en abo e, a single science o which all o he sciences a e subo dina e.10 Tha is why i is also no legi ima e o hink ha dia- lec ic deals indi ec ly wi h p ope p inciples simply because i deals wi h common p inciples in he i s place. In ac , his would imply admi ing ha p ope p inciples a e, up o a ce - ain poin , subo dina e o common p inciples and ha hey may be deduced om hem. Fo A is o le, on he con a y, he ἀξιώµατα a e no p inciples om which e e y hing can be deduced, gi en ha he sup eme gene a unde which all beings all ( he ca ego ies) a e di e en and only wi h he con ibu ion o p ope p inciples o each o hese gene a can ac ual conclusions be d awn.11 Fu he mo e, his would, once again, make dialec ic a uni e sal science, which, as we ha e seen, is p e en ed by he plu ali y o gene a i educible o each o he and o an (illuso- y) common p ima y genus, as well as by he co esponding plu ali y o pa icula sciences i educible o each o he and o an (illuso y) p ima y science. Tha is why A is o le empha ically and epea idly un- de lines ha dialec ic is no a common o a uni e sal science12 – no only, by he way, o he easons se ou abo e, bu p ima ily o he mo e basic eason ha dialec ic is no e en a science, since i does no ope a e κατ’ ἀλήθειαν, bu only κατὰ δόξαν (c . AP . A1, 24a22–b15; A30, 46a3–10; APo. A2, 72a8–14; A19, 81b18–23; Top. A1, 100a25–b23; A14, 105b30–31; Θ1, 155b3–16; Θ3, 159a11–14; SE 2, 165a38– b11; SE 11, 172a36–b1; Me aph. Γ2, 1004b17–26; see also De an. A1, 402b25–403a2). 9 The maxima gene a a e hemsel es plu al ( he hen ca ego ies): c . especially AP . A27, 42b20–43; APo. A9-22; A32, 88a36–b3. 10 The only excep ion o he ule o one genus co esponding o one science is, in ac , he case o he subo dina e sciences (see again APo. A7; A9, 75b37–76a15; A28, 87a38–b4; and also: APo. A13, 78a22–79a16; A15, 98a24–34). Bu e en his excep ion is mo e appa - en han eal, as he p inciples o subo dina e sciences a e common o he supe o dina e ones and only o his ex en a e hey sha ed by bo h (c . APo. A9). 11 C . APo. A32, 88a36–b3, eely pa aph ased he e. This hesis is ob iously in line wi h he s a emen in he same ea ise, acco ding o which he p inciples o non-con adic ion and o excluded middle only ake pa o he p emises o a demons a ion unde excep ional ci cums ances (in he i s case, when he conclusions a e ins ances o he p inciple o non-con adic ion; and, in he second case, when he demons a ion is made pe impossibile: c . APo. A11, 77a10–25. 12 C . SE 9, 170a20–b11, and 11, 172a11–15. Fo he gene al hesis ha he e is no uni e sal science, see especially Me aph. A9, 992b18–993a7, and EE A8, 1217b25–1218a1, and c . EN A4, 1096a23–29. An ónio Ped o Mesqui a Filoso i a Unisinos – Unisinos Jou nal o Philosophy – 18(1): 57-62, jan/ap 2017 60 This is hen he second majo eason why dialec ic can- no be he uni e sal me hod o es ablishing he i s p inciples o sciences. In sho : he e canno be, in gene al, a uni e sal science; and dialec ic, in pa icula , is no ha science. IV This leads us o he hi d majo eason. The uni e sali y o dialec ic, on which es s i s abili y o discuss p oblems in any domain, a om equi ing i s in e - en ion in he ealm o i s p inciples, is due o h ee mo i es ha pa ially exclude i . Fi s ly, i s ma e ial emp iness: i is because dialec ic does no ha e any genus as i s p ope objec , ha i can uni e sally eso o co e all gene a, and as a esul c osso e wi h all sciences (see APo. A11, 77a26–35; SE 11, 172a27–30; and c . SE 11, 172a11–15). Secondly, he ci cums ance ha , also o his eason, dia- lec ic is no a science, bu only a me hod o “a capaci y o u - nishing a gumen s” (δυνάµεις τινὲς τοὺ πορίσαι λόγους) – and p ecisely o u nishing a gumen s in any ield.13 Thi dly, he ac ha dialec ic add esses opics abou which, by de ini ion, e e ybody knows and anyone can gi e an opinion (c . Rh. A1, 1354a1–3), a ac ha cons i u es no only a jus i ica ion o i s uni e sali y, bu also a es ic ion in he meaning o be gi en o such uni e sali y. Gene ally, one can say ha he i s poin p e en s dia- lec ic om being a science, he hi d excludes i om being he gene al me hod o es ablishing i s p inciples, and he second explains exac ly in wha sense dialec ic has a uni e sal na u e. Gi en ha enough has been said abou he i s poin and he hi d one is he mos undamen al o he p esen dis- cussion, le us begin wi h he second poin . Wha does i imply o each one o he pa icula scienc- es o decla e ha dialec ic is “a capaci y o u nishing a gu- men s” in any ield? To “ u nish a gumen s” (na u ally o a dialec ical na u e) in any ield, and he e o e o any science, su ely canno mean submi ing o pa icula sciences he gene al a gumen a i e schemes hey would be allowed o use, ega dless o any com- pliance wi h hei p ope p inciples, in hei s ic ly scien i ic ou ine. This would be indeed unaccep able in wo espec s: i s ly, science can only p oceed demons a i ely om i s p ope p inciples; and, secondly, scien i ic demons a ion is based on ue p emises, whe eas dialec ical a gumen s a e by de ini ion based on p emises which a e simply “accep ed” as ue, i.e., on ἔνδοξα. I we bea in mind A is o le’s ac ual p ac ice in his own scien i ic ea ises, hen o u nish dialec ical a gumen s o he pa icula sciences mus a he mean o p o ide hem wi h an in oduc o y coun e balance o heses and easons p e i- ously se o h on a ce ain opic o inqui y – a coun e bal- ance which scien is s should be acquain ed wi h and ake in o due accoun when p oposing an explana o y heo y o ha opic.14 The ad an age o his p elimina y coun e balance is ha i allows he scien is o pick up he p os and con as ha , acco ding o he “accep ed” opinions, a e aised in ela ion o each p oblem and hus map ou he di icul ies ha , in his ac- ual scien i ic in es iga ion, will ha e o be aken in o accoun and esol ed. I is in e es ing ha , o he h ee u ili ies ha A is o le assigns o dialec ic (Top. A2, 101a25–36), wo ha e a di ec link wi h his applica ion o science: he eco d o ἔνδοξα; and he habi o es ing an apo ia in bo h senses (πρὸς ἀµφότερα διαπορῆσαι) in o de o be able o easily dis inguish ue om alse in e e y opic.15 Undoub edly, his doc ine explains he al eady men- ioned ac ha A is o le’s own in es iga ions in his scien i ic ea ises a e o en in oduced by an exhaus i e conspec us o he adi ion and by a de ailed s udy o he di icul ies o be aced and esol ed. Le us now look a he hi d g ound o he uni e sali y o dialec ic and he eason why i implies ha dialec ic can- no es ablish i s p inciples: dialec ic add esses ma e s which e e ybody knows abou and on which e e yone may gi e an opinion. This ce ainly gi es an a gumen o he uni e sali y o dialec ic. Howe e , i also shows in wha exac sense one can speak o dialec ic being uni e sal. In e ec , i dialec ic is uni e sal because i add esses ob- jec s which e e yone knows, i is also only while i add esses ob- jec s ha e e yone knows, ha is, only while i add esses such objec s a a le el which e e yone can access, and in such a way ha e e yone can gi e hei opinion, ha dialec ic is uni e sal. 13 Rh. A2, 1356a33–34. No e ha his cha ac e isa ion de i es om he canonical de ini ion o dialec ic: see Top. A1, 100a18–24, and SE 34, 183a37–b8; bu c . also Top. A2, 101a28–30 (and no e: µέθοδον γὰρ ἔχοντες ῥᾷον περὶ τοῦ προτεθέντος ἐποχειρεῖν δυνησόµεθα), as well as Θ14, 164b1–7. E ans summa izes pe ec ly his poin when he s a es: “Dialec ic is indeed conce ned wi h e e y hing abou e e y hing, as i s philosophy is no ; bu he p ice which i pays o his uni e sali y is ha , unlike i s philosophy, i is no scien i ic in cha ac e ” (E ans, 1977, p. 48). To his no e o gene ali y, he Eudemian E hics adds ano he no e: dialec ical a gumen s a e no only common (κοινοί), bu also undamen ally des uc i e (ἀναιρετικοί); c . EE A8, 1217b16–19. 14 Tha is why A is o le equen ly p ecedes he analy ical conside a ion o a p oblem (i.e., i s conside a ion in acco dance wi h p ope p inciples o he science in ques ion) wi h a dialec ical app oach (λογικῶς, i.e., gene al and abs ac , because i is independen om he p inciples p ope o any science and he e o e applicable o all ields), o p e ixes o he de elopmen o scien i ic a gumen s a numbe o me ely “logical” a gumen s: see pa adigma ically APo. A22, 82b37–84a30; A24, 85b23–86a30; A32, 88a19–b7; Ph. Γ5, 204b4–205a7; Me aph. Ζ4, 1029b13–1030b3; c . Cael. A7, 275b12–276a17, Me aph. A2, 982a4–b10, and he long in e egnum καθόλου in Cael. A10–12, 280a32–283b9. 15 The i s , o a “gymnas ic” na u e, consis s p ecisely in he abili y o a gue abou any opic. A is o le on Dialec ic and Fi s P inciples Filoso i a Unisinos – Unisinos Jou nal o Philosophy – 18(1): 57-62, jan/ap 2017 61 When A is o le decla es, in he Sophis ical Re u a ions, ha om common p inciples qua common dialec ic can be- come an “in es iga ion o all hings” and se s i sel as a “ce ain a ”,16 his s a emen should p obably be unde s ood in he same ein, hus implying ha such an in es iga ion will only p oceed o he ex en o he knowledge o common people and in he p opo ion o he gene al opinion. And, no doub , i is wi h his p ecep in mind ha A is o le decla es in he Me aphysics ha dialec ic discusses all opics, bu in a me ely inquisi i e ashion (πειραστική) and no wi h a iew o knowl- edge (γνωριστική).17 Now, his qui e ob iously excludes dialec ic om being an a o p inciples, as hey a e by na u e mo e knowable in hemsel es han o us and he e o e necessa ily inaccessible o common knowledge as such (see APo. A2, 71b19–72a8; Top. Ζ4, 141a23–142a21; EN Ζ3, 1139b31–35; c . AP . B16, 64b28–65a9; Top. A1, 100a25–29; A1, 100a30–b21; Θ1, 155b3–16; Θ3, 159a11–14; Θ11, 161b30–33). Bu , a he same ime, i does no exclude he possibili y o i o include in i s goal, as a “ce ain a ”, he de ence o common p inciples, in a way ye o be deciphe ed. Once again, we a e led o conclude, he e o e, ha o A is o le dialec ic is no he gene al me hod o es ablishing he p inciples o science – bu ha he de ence o common p inciples agains po en ial e is ic a acks may all unde he scope o i s mission. V The ou h and inal eason o his conclusion is i ial. I is ne e heless impo an o ecall i , gi en ha some de- ende s o he hesis unde analysis appea o o ge i a imes. A is o le epea edly men ions he p ocess by which (p ope ) i s p inciples a e es ablished and his p ocess is no dialec ic, bu a he (some kind o ) expe ience (see, o ins ance: AP . A30, 46a17–22; APo. A18, 81a38–b9; A31; B2, 90a24–30; B19, 99b20–100b5; Cael. Γ7, 306a5–9; GC A2, 316a5–14; HA A6, 491a7–14; GA Γ10, 760b27–33; EN A7, 1098b3–4; Ζ9, 1142a16–20). Rega dless o how we wish o in e p e his doc ine, wha seems o be an inescapable ac is ha o A is o le he es ablishing o he i s p inciples o science is no wi hin he powe s o dialec ic. VI I is ime o sum up. I ou accoun is igh , dialec ic plays no ole wha soe - e in he disco e y, pos ula ion o jus i ica ion o he p ope p inciples o pa icula sciences. Conce ning hese p inciples, he ask o dialec ic is limi ed o i s examina ion and discus- sion in he ligh o ἔνδοξα. The ole o dialec ic is he e o e limi ed, in p inciple, o ἀξιώµατα. Bu wha exac ly is his ole? Le us i s ecall wha i is no . Su ely, i is no up o dialec ic o disco e o pos ula e hem. I is no up o dialec ic o disco e hem, since i does no disco e any hing. Dialec ic discusses gi en p oposi ions, in o de o con i m o e u e hem, acco ding o he ole ha is as- signed o he dialec ician in he ques ion/answe exe cise, based on accep ed opinions, ecognised as such by bo h pa ies.18 Bu i is no up o dialec ic o pos ula e hem ei he . In ac , he con i ma ion o e u a ion o gi en p oposi ions ha esul s om he dialec ical exe cise, suppo ed as hey a e on pu ely endoxic p emises, only ha e alue wi hin he ame- wo k de e mined by he dependence om hese p emises. And since he ἔνδοξα a e, by de ini ion, only accep ed as ue, and may in ac be alse,19 no easoning based on ἔνδοξα can u h ully es ablish any hing a all, and he e o e no eason- ing based on ἔνδοξα can es ablish common p inciples, which a e p inciples o demons a ion and he e o e ue p inciples (c . APo. A2, 71b19–33; A9, 75b37–76a15; A19, 81b18–23; Top. A1, 100a25–29). Wha , hen, is he ole o dialec ic ega ding common p inciples? Tha ole has al eady been sugges ed. The ole o de end- ing common p inciples om an e is ic a ack, ha is, om a malicious a ack, made only ic o iae g a ia.20 In ac , con a y o p ope p inciples, common p inci- ples a e, acco ding o A is o le, necessa y in i ue o hem- 16 Ἔστιν ἐκ τούτων περὶ ἁπάντων πεῖραν λαµβάνειν καὶ εἶναι τέχνην τινά” (SE 11, 172a39–b1). 17 C . Me aph. Γ2, 1004b17–26. The Eudemian E hics, as we ha e seen (abo e, n. 13), would e en add: in a me ely des uc i e way (c . A8.1217b16–19). 18 C . Top. Θ14, 164b1–7, and Θ1, 155b3–16; bu see also: AP . A1, 24a22–b12; APo. A11, 77a31–35; A2, 72a8–14; Top. Θ2, 158a14–21; SE 10, 171a38–b2; 11, 171b3–6; 11, 172a15–21. The c i e ia o an opinion o be “accep ed”, in he echnical sense, a e clea ly de- ined a he beginning o he Topics: accep ed opinions [ἔνδοξα] a e hose “which seem so [ ue and p ima y] o e e yone, o o mos people, o o he wise – o all o hem, o o mos , o o he mos amous and es emed [τοῖς µάλιστα γνωρίµοις καὶ ἐνδόξοις]” (Top. A1, 100b21–23). Fo simila desc ip ions o ἔνδοξα, see: Top. A14, 105a34–b12, and Γ1, 116a13–20. 19 C . especially Top. Θ12, 162b27: εἰ µὲν γὰρ ἐκ ψευδῶν ἐνδόξων δέ, λογικός; bu see also APo. B19, 100b5–7. 20 As A is o le unde lines in SE 11, 171b22–34, e is ic is mo ed by he desi e o win he discussion a any cos (al hough i may also be accompanied by he desi e o mone a ily p o i om i , in which case i should be p ope ly called “sophis ic”). Tha is why, o A is o le, he e is ic con ende is, by de ini ion, a disloyal qua elle who uses any means o gi e he appea ance o ha ing won. The means ha he uses and he way o ecognising hem and o denounce hem as e is ic a e he objec o he Sophis ical Re u a ions. An ónio Ped o Mesqui a Filoso i a Unisinos – Unisinos Jou nal o Philosophy – 18(1): 57-62, jan/ap 2017 62 sel es and ecognized as such.21 Now, i common p inciples a e so, su ely hey can only be ques ioned on he basis o an e is ic a gumen . Dialec ic is he e o e equi ed, since i is i s ask o educe sophis ical e u a ions.22 A good example o his dialec ical de ence o a com- mon p inciple in he co pus is he discussion o he p inciple o non-con adic ion in book Γ 3-8 o Me aphysics. In his ex , A is o le clea ly explains he s a us o such a de ence: i consis s in e u ing he ejec ion o he p inciple o non-con- adic ion by showing he impossibili y o such a nega ion (c . Me aph. Γ4, 1006a11–28). In sho : p ope p inciples o sciences a e known in a non-dialec ical way, by a p ocedu e ha somehow in ol es expe ience; common p inciples a e acknowledged by each pa icula science wi h he pa icula con en ha applies o ha science,23 e en i dialec ic may de end hem om any a - emp o an e is ic a ack. The conclusion is no , he e o e, ha dialec ic has no unc ion as ega ds he p inciples o science. I is, a he , ha i s in e en ion is much mo e limi ed and modes han hose campaigning o he hesis unde analysis belie e, as ha unc- ion is p ima ily limi ed o common p inciples and only con- sis s o suppo ing, and no disco e ing o es ablishing, hem. In he case o p ope p inciples, dialec ic is only equi ed in a supplemen a y way, o he p elimina y analysis o he e- cei ed opinions and he acknowledgmen o he ἀπορίαι which esul om hem and which he scien i ic in es iga ion will ha e o ace. Re e ences BEKKER, I. (ed.) 1831. A is o elis Ope a, I-II. Be lin, Königlichen P eussischen Akademie de Wissenscha , 789 + 673 p. EVANS, J.D.G. 1977. A is o le’s Concep o Dialec ic. Cam- b idge, Camb idge Uni e si y P ess, 150 p. FREDE, M. 1987. The Uni y o he Gene al and Special Me a- physics: A is o le Concep ion o Me aphysics. In: M. FRE- DE, Essays in Ancien Philosophy. Minneapolis, Minneapo- lis Uni e si y P ess, p. 81-95. IRWIN, T.H. 1977. A is o le’s Disco e y o Me aphysics. Re iew o Me aphysics, 31:210-229. NUSSBAUM, M.C. 1986. Sa ing A is o le’s Appea ances. In: M.C. NUSSBAUM, The F agili y o Goodness. Camb idge, Camb idge Uni e si y P ess, p. 240-263. OWEN, G.E.L. 1961. Ti henai a phainomena. In: S. MAN- SION (ed.), A is o e e les p oblèmes de mé hode. Com- munica ions p ésen ées au Symposium A is o elicum. Lou- ain/Pa is, Publica ions Uni e si ai es de Lou ain/Édi ions Béa ice-Nauwelae s, p. 83-103. ROSS, W.D. (ed.) 1949. A is o le’s P io and Pos e io Analy ics: A Re ised Tex wi h In oduc ion and Commen a y. Ox o d, Cla endon P ess, 690 p. SMITH, R. (ed.) 1994. Topics Books I and VIII. Ox o d, Cla en- don P ess, 246 p. WEIL, E. 1975. La place de la logique dans la pensée a is o él- icienne. In: J. BARNES; M. SCHOFIELD; R. SORABJI (eds.), A icles on A is o le. I: Science. London/New Yo k, Duck- wo h/S . Ma in’s P ess, p. 88-112. Submi ed on Ma ch 14, 2017 Accep ed on Ma ch 16, 2017 21 Οὐκ ἔστι δ’ ὑπόθεσις οὐδ’ αἴτηµα, ὃ ἀνάγκη εἶναι δι’ αὑτὸ καὶ δοκεῖν ἀνάγκη (APo. A10, 76b23–24). Al hough he exp ession ἀξίωµα does no explici ly occu in his passage, he con ex clea ly indica es ha i e e s o hem; see Ross (1949, p. 541). 22 C . SE 9, 170a20–b11; 34, 183a37–b8. Hence, again, i s p edominan ly des uc i e na u e (EE A8, 1217b16–19). 23 This is a leas wha seems o esul om Me aph. Γ3, 1005a23–27 (axioms a e used in all sciences bu only wi hin he amewo k o he genus whe e hei demons a ions ope a e), and, mo e gene ically, om APo. A10, 76a37–b11 (axioms a e only κοινὰ κατ’ ἀναλογίαν, gi en ha hei applica ion di e s depending on he science in which hey in e ene and he genus o which hey apply).