After Parmenides: Studies on Language and Metaphysics in Early Greek Philosophy 3896659804, 9783896659804, 9783896659811

Table of contents :
Cover
Author’s Prefatory Note
Prefazione (M. Pulpito)
Acknowledgments
Abbreviations
Part I: Preliminaries and Fundamentals
1. Presocratic Origins of the Principle that There Are No Origins from Nothing
2. “Nothing” as “Not-Being”: Some Literary Contexts That Bear on Plato and also on Parmenides
Part II: The Discovery of Form
3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples
4. Quality, Structure, Emergence: in Anaxagoras, Philolaus, Democritus, and in Empedocles
5. The Concept of the Universal in Some Later pre-Platonic Cosmologists
6. Democritus on the Distinction Between Universals and Particulars: the Relevance of Fragments B124 and B165
7. Intrinsic and Relational Properties of Atoms in the Democritean Ontology
Part III: The Sophist’s Demurrer
8. Gorgias on the Function of Language
Bibliography
Indexes

Citation preview

Supplementa E leatica 2 Alexander P. D. Mourelatos

After Parmenides Studies on Language and Metaphysics in Early Greek Philosophy A cura di Massimo Pulpito

ACADEMIA

Supplementa Eleatica, vol. 2

Series editors: Stefania Giombini Massimo Pulpito Series founder: Livio Rossetti

Supplementa E leatica 2

Alexander P. D. Mourelatos

After Parmenides Studies on Language and Metaphysics in Early Greek Philosophy A cura di Massimo Pulpito

ACADEMIA

Questo volume è stato finanziato con il contributo della Regione Campania LR7/2003

© Coverpicture: Mosaico dei filosofi (part.) Museo Archeologico Nazionale di Napoli.

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.d-nb.de ISBN

978-3-89665-980-4 (Print) 978-3-89665-981-1 (ePDF)

British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. ISBN

978-3-89665-980-4 (Print) 978-3-89665-981-1 (ePDF)

Library of Congress Cataloging-in-Publication Data Mourelatos, Alexander P. D. After Parmenides Studies on Language and Metaphysics in Early Greek Philosophy Alexander P. D. Mourelatos 239 pp. Includes bibliographic references and index. ISBN

978-3-89665-980-4 (Print) 978-3-89665-981-1 (ePDF)

Onlineversion Nomos eLibrary

1st Edition 2022 © Academia Verlag within Nomos Verlagsgesellschaft, Baden-Baden, Germany 2022. Overall responsibility for manufacturing (printing and production) lies with Nomos Verlagsgesellschaft mbH & Co. KG. This work is subject to copyright. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. Under § 54 of the German Copyright Law where copies are made for other than private use a fee is payable to “Verwertungs­gesellschaft Wort”, Munich. No responsibility for loss caused to any individual or organization acting on or refraining from action as a result of the material in this publication can be accepted by Nomos or the author. Visit our website academia-verlag.de

ǽ੾ȞȦȞȠ੤IJȠȢIJ੽ȞʌȡંIJİȡȠȞȝ੻Ȟ੥੼ȜȘȞ੢ıIJİȡȠȞį੻ ਫȜ੼ĮȞĭȦțĮ੼ȦȞȠ੣ıĮȞਕʌȠȚț઀ĮȞ Įਫ਼IJȠ૨ į੻ ʌĮIJȡ઀įĮʌંȜȚȞİ੝IJİȜો țĮ੿ ȝંȞȠȞਙȞįȡĮȢਕȖĮșȠઃȢIJȡ੼ijİȚȞਥʌȚıIJĮȝ੼ȞȘȞ Zenone amò la città prima detta Iele e poi Elea, che era una colonia dei Focesi, sua patria, una città modesta, capace soltanto di crescere uomini virtuosi. (Diogene Laerzio, IX 28)

Dai giardini di Palazzo Alario risuonano – sapientemente impresse su pietra – le parole di Diogene Laerzio, a dare il benvenuto al visitatore in terra di Elea, città natale dei filosofi Parmenide e Zenone, un tempo fiorente colonia focea della Magna Grecia, oggi parco archeologico nel comune di Ascea. La descrizione dello storico greco ben rappresenta, ancora a distanza di secoli, lo spirito dei luoghi: il borgo costiero di Ascea – nell’area protetta del Parco Nazionale del Cilento, Vallo di Diano e Alburni, che vanta i riconoscimenti UNESCO come Patrimonio Culturale, Geoparco, Comunità Emblematica della Dieta Mediterranea, Riserva di Biosfera – si presenta al viaggiatore moderno come una cittadina modesta, abitata da uomini virtuosi e orgogliosi del passato glorioso di Elea-Velia. Ed è proprio la volontà di valorizzare la memoria della civiltà eleatica, quale fondamento razionale del pensiero occidentale, uno degli obiettivi statutari che informa la mission della Fondazione Alario per Elea-Velia. Istituzione di promozione e di organizzazione della cultura – riconosciuta quale Fondazione culturale di preminente interesse regionale – ha sede nello storico Palazzo Alario, antica masseria fortificata del XIX secolo, all’interno di un imponente complesso strutturale a pochi chilometri dal sito archeologico, con sale attrezzate, auditorium all’aperto, cineteatro, laboratorio linguistico, biblioteca specialistica e foresteria. Fin dalla sua costituzione nel 1986, la Fondazione – le cui attività istituzionali interessano i settori della formazione, ricerca, promozione culturale e sviluppo locale – si dedica allo studio e all’approfondimento dei temi che riguardano la filosofia eleatica e, più in generale, il pensiero presocratico. In particolare, dal 2004 – prima annualmente, poi con cadenza biennale – la Fondazione organizza Eleatica, la sessione internazionale di studio sull’Eleatismo e i suoi massimi esponenti che

coinvolge studiosi provenienti da tutto il mondo, tanto da essere inserita dal Domenicale del Sole24Ore tra gli eventi culturali più significativi in Italia. Da Eleatica, d’intesa con il Collegio dei Cittadini Onorari e grazie alla prestigiosa collaborazione con Academia Verlag, nasce la collana omonima delle pubblicazioni scientifiche, che rendono conto delle giornate di studio e degli ampi dibattiti di respiro internazionale cui è collegata l’attività convegnistica. Il viaggiatore attento alle parole di Diogene Laerzio non mancherà di cogliere la tenacia e l’impegno della Fondazione nel perseguimento degli obiettivi di valorizzazione di un’area – qual è appunto quella dell’antica Chora Velina e del Cilento in generale – storicamente e culturalmente determinante per lo sviluppo dei tratti identitari della civiltà occidentale. Un’area che ci auguriamo possa rappresentare una tappa obbligata nel percorso verso la conoscenza. Marcello D’Aiuto Presidente

Table of contents

Author’s Prefatory Note .................................................................................

10

Prefazione (M. Pulpito) ..................................................................................

11

Acknowledgments ..........................................................................................

17

Abbreviations .................................................................................................

18

Part I: Preliminaries and Fundamentals 1. Presocratic Origins of the Principle that There Are No Origins from Nothing ..................................................................................................

21

2. “Nothing” as “Not-Being”: Some Literary Contexts That Bear on Plato and also on Parmenides.................................................................

36

Part II: The Discovery of Form 3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples..................................................................................................

49

4. Quality, Structure, Emergence: in Anaxagoras, Philolaus, Democritus, and in Empedocles .................................................................................

79

5. The Concept of the Universal in Some Later pre-Platonic Cosmologists..........................................................................................

127

6. Democritus on the Distinction Between Universals and Particulars: the Relevance of Fragments B124 and B165...............................................

146

7. Intrinsic and Relational Properties of Atoms in the Democritean Ontology ................................................................................................

159

Part III: The Sophist’s Demurrer 8. Gorgias on the Function of Language....................................................

183

Bibliography...................................................................................................

213

Indexes ...........................................................................................................

227

For Olive Elizabeth Glaeser Forbes

Author’s Prefatory Note In autumn 2019, Stefania Giombini and Massimo Pulpito, in their capacity as general editors of the series Supplementa Eleatica, invited me to assemble a set of thematically related essays of mine, with a view to having these essays republished as a single volume in that series. The present book is the outcome of the solicited project. I reaffirm here my strongly felt thanks to the two series-editors for their very welcome invitation. On several of the occasions of my attending Eleatica conferences, the Fondazione Alario per Elea-Velia very kindly hosted me and my wife in the Fondazione’s beautifully situated campus in Ascea – just a few kilometers from the site which in antiquity was the city and home of Parmenides and Zeno of Elea. I am grateful for those occasions of memorable hospitality; and I also record here my keen and sincere appreciation, both academically and personally, of the Fondazione’s sponsorship of the entire Supplementa Eleatica series. Massimo Pulpito also offered to serve as editor of the solicited volume, and his contribution in that role has been invaluable. Most of the papers had to be digitally transcribed from hard-copy versions, which conformed to the variously different styles (European and North American) of the venues in which the papers had originally appeared. So, fresh copy-editing, and – in more than a few cases – revision and updating of content were required. I thank Myriam Bittner of Academia Verlag, for her assistance and for her understanding in a process that took longer than the publisher had envisaged. In the collaboration between author and volume-editor, Massimo Pulpito has been careful, thorough, generous with his time, also congenially patient, and astutely judicious when offering criticism or suggestions. My profound thanks to him for his work as editor of the present volume. Austin, Texas, November 2021 A.P.D.M.

Prefazione Massimo Pulpito

Il campo di studi principale di Alexander P. D. Mourelatos è stato il pensiero presocratico, assieme ad alcuni aspetti della filosofia di Platone e di Aristotele. Il suo primo libro, The Route of Parmenides, è stato pubblicato nel 1970, e poi riedito in una versione riveduta e ampliata nel 2008. Quel volume, apparso quando l’autore aveva solo trentatré anni, è ritenuto oggi uno dei testi più importanti della letteratura critica su Parmenide. Un posto particolare tra i Presocratici lo ricopre dai primi anni 2000 un pensatore pre-parmenideo, Senofane di Colofone, la cui fisica, talvolta marginalizzata, è stata invece l’oggetto di una originale reinterpretazione da parte di Mourelatos. Proprio a Senofane (e alla sua affascinante “astrofisica delle nuvole”) furono dedicate le lectiones magistrales che lo studioso tenne in italiano per Eleatica 2010, e che gli valsero l’anno dopo la Cittadinanza Onoraria dell’Antica Città di Elea. Le lezioni si intitolavano La “natura delle cose” prima di Parmenide: il mondo visto da Senofane, e offrivano uno sguardo lucido su una delle fasi più interessanti della riflessione naturalistica pre-parmenidea. Ma Mourelatos, già nei tre decenni precedenti, aveva riservato una serie di studi altrettanto significativi a quel che era avvenuto dopo Parmenide. E come avrebbe poi fatto nel caso di Senofane, in quegli studi egli problematizzò la lettura, fino troppo schematica, secondo cui tutto era legato (anche retrospettivamente) a Parmenide. Se nel caso del Colofonio Mourelatos avrebbe rivelato gli errori interpretativi generati dal “mito eleatico”, che aveva condotto a sovrainterpretare (e in ultima analisi a fraintendere) i versi di Senofane, degradandolo a mero precursore, nel caso dei filosofi post-parmenidei si era proposto di valutare il reale condizionamento che Parmenide (e l’eleatismo in genere) avevano determinato sulla filosofia successiva. Seppure, infatti, tale condizionamento appare innegabile, è alto il rischio di una sua sopravvalutazione, allorché tutta la riflessione successiva viene forzatamente ricondotta ad una mera replica all’eleatismo. Mourelatos si è sforzato di mostrare come il quadro fosse sensibilmente più complesso, cercando di far emergere anche i contributi più originali di questi pensatori (e cioè i cosiddetti Pluralisti, i Pitagorici – più nello specifico Filolao – ma anche figure più prossime all’eleatismo, come Melisso o, seppure polemicamente, Gorgia). È a questa fase della sua ricerca, incentrata sul pensiero post-parmenideo, che è dedicata principalmente questa raccolta. Il volume contiene otto saggi, di cui sette già pubblicati in inglese, e uno (il terzo di questo libro) che originariamente, in una

12

Massimo Pulpito

versione ridotta precedente – qui sviluppata e ampliata – era stato pubblicato in greco moderno. Il libro è suddiviso in tre parti. La prima, intitolata Preliminaries and Fundamentals, si apre con il saggio “PRESOCRATIC ORIGINS OF THE PRINCIPLE THAT THERE ARE NO ORIGINS FROM NOTHING”, che si propone di condurre una breve indagine sulle radici del celebre principio ex nihilo nihil. L’attenzione si concentra naturalmente sull’eleatismo. Dapprima Mourelatos esamina i due diversi argomenti a cui Parmenide ricorre in favore del suddetto principio: l’uno che si appella all’indicibilità e impensabilità del non essere; l’altro che evoca il “principio di ragion sufficiente”. Quindi, l’autore volge la sua attenzione all’uso che dell’ex nihilo nihil (e del complementare rifiuto dell’annichilimento) faranno due pensatori post-parmenidei, Melisso ed Empedocle. Contro la tesi dominante secondo cui il principio sorgerebbe con Parmenide (condizionando così tutta la filosofia successiva), e in parte riabilitando la tesi aristotelica che vedeva il principio antico quanto il pensiero filosofico stesso, Mourelatos mostra come Melisso sembra non solo non essere condizionato dagli argomenti parmenidei (specialmente quello epistemologico-semantico) ma rispondere a una richiesta di intellegibilità certamente preeleatica. Quanto a Empedocle nella sua versione del principio sono palesemente in gioco anche credenze di tipo religioso (come quella riguardante l’immortalità) più antiche del poema di Parmenide. Il secondo capitolo si intitola “‘NOTHING’ AS ‘NOT-BEING’: SOME LITERARY CONTEXTS THAT BEAR ON PLATO AND ALSO ON PARMENIDES”. Qui lo studioso riprende alcuni passi della letteratura greca classica che mostrano come il senso di termini come “niente” e “nessuno” (e il corrispondente “non essere”) non fosse solo di tipo esistenziale, come si è tradizionalmente creduto, ma che accanto ad esso (pur prevalente negli usi letterari) vi fosse un significato di tipo “caratterizzante”. L’essere niente o nessuno indicherebbe non l’assenza, ma una valutazione negativa dei caratteri di un certo oggetto o individuo. Poiché questo uso minoritario compare in passaggi di grande drammaticità, il suo valore concettuale, secondo Mourelatos, si sarebbe rivelato di particolare rilevanza nelle elaborazioni filosofiche di Parmenide e Platone. In Parmenide, infatti, come l’autore ha mostrato in The Route of Parmenides, al centro delle riflessioni parmenidee più che la mera esistenza vi è la caratterizzazione della natura delle cose. In Platone, invece, tale valore sarebbe riemerso nell’idea secondo cui la realtà è organizzata in gradi. Non è forse un caso se l’ultima parola del capitolo che chiude la prima parte è “form”, nozione che sembra fare da ponte con la seconda sezione del libro. Questa parte, infatti, affronta il nucleo tematico principale dell’intera raccolta: The Discovery of Form. Il terzo capitolo, che apre questa sezione e si intitola “FORM AS SHAPE, ARRAY, ORDER: EARLIEST APPROACHES AND ELEATIC SCRUPLES”, introduce tale linea di indagine, offrendo uno sguardo di insieme sulla genesi di questa “scoperta” nel pensiero presocratico. Mettendo in discussione l’interpretazione classica che opponeva il materialismo degli Ionici alla centralità degli aspetti formali e strutturali della realtà riconosciuta dai Pitagorici, Mourelatos mostra come la “scoperta della forma” possegga in realtà una storia ben più complessa, articolata in quattro fasi. Nella prima, corrispondente a quella in cui vengono

Prefazione

13

elaborate le prime cosmologie ioniche del sesto e del quinto secolo, emerge un graduale riconoscimento del valore degli aspetti formali nella spiegazione del mondo. Il secondo momento è quello più complesso e vede come protagonista Parmenide. Da un lato l’Eleate sarà responsabile di una vera “crisi” concettuale attraverso il rifiuto di ogni genesi, cambiamento, divisibilità, che Melisso – figura che in questa storia assume un ruolo decisivo – estenderà anche alle “forme”. Dall’altro, però, nella problematica sezione sulla Doxa, Parmenide ha fatto ricorso proprio alla forma e alla struttura nella spiegazione del mondo fisico. È a quest’ultimo contributo parmenideo che, in modo più incisivo, nella terza fase, sembrano rifarsi Empedocle e Anassagora, con le loro prospettive diverse, ma incentrate in ogni caso sulla crucialità degli aspetti formali e strutturali nel loro modello esplicativo. Ciò però non supera il problema emerso nel contributo critico di Parmenide, sviluppato da Melisso, a cui sembrano invece rispondere, nella quarta e ultima fase, da un lato Filolao, con la parificazione metafisica – ma certo non assiologica – tra il piano della materia e quello delle forme (le “cose illimitate” e le “cose limitanti”), e dall’altro Democrito, con l’unificazione di materia e forma nei costituenti atomici della realtà. Questo lungo e articolato percorso intellettuale, che sfocia nella risposta filolaica e democritea alla sfida lanciata a Elea, aprirà la strada alle due maggiori “filosofie della forma” dell’antichità: quella di Platone (che in un certo senso si porrà sulla traccia di Filolao) e quella di Aristotele (che invece avanzerà una proposta da questo punto di vista analoga – sebbene diversa e persino rivale – alla filosofia di Democrito). Nel lungo e cruciale quarto capitolo (quasi una monografia), “QUALITY, STRUCTURE, EMERGENCE: IN ANAXAGORAS, PHILOLAUS, DEMOCRITUS, AND IN EMPEDOCLES”, Mourelatos si concentra su un problema ancora più specifico. Se è vero che i cosiddetti Pluralisti hanno ammesso enti fondamentali immuni, more eleatico, alla generazione, alla corruzione e al cambiamento, in che modo, si chiede l’autore, le loro teorie resistono alle obiezioni del “secondo eleatismo” (Zenone e Melisso), che colpiscono l’emergere ex nihilo di qualità e strutture derivate dalla ricomposizione degli elementi fondamentali? Se infatti tali entità non nascono, non muoiono e non cambiano, a essere soggetti a nascita, morte e cambiamento sembrerebbero essere gli aspetti qualitativi e strutturali dei loro composti. In altre parole, possiamo ritenere filosofi come Anassagora, Empedocle, Filolao e gli Atomisti degli “emergentisti” (come efficacemente li designa Mourelatos), che cioè conferiscono agli aspetti formali uno status ontologico speciale e sopravveniente, diverso da quello degli enti fondamentali? In questo capitolo Mourelatos mostra come tali aspetti secondari siano, in realtà, riconducibili a caratteri primari degli enti fondamentali come la loro prevalenza quantitativa (in Anassagora), l’armonia di tipo matematico (in Filolao), la forma atomica (in Democrito), o la presenza di poroi e la meccanica erotica degli elementi (in Empedocle) – escludendo, dunque, qualunque forma di ingenuo “emergentismo” in questi autori. Particolarmente ingegnosa la soluzione del problema ideata da Empedocle, a cui, non a caso, lo studioso dedica una particolare attenzione. Empedocle è qui presentato come ben consapevole della questione degli effetti dell’“emergenza”, ma senza che egli conceda che tali effetti siano in definitiva coinvolti nello schema

14

Massimo Pulpito

esplicativo da lui immaginato. Egli non avrebbe adottato una metafisica fondata sulla dicotomia realtà/apparenza, né anticiperebbe l’alterazione qualitativa aristotelica, e tantomeo immaginerebbe una sorta di principio di preformazione. Egli, piuttosto, avrebbe cercato di rendere intelligibile la genesi sfruttando in modo sagace una fisica dei canali flessibili e deformabili (i poroi, appunto), in combinazione con una dottrina dell'illusione percettiva. Un altro dei significati dei termini eidos e idea, utilizzati per indicare la forma, è quello che fa riferimento al tipo o al genere di qualcosa, ossia a ciò che nella storia della filosofia verrà chiamato “universale”. A questa nozione sono dedicati i due capitoli successivi di questa sezione. Il quinto saggio ha per titolo “THE CONCEPT OF THE UNIVERSAL IN SOME LATER PRE-PLATONIC COSMOLOGISTS”, ed è volto a tratteggiare i lineamenti di una sorta di preistoria del concetto filosofico di universale. Facendo riferimento principalmente alla nota distinzione di Charles Sanders Peirce tra le nozioni di “type” e “token”, Mourelatos prende in esame alcuni frammenti tratti dalle opere di tre autori, ossia Empedocle (circa la composizione degli oggetti sulla base delle radici, le forme biologiche così come i modelli che si ripetono nei cicli cosmici), Filolao (circa la relazione tra i generi, le specie e i particolari) e Democrito (soprattutto a proposito delle forme atomiche). Proprio a quest’ultimo autore (il più giovane tra i cosiddetti Presocratici) e in particolare ai frammenti nei quali sembra farsi avanti una separazione consapevole tra universali e particolari, l’autore dedica il capitolo successivo, “DEMOCRITUS ON THE DISTINCTION BETWEEN UNIVERSALS AND PARTICULARS: THE RELEVANCE OF FRAGMENTS B124 AND B165”. Qui Mourelatos approfondisce le analisi già condotte nel quinto capitolo, mostrando come nel frammento B124 Democrito sembra tematizzare la separazione metafisica tra il livello dell’universale (l’uno) e quello dei particolari (i molti), mentre in B165 la stessa distinzione è presentata i termini epistemologici. Ancora a Democrito è dedicato il settimo capitolo, l’ultimo di questa sezione centrale: “INTRINSIC AND RELATIONAL PROPERTIES OF ATOMS IN THE DEMOCRITEAN ONTOLOGY”. Qui, a partire dall’analisi di un passaggio della Metafisica di Aristotele ed esplorando le discrepanze lessicali tra la testimonianza aristotelica e il vocabolario democriteo, lo studioso si propone di indagare analiticamente la struttura logica soggiacente alle proprietà degli atomi di Democrito, distinguendo tra le proprietà intrinseche che gli atomi possiedono, quindi indipendentemente dagli altri atomi (come ad esempio la forma e la dimensione), e quelle che, invece, dipendono dalla relazione e interazione con essi (come l’ordine e la posizione). Ciò consente di mostrare in dettaglio (attraverso una inaspettata ramificazione di proprietà) il complesso modello esplicativo soggiacente all’ontologia atomistica. L’ultimo capitolo del volume – “GORGIAS ON THE FUNCTION OF LANGUAGE” – possiede una sezione a sé, che funge da epilogo: The Sophist’s Demurrer. Qui Mourelatos si occupa della terza parte del trattato Sul non essere o sulla natura di Gorgia, affrontando la questione della concezione del significato e quindi della funzione del linguaggio secondo il Sofista di Lentini, di cui l’Encomio di Elena è il testo di riferimento. Mourelatos mostra che tra una referential conception of meaning (secondo cui il significato si reduce al riferimento a qualcosa di esterno) e una

Prefazione

15

ideational conception (per la quale, invece, il significato corrisponde ad una immagine mentale) bisogna invece riconoscere una terza via, adottata da Gorgia, e cioè la behavioral conception, che rimarca la funzione “comportamentale” del linguaggio. Come si evince anche dal dettagliato esame degli argomenti utilizzati nell’ultima parte di Sul non essere (a dispetto della finalità critica e dialettica del trattato) e soprattutto in una testimonianza di Sesto Empirico (e dunque non solo nell’Elena), il linguaggio per Gorgia parrebbe essere una forma di interazione volta a ottenere precisi effetti negli interlocutori. Come si vede, ognuno di questi saggi dedicati a ciò che è accaduto filosoficamente dopo Parmenide (e composti dall’autore dopo il suo giustamente celebrato libro su Parmenide), costituisce un contributo prezioso alla comprensione di aspetti cruciali della filosofia presocratica. Ma, soprattutto nella sezione centrale del libro, è il loro insieme che consente di ricavare da questi studi una nuova e convincente prospettiva generale sulle (direi, una nuova e convincente narrazione delle) linee di sviluppo della prima filosofia greca. Ed è questo, a mio avviso, il valore principale di questa parte della produzione scientifica di Alexander Mourelatos.

Acknowledgments “PRESOCRATIC ORIGINS OF THE PRINCIPLE THAT THERE ARE NO ORIGINS FROM NOTHING” was originally published in The Journal of Philosophy, 78 (1981), 649-65. It is reprinted here with the kind permission of the journal. “‘NOTHING’ AS ‘NOT-BEING’: SOME LITERARY CONTEXTS THAT BEAR ON PLATO AND ALSO ON PARMENIDES” was originally published in Arktouros, Hellenic Studies Presented to Bernard M. W. Knox, edited by G. W. Bowersock, W. Burkert, and M. C. J. Putnam, Berlin and New York: Walter De Gruyter, 1979, pp. 319-39. It is reprinted here with the kind permission of Walter de Gruyter. An early version (here revised, adapted, and expanded) of “FORM AS SHAPE, ARRAY, ORDER: EARLIEST APPROACHES AND ELEATIC SCRUPLES” was published in Greek under the title “Ǿ ਕȞĮțȐȜȣȥȘ IJોȢ ȂȠȡijોȢ ıIJ੽Ȟ ʌȡȫȚȝȘ ਦȜȜȘȞȚț੽ ijȚȜȠıȠijȓĮ” [The Discovery of Form in Early Greek Philosophy], in ȆĮȜȓȝȥȘıIJȠȞ [Palimpsiston], 5 (1987), 93-114. “QUALITY, STRUCTURE, EMERGENCE: IN ANAXAGORAS, PHILOLAUS, DEMOCRITUS, AND IN EMPEDOCLES” was previously published in John J. Cleary, ed., Proceedings of the Boston Area Colloquium in Ancient Philosophy, vol. 2, New York: University Press of America, 1987, pp. 127-94. Reprinted here with some revisions, “THE CONCEPT OF THE UNIVERSAL IN SOME LATER PREPLATONIC COSMOLOGISTS” was previously published in Mary Louise Gill and Pierre Pellegrin, eds., A Companion to Ancient Philosophy, Blackwell Companions to Philosophy, Oxford: Blackwell Publishing, 2006, pp. 56-76. “DEMOCRITUS ON THE DISTINCTION BETWEEN UNIVERSALS AND PARTICULARS: THE RELEVANCE OF FRAGMENTS B124 AND B165” was originally published in Andreas Bächli and Klaus Petrus, eds., Monism, Philosophische Analyse 9, Frankfurt and London: Ontos Verlag, 2003, pp. 43-56. It is reprinted here with the kind permission of Walter de Gruyter. “INTRINSIC AND RELATIONAL PROPERTIES OF ATOMS IN THE DEMOCRITEAN ONTOLOGY” was originally published in Metaphysics, Soul and Ethics: Themes from the Work of Richard Sorabji, Ricardo Salles, ed., Oxford: Clarendon Press, 2005, pp. 39-63. It is reprinted here with the kind permission of Oxford University Press. “GORGIAS ON THE FUNCTION OF LANGUAGE” originally appeared in Gorgia e la sofistica, eds., Luciano Montoneri and Francesco Romano, (series) Siculorum Gymnasium, N. S. a.XXXVIII, nos. 1-2 (1985). It was subsequently republished in a revised and expanded version in Philosophical Topics, 5 (1987), 135-170. It is reprinted here with the kind permission of the editors of both the Siculorum Gymnasium series and of Philosophical Topics.

Abbreviations

DK, also Diels-Kranz: Diels, H. and Kranz, W. 1952. Die Fragmente der Vorsokratiker. 6th ed. revised by W. Kranz, 3 vols. Berlin: Weidmann. ENN: ex nihilo nihil, “nothing [comes to be] out of nothing.” LSJ: Liddell, H. G. and Scott, R. 1940. A Greek-English Lexicon. 9th ed., rev. H. S. Jones and R. McKenzie. Oxford: Oxford University Press. MXG: De Melisso, Xenophane, Gorgia (anonymous Peripatetic work, received as part of the Aristotle corpus). PSR: (the) Principle of Sufficient Reason. UM: Unmoved Mover (Aristotle).

Part I: Preliminaries and Fundamentals

1. Presocratic Origins of the Principle that There Are No Origins from Nothing

Even those who might question the truth of the ex nihilo nihil principle would readily concede that this principle itself could not have sprung from nothing. The origins are in Presocratic philosophy. But in what context, and with what antecedents or rationale? Aristotle was convinced that the principle was as old as philosophy itself. He frequently speaks of it as the “common assumption’’ (NRLQƝGR[D, or koinon dogma; cf. homognomousi hapantes) of all the physikoi, of all who wrote “on nature.” 1 This global attribution became standard in the Peripatetic tradition and is reflected in many modern interpretations of the Presocratics. 2 But the earliest text with a recognizable version of the ex nihilo nihil (henceforth ENN) is Parmenides DK28B8.7-10. Now since a concern with relations between Being and Not-Being (or Nothing) 3 is saliently characteristic of Elea, the question arises whether Aristotle’s attribution of ENN to philosophers before Parmenides may not be anachronistic. The suspicion of anachronism is reinforced when we consider the rationale for ENN Aristotle projects to his predecessors, all the way back to Thales. They held, Aristotle says, that “from what-is-not nothing could have come to be, because something must be present as a substratum” (Phys. I.8.191a30-31). The “because” clause here blatantly invokes Aristotle’s own triadic ontology of matter-privationform. If that is why Aristotle declares ENN to be pre-Eleatic in its origins and archaion, “ancient” (Metaph. I.984a27-984bl), why should we believe him? The antithesis to the Peripatetic thesis would tend to make high drama out of the story of early Greek philosophy. Before Elea spoke, birth, death, change, or indeed – in Heraclitus – flux, are metaphysically rampant. Then Parmenides proclaims ENN and immediately broadens it into a ban on all change. Melissus copies Parmenides, adding a few touches of his own. An austere and paraOriginally published in The Journal of Philosophy, 78 (1981), pp. 649-65, for the paper’s presentation at a “symposium session” at the meeting of the American Philosophical Association (Eastern Divion) of 30 December 1981. It is reprinted here with the kind permission of the journal. On the December 1981 occasion, David Gallop served as Commentator: see Gallop 1981. 1

Physics I.187a27-29 and 34-35, Metaphysics XI.6.1062b24-25; cf. Phys. I.8.191a24-31, Metaph. VII.7.1032b30-31. For the cognate view, that philosophers from Thales on have denied that the physis or archƝ comes to be or perishes, see Metaph. I.3.983b6-21, and 984a27-984bl. 2 See, e.g., Gomperz 1901, I, p. 171. See also Casertano 1978, pp. 100 and 248-49. 3 The terms PƝ HRQ and PƝGHQ are often used interchangeably by the Presocratics, as are also the terms ouk eon and ouden. See below, Ch. 2 “‘Nothing’ as ‘Not Being’”.

22

Alexander P. D. Mourelatos

doxical monism is the Eleatic upshot. Repelled by Eleatic paradoxes, yet thoroughly captivated by Elea’s ENN principle, the non-Eleatic philosophers of the fifth century variously and ingeniously attempt to “save the phenomena” of diversity and change. My statement of the antithesis is compendious to the point of caricature. But the reader will perhaps recognize in it several distinctive themes found in widely read accounts of Presocratic philosophy. 4 Without subscribing to Aristotle’s projected rationale of ENN, I propose nevertheless to rehabilitate Aristotle’s thesis. My approach, in the main, will be indirect. I shall analyze texts in Parmenides, Melissus, and Empedocles that give us the earliest statements of ENN. I shall be looking for arguments or other clues to rationale. From the comparison between Parmenides and Melissus it will emerge that there is no monolithic Eleatic argument or rationale for ENN. Yes, there is a uniquely Parmenidean argument; but already in Parmenides there is an alternate and ancillary argument that betrays ancient roots. In Empedocles we find variants of that alternative, as well as evidence of a religious-affective rationale that also has ancient roots. This will not be a complete story of the origins of ENN, but I hope enough will be said to clear the way for renewed appreciation of the tenor of Aristotle’s thesis. My concern is not to vindicate Aristotle but to bring out conceptual connections and implications in Presocratic fragments. I Parmenides’ formulation of ENN is embedded in these lines: For what birth will you seek of it? having grown how [or “whither”]? from what? Neither [or “nor”] will I allow you to say or to think “from (ek) what-is-not.” For it is not sayable or thinkable that it is not. Besides, what requirement has impelled it, later or sooner [or “later rather than sooner”], having started from nothing, to grow into being (phyn)? So, it is right either that it should altogether (pampan) be (pelenai) or not (at all). Nor will the hold of trust permit that out of what-is-not [or “out of what-is”] may come to be something beside itself. (B8.6-13)

Parmenides’ statement of ENN comes across clearly in spite of the textual uncertainties indicated by the alternative translations. The statement is formed by the opening three rhetorical questions together with the injunctive assertion that immediately follows. Lines 12-13 (last sentence in the translation) may be a reiteration of ENN, if we retain the manuscript reading HNPƝHRQWRV “out of what-is-not.” With the emendation ek tou eontos, “out of what-is,” which is 4

See, e.g., Stokes 1971, esp. pp. 31-36 and 253-55; and Barnes 1979a, vol. II: Empedocles to Democritus, esp. pp. 11-15 and 130-40.

1. Presocratic Origins of the Principle that There Are No Origins from Nothing

23

favored by most editors, the lines may be interpreted as constituting the statement that there is no coming-to-be out of what-is, either. 5 It has often been observed that Parmenides offers two distinct arguments here in support of ENN. The most perceptive differentiation of the two arguments remains that by G. E. L. Owen. 6 The first argument is that we cannot speak or think of generation from what-is-not simply because “it is not sayable or thinkable” that X (the subject in question) should not be. This invokes the characteristic Parmenidean prohibition of the route of “X is not.” It is a matter of controversy precisely what the rationale or the full scope of that prohibition is. But it suffices for us to note that the genre of the rationale laid out in B2-B7 is epistemological-semantic, and that, in the course of the deductions in B8, not just generation ex nihilo but coming-to-be as such, perishing, and change generally, are shown to involve impermissible recourse to “what-is-not.” In effect, in the light of the epistemological-semantic rationale, ENN in Parmenides is hardly a principle; it is a special application of a sweeping ban on coming-tobe. The second argument Parmenides offers for ENN is quite different. Marked off from the first argument by de, “besides,” the second is implied in this rhetorical question: “What . . . has impelled it (ǀUVHQ) . . . to grow into being?” The causal force of “impelled” is here deliberately ironic, matching the effect produced by the use of “having grown” in the first argument. Certainly the hypothesis that is for the moment being entertained of a generation ex nihilo directly excludes causes. The subject of “impelled” is, in fact, not a thing or agent but an abstraction, a “requirement” (chreos). As Owen saw, Parmenides is invoking not the Principle of Causality but the Principle of Sufficient Reason – abbreviated PSR henceforth. 7 This is shown by two details in the text. I focus first on the noun chreos, “requirement.” This is a cognate of the impersonal expressions FKUƝ and chreon esti. The latter are Parmenides’ favorite modal devices for conveying the logical propriety that attaches to the route “is” and to the goddess’s discourse, or, when negatived, the impropriety that attaches to the route “is not” and to mortal doxai, “opinions.” Modern readers of Parmenides are often put on the wrong track by the translations 5

See Stokes 1971, pp. 131 and 253-55. For another interpretation of the emended text, see below, note 11. 6 Owen 1966, pp. 325-29. 7 Cf. Santillana 1967, p. 63. Cf. also Barnes 1979a, I, pp. 187-88. Barnes in the end decides that PSR is not involved. His “simpler” interpretation of lines 9-10 is that “if O does not exist at t, then nothing can ‘rouse’ it into existence, for it is not there to be ‘roused’” (188). In adopting the weak translation “at t” or “ever” for K\VWHURQƝSURVWKHQ Barnes follows Stokes 1971, p. 254. Both scholars overlook the fact that polar disjunctions typically have pregnant use in Parmenides: cf. B8.11, 8.16, 8.36, 8.45 (the point is brought out nicely in Austin 1979, ch. 2, passim). The suggestion conveyed by such disjunctions is of significant contrast and momentous choice. The Stokes-Barnes translation weakens K\VWHURQ Ɲ SURVWKHQ to the point of making it otiose: “What necessity would ever have caused it to grow?” (Stokes 1971, p. 254). See also below, note 10. For a more recent analysis of Parmenides B8.6-13, see Mourelatos–Pulpito 2018.

24

Alexander P. D. Mourelatos

“need” for chreos and “is necessary” for the impersonal modals. It can be shown from the use of chre- words that justification and intelligibility lie at the heart of the modality these words envisage. 8 That Parmenides is not in our passage using chreos loosely is shown by his following it up with FKUHǀQHVWL, “it is right,” in the immediately succeeding assertoric corollary (line 11) of the rhetorical question asked. The other telltale sign of appeal to PSR is the disjunction “later or sooner.” 9 The argument is familiar: Suppose X came to be 10,000 years ago. Why not, say, 10,001 years ago? or, why not 9,999 years ago? But it cannot have come to be at two times. And since the demand for sufficient reason can be used over and over again to discredit not only our original supposition but also all alternatives to it, it follows that we have no sufficient reason for selecting any time as the time at which X comes to be. Parmenides gives the upshot of the argument in the disjunction: “Either it is altogether or not (at all).” 10 There is, unfortunately, some uncertainty as to the import of the first disjunct. For, in fact, the conclusion that there is no time at which X comes to be entails a triple disjunction: either X is not; or X always is; or X atemporally is. This thorny issue we can skirt; its resolution does not bear on our theme of the origins of ENN. Brief exploration of another point, however, will have repercussions when we come to study Empedocles. Note that – even apart from Eleatic arguments against the reality of place – this PSR-type argument for ENN could not be offered if the relevant determination is place rather than time. Place, as generally understood by the ancients, is determined by the presence of a body other than the body in place. So, if we should ask, “Why should it have come to be at place l1, rather than at l2?” the supposition of a generation ex nihilo would be immediately contradicted. It would not have been as obvious – perhaps it is in fact false – that positing time would contradict equally the supposition of a generation ex nihilo. In comparison to the first argument, the second is characterized in Owen 1966 as “more natural and charitable,” also as “conventional” and “very Greek” (pp. 329, 325). To suppose that the first argument had occurred to any philosopher before Parmenides would be gratuitous. But matters are different 8

See Mourelatos 1970/2008a, pp. 277-78. Cf. Mourelatos–Pulpito 2018. Cf. Aristotle, De caelo I.12.283a11-12. For other relevant parallels, see Owen 1966, pp. 325-26. Corresponding to the two possible construals of K\VWHURQƝSURVWKHQ, viz., “later or sooner” and “later rather than sooner,” Owen distinguishes and discusses two versions of the argument. On philological grounds, the second construal seems unlikely: see Stokes 1971, p. 340, n. 1. 10 Barnes 1979a objects that sufficient reason can, after all, be given why X, or “O” in his notation, should come to be at a select time: “at t1 something may hold of it which does not hold at t2 – the demiurge may have determined upon t1 as the appropriate time to create O” (p. 188). It is precisely this recourse to the arbitrary fiat of a demiurge that arguments via PSR, as classically conceived, are intended to block. Besides, when Parmenides says WRXPƝGHQRVDU[DPHQRQ(B8.10), a demiurge is excluded along with causes of any sort. 9

1. Presocratic Origins of the Principle that There Are No Origins from Nothing

25

with the second argument. Each of its two major components is attested before Parmenides. Almost certainly Anaximander’s apeiron has no temporal origin. The same can be said of Anaximenes’ air, or of Xenophanes’ one God. Recall, too, Heraclitus’ kosmos which “no man or god did create, but it ever was, is, and shall be” (DK22B1). The second component, sufficient reason, is built into the demand for intelligibility in explanation which is part of the quarrel between nascent philosophy and the tradition of poetry and myth. In any event, PSR is conspicuous in an important context of Milesian cosmology: it was employed by Anaximander to make intelligible the position and stability of the earth (DK12A11, A1, A26). A natural complement of ENN is the principle that nothing is ever annihilated, nothing is reduced to nothing. It is a curious feature of Parmenides’ account – and, as we shall see, of Melissus’ account as well – that the complement is not supported by a separate argument. Indeed, it is not certain that the complement even gets a separate statement. 11 Of the many explanations of this asymmetry, the one that still appeals to me most is that Parmenides regards it as so obvious that, when we say “X has perished” we mean “X has ceased to be,” i.e., “X is not,” that he deems it unnecessary to argue the point. This attitude is revealed in the rhetorical asymmetry of B8.21: “Thus is coming-to-be quenched and the unheard of (apystos) perishing.” There is, however, a significant difference in logical strength between ENN and its annihilation complement. The latter carries a heavier burden of proof. For there are two ways in which a given X might be annihilated: it could just perish, drop out of existence, cease to be; or it could be destroyed through the causal influence or agency of something else. To dispose of the latter possibility is no easy matter – as is evident from Plato’s long search for a cogent proof of the immortality of the soul. When we entertain the supposition of a generation ex nihilo, the supposition itself tells us that causal agency has been excluded. 12 But when we entertain the supposition that X might be annihilated by a destroyer, no limits are set on the imagined destroyer’s power and efficacy. In effect, the principle that nothing is ever annihilated corresponds strictly not to ENN but to a principle that nothing can come to be out of anything. Parmenides, of course, has no difficulty certifying either this stronger version of ENN or the annihilation complement. But any philosopher who does not base his argument on the injunction against recourse to “what-is-not” will have a hard time arguing for the annihilation complement. We shall see that, 11

Diels-Kranz read epeit’ apoloito at B8.19, but that is an emendation of the manuscript reading epeita peloi to, which is what nearly all recent editors have adopted. If the emendation ek tou eontos for line 12 is right, then Barnes offers us the possibility of an indirect separate statement (with just the faint whiff of an argument): Barnes 1979a, I, p. 185; cf. Stokes 1971, p. 132. Otherwise we have only the two QEDs that pair coming-to-be with perishing (B8.13-15, and B8.21). 12 Note that the possibility envisaged is not that of a demiurge creating things out of no pre-existing matter. Cf. above, note 10.

26

Alexander P. D. Mourelatos

when arguments against annihilation are offered, they deal with the “easier” of the two possibilities of annihilation, just plain perishing. II Let us next turn to Melissus. ENN appears explicitly, and in a formulation that comes close to the perennial Latin version, at the end of DK30B1: There always was whatever was, and it always will be. For if it came to be, then it is necessary that before it came to be it was nothing (PƝGHQ); and if it were nothing, in no way (oudama) could anything (ouden) come to be out of nothing.

The annihilation complement is not expressed, but it must somehow be implicit. For it is needed to provide the second half of the conclusion – which is reiterated in the opening sentence of B2 – that “it will always be.” At a later stage of the argument, having established that X (the bare subject of the deduction, as in Parmenides) is “one and all of it the same (homoion)” (B7[1]), Melissus seems to offer a separate proof against annihilation. He announces that X “could not perish” (B7[2]), and immediately appends four other denials of forms of change. He indicates what his general strategy of proof is: “If it were to undergo any of these, it would no longer be one.” This proof strategy is, in fact, never deployed against “perishing.” The closest Melissus comes is in this argument against alteration: “For if it alters (heteroioutai), it is necessary that what-is should not be the same (homoion), but rather that what was previously should perish and that what-is-not should come to be.” Ostensibly it is the double disruption involved in alteration that violates sameness, and hence unity; we are not shown that coming-to-be all by itself or perishing all by itself would violate unity. 13 As in Parmenides, the annihilation complement to ENN is presented obliquely and allusively. Embedded in B7(1)-(2) is a statement of this proposition, or perhaps more aptly, of this rule of linguistic transformation: 14 If X, from being F, becomes G, then X’s G-ness has come into being and X’s F-ness has perished. The section that follows, B7(3), gives the rule’s corollary: If X, from having arrangement Q, comes to have arrangement R, then X’s R-ness has come into being and X’s Q-ness has perished. For our theme this Melissan rule is significant in two ways. It shows, first of all, an essential agreement between Parmenides and Melissus with respect to the scope of the denial of genesis: the objection is not just against birth-death, but against all conceivable forms of change. But the 13

Barnes’s observation (1979a, I, pp. 216 and 181), that “[the proposition ‘O is not destroyed’], in Melissus, is only established by way of [‘O is unique’],” is not quite right. 14 The significance of the rule is stressed and brought out effectively by Barnes (1979a, I, p. 216; II, pp. 140 and 130).

1. Presocratic Origins of the Principle that There Are No Origins from Nothing

27

rule also points up an important difference between the two thinkers. For Parmenides the rule would be otiose. The crucial point for him is not so much that alteration, motion, transfiguration reduce to birth and death – or, for that matter, to disunity – it is rather that all forms of change, including birth and death, involve recourse to “what-is-not.” The difference just noted is symptomatic of a fundamental difference between the two Eleatics. Melissus makes no general use of Parmenides’ injunction against speaking or thinking of what-is-not, let alone of the epistemological-semantic considerations that motivate the injunction. This difference is conspicuous not only in the two philosophers’ respective arguments against change, but in their respective arguments for monism. Assuming that Parmenides does argue for some version of monism, 15 what makes that argument possible is, almost certainly, a rejection of contrariety or of difference; and this rejection is a special application of the injunction. In Melissus, by contrast, monism is the consequence of an intermediate deduction, indeed one that is notoriously precarious and un-Parmenidean, viz., that X is apeiron, “infinite” (B3-B6). Now even if Melissus did not have Parmenides’ strategy of attacking all forms of change by invoking regularly the injunction on recourse to “what-isnot,” might he not still have invoked the injunction occasionally and ad hoc? He gives at least the appearance of doing so in two texts. In B7(7) we find this argument: “Nor is there any void; for void is nothing (ouden estin), and a ‘nothing’ (WRJHPƝGHQ) could not be.” Other translations are possible, 16 but it is the translation just given that makes the strongest case for the view that Melissus is here invoking the Parmenidean injunction. The other text is the opening of Simplicius’ paraphrase of the main Melissus fragments. Several scholars are convinced that this opening is an actual quotation: “If, to begin with, there is nothing, what can be said of it as of something that is (KǀV RQWRV WLQRV)?” 17 If this is a quotation, or even a reliable paraphrase, it might do more than furnish another instance of the occasional use of the Parmenidean injunction; it could furnish a model for reconstructing Melissus’ Parmenides-inspired rationale for ENN. To take B7(7) first: Melissus’ point is simply that, if void does not exist, it cannot be an attribute of any part of X. No injunction against speaking or thinking of what-is-not is issued. Melissus is not invoking the “much-contending challenge” spoken by Parmenides’ goddess; he is uttering a truism that is mediated by the widely current assimilation of “nothing” to “what-is-not” (see 15

I am not persuaded by Barnes (1979a, I, pp. 204-13, and 1979b) that Parmenides is in no sense a monist; but I do agree with Barnes that he is not the sort of monist Melissus is. See Mourelatos 1981a, pp. 70-71. 16 See Barnes 1979a, I, pp. 214 and 218 with n. 22. 17 Simplicius, In Phys. 103 (reproduced in Diels-Kranz 1952, vol. 1, p. 268). See Burnet 1957, p. 321; Reale 1970, p. 368; Vitali 1973, p. 130. Cf. Barnes 1981, pp. 78-79.

28

Alexander P. D. Mourelatos

above, this Chapter, note 3). The first sentence of the Simplicius paraphrase is closer to the Parmenidean approach. Still, compared to the urgency of Parmenides’ “thou canst not come to know what-is-not” (B2.7-8), the rhetorical question purportedly asked by Melissus seems complacently dismissive. When read in its full context in Simplicius, the Melissan question has this force: What’s the use of considering the possibility that there is nothing? What real attributes (KǀVRQWRV) might we assign to this situation of utter cosmic nothingness? One can ask this sort of question without being in the least inclined to forbid references to what-is-not. What, then, is Melissus’ rationale for ENN? The evidence suggests that he would not have been particularly moved by the first of Parmenides’ two arguments. To suppose that he was moved by Parmenides’ second argument would be gratuitous. Besides, that second argument itself points to pre-Eleatic concerns and themes. The honest answer is that we do not know Melissus’ rationale for ENN. Still we should not miss two clues. The fact that Melissus cites ENN without proof, either at the start or very near the start of his treatise, and the fact that ENN from that point on provides the indispensable basis for all his deductions – these two facts suggest that he counts on his readers’ and hearers’ easy acceptance of ENN, as of something familiar and obvious. Simplicius, adhering to the Peripatetic interpretation, speculates that “with respect to generation and perishing, Melissus made use of the postulates (D[LǀPDVLQ) of the physikoi (loc. cit.). That it was Parmenides who provided, in a general way, the inspiration and the model for Melissus’ deductions there is no doubt. What is dubious is that he provided specific, identifiable premises. What seems to lie behind Melissus is not the Parmenides of the “much-contending challenge” but a Parmenides naturalized in Ionian Samos. III The combined statement of ENN plus its annihilation complement appears for the first time in Empedocles DK31B12.1-2: There is no contriving (DPƝFKDQRQ) of a coming-to-be out of whatis-not, and it is not feasible (DQƝQ\VWRQ) and not to be heard of (apyston) that what-is should utterly perish. 18

A closely related statement is B11: Fools, for they do not have thoughts that reach far, who expect that what formerly was not should come to be, or that something should die and altogether perish. 18

The general sense conveyed by this translation of lines 1-2 is secure in spite of several textual problems. For line 3, see below.

1. Presocratic Origins of the Principle that There Are No Origins from Nothing

29

Also apposite is B17.30-33: And nothing supervenes on these things [the four elements and two forces], nor does [any one or part of them] fail (DSROƝJHL). For if they kept perishing (ephtheironto) through and through (diamperes), they would no longer be. And what might augment this All? having come from where? and how could [this All, or something] utterly perish (exapoloito), since nothing is bereft of these things? 19

The rhetorical questions here are strongly reminiscent of those in Parmenides’ ENN passage (B8.6-7). Indeed, there are many echoes of Parmenidean phrasing in all three of the texts cited. But of Parmenides’ first argument for ENN there is only a faint hint: B12.2 apyston, “not to be heard of,” with reference to the annihilation complement, which recalls Parmenides’ B8.21 apystos olethros, “perishing not to be heard of.” The hint is faint because, even assuming that apyston is the correct reading, 20 it gets no support either from its context or from the two other cognate texts. To the contrary, we find evidence that Parmenides’ injunction on recourse to “what-is-not” had made no impression on Empedocles. In B17.31 we find Empedocles arguing: “If they kept perishing . . . they would no longer be.” The context shows clearly that the relevant denial of the consequent is: “But they still are.” The relevant denial for Parmenides would have been: “But you cannot say or think that they would no longer be.” Deployment, or even mention, of Parmenides’ injunction against “what-is-not” is not found in the rest of Empedocles’ fragments either. True, in B8 and B9 we have Empedocles’ famous critique of the ordinary language of birth and death, but the contrast with Parmenides is, even there, more significant than the similarity. In Parmenides, certain epistemological-semantic considerations lead to ontological conclusions; in Empedocles, it is the ontological pronouncements that provide grounds for the disparagement of ordinary language. The movement of thought is exactly the reverse. Does Empedocles have another rationale for ENN or for its annihilation complement? There is at least the rhetoric of an argument at B17.31-33. The logical content is, however, disappointing. We are told that “nothing supervenes” because the set of four elements and two forces jointly constitute “the All.” Now we are not entitled to declare that set closed unless we are already assured of the truth of ENN. Correspondingly, to suppose that perishing would leave a gap, whereas “nothing is bereft of these things,” appeals to a tacit premise that no gaps can ever develop, which in turn can be established only 19

Cf. B13, B14. The imperfect ephtheironto has the aspectual nuance of temporal distribution. The adverb diamperes must have the corresponding sense of spatial distribution. 20 The reading apyston in Empedocles is, in fact, a modem emendation that attempts to make sense of the nonsensical apauston, “unceasing,” which appears in one of our two sources. The reading DSUƝNWRQ“not to be done,” found in our other source, is redundant after DQƝQ\VWRQand thus philologically a lectio facilior.

30

Alexander P. D. Mourelatos

by appeal to the annihilation complement. Finally, the argument that “if they kept perishing through and through, they would no longer be” seems, on the face of it, irrelevant. What we need is proof that there is not even a single occurrence of the perishing of a tiny bit. We shall, however, shortly find evidence that, for Empedocles, the occasional perishing of a bit entails perishing that is spatially widespread and temporally endemic. So at least this one argument may be rehabilitated. I have so far omitted translation of the third line of Empedocles B12. That line, which is a justificatory clause for at least the annihilation complement, will repay careful study. Certain grammatical problems argue against our accepting the line in the form in which it has been preserved in our manuscripts. 21 The best defense of the received text yields this translation: 22 For it will always be posited wherever one may at one time lean (B12.3).

Assuming that the “leaning” takes place in a vain attempt to extirpate “whatis,” the point would be: in destroying something one must lean on something else. Yet this, too, hardly gives Empedocles a cogent reason in support of the annihilation complement. He has just claimed that no what-is is possibly destroyed; he cannot support this with evidence that some what-is must be preserved to serve as Archimedean fulcrum. Besides, unless one obtains the allusion to a destroyer out of a glossing of HUHLGƝL, which would play up aggressive connotations of “lean,” B12 as a whole would not suggest that a destroyer is being envisaged. In B12.1-2 the formulation is precise and symmetrical. As though to make it doubly clear that the case of coming-to-be is truly ex nihilo, involving no antecedents whatsoever, Empedocles employs an impersonal construction. In the corresponding case of annihilation, the construction is not passive but middle. The only entity mentioned is a single “what-is,” which seems to rule out a separate destroyer. 23 As we saw in connection with Parmenides, a causal proof against all perishing, one that would eliminate all possible destroyers, is inherently more difficult. Empedocles limits himself to the more manageable task of arguing against the exact counterpart of generation ex nihilo, viz., that something should simply “perish” spontaneously, out of its merely being. Nearly all editors have opted for a simple and highly attractive emendation of B12.3 which was proposed by Friedrich Panzerbieter more than a century ago. Instead of reading, with the manuscripts, WKƝVHVWKDL, “be posited,” which 21

Įੁİ੿ Ȗ੹ȡ șȒıİıșĮȚ ੖ʌૉ țİ IJȚȢ Įੁ੻Ȟ ਥȡİȓįૉ. The place of the verb in the main clause is occupied by the infinitive WKƝVHVWKDLThis is tolerable in the indirect discourse employed by the author of our source for the fragment; but it yields a harsh “absolute” use of the infinitive when the line is read independently. Meter precludes a simple adjustment to the corresponding finite form. The required passive construal of the middle form of the verb would be awkward even with the correction theseitai proposed by Bollack (1969, III, p. 83). 22 Cf. Bollack 1969, II, pp. 22-3; III, pp. 80-84. 23 Cf. B11. The ephtheironto of B17.31 also has middle sense, since it explicates DSROƝJHL, “fail.”

1. Presocratic Origins of the Principle that There Are No Origins from Nothing

31

would require a quite awkward construal of this infinitive, we should read: WƝL g’ estai. The translation is: For it will nonetheless (ge) at any time be (estai) there (WƝL), wherever one may at one or another time set it. 24

The rich philosophical potential of the Panzerbieter text and of the translation just given have not been fully realized. Many modern readers are content with the condensed translation, “For it will always be wherever one may set it.” The point cannot, of course, be that things will stay forever where they may be initially placed; rather, that we can generally count on finding things again where we have at any earlier time placed them – provided no one else has meanwhile placed them elsewhere. In effect, Empedocles would be saying that, if things could simply drop out of existence, the world would be thoroughly unreliable. Even simple references involving nothing more than the displacement of objects would be in doubt. Upon not finding my cloak where I left it, why should I search for the culprit who perhaps made off with it? or why should I search my own memory for other places where I may have mislaid it? There would always be the possibility that the cloak simply went out of existence. (It goes without saying that I could not search for a cause of its destruction – fire, for example – since the supposition I am entertaining is precisely that at any time and in any place a thing may perish even in the absence of a destroyer.) This pragmatic argument easily suggests a parallel argument for ENN: “Things will never show up where no one has previously set them.” If I put a single drachma in my purse, I had better not expect to find ten drachmas when I next reach into it. There is, however, a flaw in the argument in either version. In order for us to have reliable expectations as to what we may find in our environment it is not necessary that no spontaneous generation or annihilation should ever take place; it would suffice if such occurrences were merely infrequent. Could this flaw be overcome? It could if we introduced a PSR-type premise: if it should happen here and at this time, it should happen at all times and everywhere. Now, as soon as we consider this supplement to the argument, the possibility strikes us that B12.3 may itself constitute an argument via PSR. The condensed translation, “For it will always be wherever one may set it,” suppresses a striking rhetorical feature of the Panzerbieter text: the adverbs of time and place are doubled up in chiasmus, “at any time (aiei) . . . there (WƝL), wherever (KRSƝL) . . . at one or another time (aien).” 25 What is the point of this emphasis on the ideas of place and time? Let me attempt an answer through an 24

Cf. Burnet 1957, p. 207: “for it will always be, wherever one may keep putting it.” Burnet rightly captures the aspectual effect of aien + present subjunctive HUHLGƝL – a detail often overlooked. Empedocles certainly does not say SRWHHUHLVƝL(aorist). 25 The distributive sense of aien is brought out by the verb aspect of HUHLGƝL; see preceding note.

32

Alexander P. D. Mourelatos

imaginary Empedoclean meditation. Suppose a thing has perished on the autumnal equinox at Acragas. The exclusion of causal antecedents and causal agents does not exclude a demand for reasons. The only relevant determination would be the time and place of the occurrence. But why should the thing in question have perished in Acragas on that day and not when it was in Syracuse on the previous summer solstice? Since it cannot have perished first in Syracuse and then again in Acragas, it follows that it has perished in neither place: so, it survives at both places and times, and similarly for any pairs of places and times we care to choose. Here is a precise statement of what Empedocles is saying in B12.3: With respect to any times t1 and t2 and places l1 and l2, if a thing X exists in l1 at t1, then, if X were set in l2 at t2, X will equally well exist there and then.

The argument can be made still more general. Assuming always that causal considerations are excluded and that the only relevant conditions are place and time, if our X is a tiny bit of one of the four elements, say, earth, then if the presence of that bit of earth in place l1 at time t1 is reason for its perishing, then there would be reason for the perishing of all other bits of earth situated at any other place at time t1. Moreover, there would be reason for the perishing not just of this bit of earth but equally of all bits of any of the four elements and two forces at their corresponding places at this or any other time. To state the generalization in language that shows some Empedoclean color: If a given place and time is baneful to anything, then all places and all times are baneful to everything. Now we see why at B17.31 Empedocles says, “If they kept perishing through and through, they would no longer be.” To suppose the isolated perishing of a tiny bit is tantamount to supposing perishing that is persistent and rampant. If the interpretation I am proposing is right, Empedocles not only has converted Parmenides’ PSR-type argument into a corresponding argument against annihilation, he has also embellished the argument by extending the application of PSR to the determination of place. Why he should have done this embellishing is not clear; a strictly temporal PSR-type argument against annihilation would have sufficed. One guess might be that he thought – wrongly – that, by arguing that no place (as well as no time) is more baneful than any other, he was, in effect, procuring a defense against Eleatic attacks on the reality of locomotion. What is more immediately relevant to our concerns here is that, with the embellishment retained, the argument cannot be reconverted into an argument for ENN. For, as we saw in connection with Parmenides, if generation is truly ex nihilo, we cannot even speak of the place in which the thing comes to be. The syntactical connection between B12.3 and B12.2 makes

1. Presocratic Origins of the Principle that There Are No Origins from Nothing

33

it, in any event, clear that Empedocles intended the argument to support only the annihilation complement. Let us take stock. Empedocles is the first to state both ENN and its annihilation complement, and he is the first to offer distinct arguments against annihilation. Indeed, the most attractive reading of B12.3 yields an argument that could not be adapted into an argument for ENN. This emphasis on the annihilation complement contrasts sharply with what we found in Parmenides and Melissus. We can immediately understand this difference of emphasis if we recall that Empedocles is the author not only of the poem “On Nature” but also of the religious poem “Purifications” which preaches a doctrine of transmigration. One of the texts that proclaim the twin principles of ENN and No Annihilation, B11, caustically addresses “fools” who “do not have thoughts that reach far.” It is likely that B11 comes from “Purifications,” as does probably this other text which is associated with B11 by Plutarch, our source for both: 26 A wise man would not in his mind anticipate such things, that for as long as mortal men (brotoi) live what they call a lifetime, for that time they are and share in good and ill, but before they were formed and after they have been decomposed they are nothing (B15).

The mental myopia of B11 and the failure of anticipation of B15 both refer to the predicament of mortal men who do not look beyond the parochial limits of their personal death and birth. Both of these texts should, in turn, be read with B117, which certainly does come from the religious poem: For already I have at one time been born (JHQRPƝQ) a boy and a girl, and bush, and bird, and mute fish of the sea.

Studies of Empedocles in the second half of the twentieth century have shown that the two poems, far from being irreconcilable, are in fact inextricably connected. As M. R. Wright has put it, “The Physics and the Katharmoi break down the division between men and long-lived gods, and between plants and animals and men, and as a corollary to this they question the accepted frontiers of birth and death” (Wright 1981, p. 61, emphasis mine). 27 So the allusions to Parmenides, even the invention of arguments invoking PSR – all this is essentially ancillary. Empedocles’ doctrine concerning coming-to-be and annihilation has its origin in a religious or affective response to the universe. First comes a strong psychological denial of the reality of death; this is soon paired with intimations of pre-existence; the traditional frontiers of birth and death break down along the lines suggested by Wright; and at the end we have the

26 27

See Wright 1981, pp. 268-69. Cf. Solmsen 1975, pp. 123-24 and 144.

34

Alexander P. D. Mourelatos

positing of four elements and two forces that are not just “long-lived” but absolutely without birth and death. The first two of the three steps just traced had been taken long before Empedocles: not by Parmenides, of course, but by Pythagoras. And it is probably Pythagoras that Empedocles acknowledges as his ultimate inspiration, in B129. Significantly, the man anonymously praised in that fragment is a foil for the mortals of B11 and B15 who are beset with mental myopia. Empedocles’ hero is “endowed with the longest range of mind (cf. PƝNLVWRQ SUDSLGǀQ    plouton),” a man who could “mentally reach across ten and twenty human lifetimes to descry each and all the things that are (KHNDVWRQWǀQRQWǀQSDQWǀQ).” IV Other arguments for ENN or for its annihilation complement are possible and were eventually offered. The exclusion of causes envisaged by spontaneous generation and spontaneous annihilation could be made directly the brunt of a reductio argument. Thus Epicurus will argue that “nothing comes to be out of what-is-not; for otherwise any thing would come to be from anything without the need of seeds,” 28 and Lucretius will elaborate with six proofs for ENN and four for the annihilation complement (I.150-214, I.215-264). The common strand in all Epicurean-Lucretian proofs is that denial of the twin principles would posit a world of no certainty and no regularity. 29 As we saw in connection with Empedocles B12.3, this pragmatically unwelcome consequence can be secured only via this premise: that what might spontaneously happen at one place and time should happen everywhere and at all times. In the Epicurean tradition, too, PSR is the ultimate rationale. The account I have offered bears in important ways on a much-discussed topic of “answers to Elea” by the “Neo-Ionians” – as Barnes very aptly calls the non-Eleatic philosophers of the fifth century. These Neo-Ionian answers have sometimes been judged harshly: a “complete logical failure” (Stokes 1971, p. 35); a “flop.” 30 The depreciatory judgment assumes that the message of Elea was, in most respects, clear, unequivocal, and delivered in logical unison by the three Eleatics. It assumes further that Elea sets the measure for NeoIonian philosophic performance. We often allow – and notably in the case of Plato and Aristotle – that a philosopher has his own philosophical purposes, and so has the right to disambiguate, formulate, interpret, and appropriate the sayings of his predecessors. Why should each of the Neo-Ionians not have the same hermeneutic right vis-à-vis Elea? Indeed, why should Melissus and Zeno

28

Letter to Herodotus 38. Cf. Solmsen 1977, pp. 274-77. 30 Barnes 1979a, II, p. 140; cf. pp. 11-15, 125-145, and I, p. 216. 29

1. Presocratic Origins of the Principle that There Are No Origins from Nothing

35

not have such a right vis-à-vis Parmenides? 31 I have argued that, in the specific case of ENN, Empedocles and Melissus reached behind Parmenides to a fundamental pre-Eleatic demand for intelligibility, and that Empedocles – furthermore or in the first instance – responded to a yearning for personal immortality and universal affinity. Either consideration must have seemed more compelling than Parmenides’ epistemological-semantic rationale. Perhaps it was only against such backgrounds, respectively, that they could make sense of Parmenides. Given the obscurities and ambiguities of Parmenides’ argument generally, given his two distinct arguments for ENN, it would be hardly fair or perceptive to charge that Melissus’ and Empedocles’ reading of Parmenides is tendentious – or, for that matter, to object that the charge is implied in the interpretation I have developed here. In any event, a properly critical evaluation of Empedocles’ answer to Elea, one that starts by noting differences between Parmenides and Melissus, will surely be too complex to be helpfully summed up with either a “fail” or a “pass.”

31

For an account that brings out differences between Melissus and Parmenides, see Solmsen 1969; for Zeno and Parmenides, see Solmsen 1971.

2. “Nothing” as “Not-Being”: Some Literary Contexts That Bear on Plato and also on Parmenides

It has often been noticed that Plato, and before him Parmenides, assimilates “what is not” (WRPƝRQ) to “nothing” (PƝGHQor ouden). 1 Given that the central use of “nothing” has important ties with the existential quantifier (“Nothing is here” = “It is not the case that there is anything here”), it has widely been assumed that contexts that document this assimilation also count as evidence that both within them and in cognate ontological contexts the relevant sense of “being” or “to be” is that of existence. That this assumption is not to be granted easily, has been compellingly argued by G. E. L. Owen. 2 His main concern was to show that the assumption is particularly mischievous in the interpretation of Plato’s Sophist, where he found it totally unwarranted. My own concern is to attack the assumption on a broader plane. “Nothing” in English has uses that do not depend on a tie with the existential quantifier. So too in Greek: PƝGHQ or ouden can often be glossed as “what does not exist”; but not infrequently it is more appropriately glossed as “not a something” – or in Owen’s formulation, “‘what is not anything, what not-in-any-way is’: a subject with all the being knocked out of it and so unindentifiable, no subject.” 3 In fact, the assimilation of “what is not” to “nothing” may – in certain contexts – work in the opposite direction: not from “nothing” to “nonbeing” in the sense of non-existence; rather from “non-being” as negative specification or negative determination to “nothing” as the extreme of negativity or indeterminacy. To convey the sense involved in this reverse assimilation I borrow Owen’s suggestive translation “not-being” for mƝ on, a

Originally published as “‘Nothing’ as ‘Not-Being’: Some Literary Contexts That Bear on Plato” in Arktouros, Hellenic Studies Presented to Bernard M. W. Knox, edited by G. W. Bowersock, W. Burkert, and M. C. J. Putnam, Berlin and New York: Walter De Gruyter, 1979, pp. 319-39. – Repr. in J. P. Anton and A. Preus, eds., 1983, Essays in Ancient Greek Philosophy, Volume Two, Albany: State University of New York Press, pp. 59-69. – It is reprinted here with the kind permission of Walter de Gruyter. 1

See Parmenides DK28B6.2, cf. B7.1, B8.7-13, B9.4; Plato Resp. 478B12-C1, Thaet. 189A10, Soph. 237C7-E2. Cf. Owen 1971, pp. 225-27. 2 Owen 1971, pp. 241-48 and passim. For use of this assumption in interpreting Parmenides, see Furley 1973, p. 12. 3 Owen 1971, p. 247.

2. “Nothing” as “Not-Being”: Some Literary Contexts

37

rendering which makes use of an incomplete participle, rather than the complete gerund, of the verb “to be.” To make alternatives clear, I begin with an exploration of the relevant English uses. The semantic spectra of “nothing” and of the personal-pronoun form “nobody” are, of course, wider than logic books might suggest. Only two uses concern us, and these can be quickly formulated if we take the expressions “nobody” and “nothing” as answers to two types of questions. (a)

1. Who is in the house now? – Nobody. 2. What is in the box? – Nothing.

(b)

1. Who is the gentleman over there? – Nobody. 2. What is that shape over there? – Nothing.

The one-word answers in exchanges (a) are clearly equivalent to more perspicuous expanded paraphrases of the form “There is no K in L,” where K can be replaced by an appropriate classifier or natural-kind expression, and L can be replaced by a locative word or phrase. This is the use of “nobody” or “nothing” that is closest to the negative existential quantified expressions in IRUPDOORJLFa‫ [ژ‬Fx. For suggestive convenience, in spite of some infelicity – since (a)-type sentences are tensed, whereas the existential quantifier imports no time implications – I shall call (a) the “existential” use of “nobody” and “nothing.” It is not as easy to give a single perspicuous paraphrase of the one-word answers in type (b). The respondent may intend to dispute the interrogator’s presupposition that there actually is something, let alone a gentleman or a shape, in the region referred to. In that case, the question receives in effect a type (a) answer. A distinct type (b) is defined in the case in which the respondent does not dispute the interrogator’s existential presupposition. He says, roughly: “I notice who you are referring to, but he’s nobody,” or “I notice what you’re referring to, but it’s nothing.” In that sort of case the respondent may be using the words “nobody” or “nothing” to convey the following message: (i) The subject is “of no account”: he or she or it lies outside the interlocutors’ scope of interests, in whatever way and however broadly these interests are defined by the interlocutors’ current endeavor.

Thus the snobbish respondent might say “He’s nobody” with reference to someone of inferior status; or “It’s nothing” may be the reassuring response of a doctor to a patient concerning certain physical changes in the patient’s body they have both observed. With the term “nothing,” but in reference either to a thing or to a person, the respondent can also convey this message:

38

Alexander P. D. Mourelatos

(ii) The subject is an extremely poor, worn out, or reduced specimen of the kind it prima facie appears to be.

The point is that the subject is extensively deprived of the normal characteristics of its kind, so it can “barely” or “hardly” be called, as for example in exchanges (a) and (b), a “gentleman” or a “shape.” One can more aptly speak of what the subject “is not” rather than of what he, she, or it “is.” In the limiting case, to borrow one of the definitions in Webster’s Third New International Dictionary, the subject is “something that is characterized by utter absence of determination: perfect indistinguishableness.” Let me refer to (b), in either of its two subtypes, as the “characterizing” use of “nobody” or “nothing.” What underlies type (a), the existential use, is a conception of non-being as emptiness: I search the house and find that it is vacant; or I open the box and find that it is empty. What underlies type (b), the characterizing use, is a conception of non-being, or properly not-being, as (i) lack of standing (almost in that juridical sense in which someone may have “no standing” in a particular court and case) or (ii) privation or attenuation. Just as the existential use of “nobody” or “nothing” corresponds to the existential use of “to be,” the characterizing uses correspond to copulative “to be.” The person or thing deprecated as a “nobody” or a “nothing” is certainly no fictive or non-existent entity. One might object at this point that there is no need to import a copulative construction in the analysis of the characterizing use. For may we not explain the use by leaning on the possible paraphrase, “the subject is treated as though he, she, or it did not exist”? The answer is that the paraphrase just given is really an elaboration, not an analysis, of the characterizing use. For it suppresses an important implication the characterizing use proclaims: the reason why the person or thing disparaged as “of no account” may be said not to exist is that he or she is a “no-body,” or it is “not-a-thing”: the subject is not identical with any particular person/thing we are prepared to deal with, nor classifiable under any of the kinds of persons/things that are of concern to us. Similarly, to speak of a person or thing as a “nothing,” in the sense that the subject is a very impoverished, enfeebled, or attenuated specimen of its kind admittedly could be paraphrased by “might just as well not exist.” But the paraphrase will not do as an analysis of the locution in the characterizing use. For the latter alludes to the rationale of the “just as well” version: the subject is no-thing, not-being in that it is “not-F1,” “not-F2,” “not-F3,” and so on, where the Fs stand for characteristics familiar to us, and which the subject normally ought to have. The characterizing use of mƝden and cognate negative terms, such as ouden and the personal forms PƝGHLV RXGHLV, has been studied in detail by A. C. Moorhouse. 4 Let me review some of his major points and expand with my own comments on two of his most suggestive examples. 4

Moorhouse 1965.

2. “Nothing” as “Not-Being”: Some Literary Contexts

39

In Homer the characterizing sense 5 is served by the adjective of disparagement outidanos, “a no-somebody-fellow,” a person “of no account, worthless, sorry, good for nothing.” The Homeric indefinite pronoun outis, from which outidanos is derived, has only the existential sense of “nobody,” never the characterizing sense. 6 The usual forms oudeis and PƝGHLV are also restricted to existential uses in Homer. But beginning in the Archaic period first oudeis, ouden, and then also PƝGHLVPƝden, expand their use by taking over the characterizing sense of Homer’s outidanos, now in reference both to persons and to things. A remarkable early example of the developing new use of oudeis is in the opening of Pindar’s Sixth Nemean Ode: There is one race of men, one race (genos) of gods; both have breath of life from a single mother. But a power that has sundered itself through and through holds us divided (įȚİȓȡȖİȚ į੻ ʌ઼ıĮ țİțȡȚȝȑȞĮ įȪȞĮȝȚȢ), so that the one (scil. genos, “race”) is nothing (ouden), while for the other the brazen sky is established their sure citadel forever. Yet we have some likeness (prospheromen), either in great intelligence (megan noon) or in strength (physin), to the immortals, though we know not (ouk eidotes) what the day will bring, what course after nightfall destiny has written that we must run to the end. 7

The race of men is certainly not a “nothing” in the sense of non-existence. It is a “nothing” in the sense of lacking many of the attributes the gods have, and exemplifying poorly or marginally those attributes that the two races possess in common. Pindar illustrates with reference to one such attribute: “Like the gods we have intelligence, and yet we do not know. . . .” It is relevant to note here that in another ode Pindar conveys this not-being of humanity using the very image and turn of phrase that Plato was to use in formulating a metaphysics of degrees of reality: 5

This is my terminology, not Moorhouse’s. It is fascinating to notice how deftly Homer exploits the semantic distinctness of outis and outidanos in the episode of the blinding of Polyphemus, the Cyclops, in Od. 9.364ff. The Cyclops naively believes that Odysseus’ name is Outis. So, his screams, when Odysseus is blinding him, bring no help from the other Cyclopes. For they assume he is complaining of a natural illness, hearing him, as they do, cry: “No one (outis) is killing me” (9.403-414). Homer’s finishing touch to this dramatic pun is at 460. The powerful monster, now reduced to ineffectual groping, yet still confident he will capture Odysseus and take revenge, speaks to his favorite sheep about the ills wrought on him by that “outidanos Outis.” In putting the two words together has Polyphemus now finally understood the fateful ambiguity of the sound “outis”? Or does he now understand Outis – wrongly – merely as a variant of outidanos? Or had he foolishly lulled himself from the start with the belief that Odysseus had called himself Outis (= outidanos) in abject humility and self-deprecation? Homer leaves us to wonder. In the end, Polyphemus’ favorite animal carries Odysseus safely out of the cave and beyond the Cyclopes’ reach. The crucial point, however, is clear: Odysseus not only is not, as Polyphemus’ fellow Cyclopes thought, “no one,” he also is not, as Polyphemus himself – at one time or another – thought, “a man of no account.” 7 Lattimore 1947, p. 111, with my own fuller translation of line 2. 6

40

Alexander P. D. Mourelatos

ਥʌȐȝİȡȠȚIJȓįȑIJȚȢIJȓįૃȠ੡ IJȚȢıțȚ઼Ȣ੕ȞĮȡਙȞșȡȦʌȠȢ We are things of a day. What are we? What are we not? Man is the (but the) shadow of a dream. (Pyth. viii. 95-96)

The characterizing use, amply attested in other fifth-century authors, including Aeschylus and Aristophanes, becomes especially prominent in Sophocles and Euripides. Most instructive for our purposes is a passage from Sophocles, Electra 1163-67: Dearest one, how you have destroyed (DSǀOHVDV) me! Surely (GƝWD) you have destroyed me, beloved brother! So then receive me into this your shelter (ਥȢ IJઁ ıઁȞ IJȩįİ ıIJȑȖȠȢ), me who is nothing into that which is nothing (IJ੽Ȟ ȝȘį੻Ȟ ਥȢ IJઁ ȝȘįȑȞ), that I may dwell with you below from now on.

Electra is here addressing the urn she thinks contains the ashes of her dead brother Orestes. A recent critic confesses bewilderment (I append translations of the lines he cites): Electra seems to assert in 969 that the dead are living [“you’ll win the praise of piety from your dead father below”]; then she seems to deny it in 1166 [“nothing to nothing,” quoted above] and 1170 [“dead men do not suffer pain”] . . . [t]he problem of the ontological status of ghosts in this play seems to me very obscure. 8

I believe there is no such problem; the supposed obscurity and the critic’s bewilderment arise simply from a failure to appreciate a certain natural or inevitable, but logically transparent, semantic complication, one that is inherent in discourse about the dead – in Greek or in English. The point is worth a digression, for it has implications for an understanding of the concepts of “nothing” and “non-being” that go beyond the Electra passage. What, in the event of a person’s death, makes semantically possible the locutions “he is no more,” “he does not exist,” “he is nothing (existential sense),” and their respective counterparts in Greek, is the absence of that person from the world of the living. 9 But if we should picture a world of the dead, to which persons who are no longer alive recede, and where they take permanent abode, denizens of that world can be said “not to exist” only by a sort of semantic 8 9

Sale 1973, pp. 139 and 834. There is, of course, a “vital” use or nuance of “to be,” “to exist,” and correspondingly of einai. But the fact that the paraphrases or translations “to live” or “to be alive” and “to die” or “to be dead” are available for positive and negative predications, respectively, in this type of use does not warrant saying that “to be,” “to exist” or einai in such cases simply means “to live” or “to be alive.” At the vary least, the vital use of these verbs carries greater rhetorical force, greater pathos, than is carried by corresponding uses of the predicates “lives” or “dies.” This is especially clear in literary examples that might otherwise be taken as pleonastic: ਵ ʌȠȣ ȗȫૉ IJİ țĮ੿ ਩ıIJȚȞ (Homer Od. 24.262); ȗȫȞIJȦȞ țĮ੿ ੕ȞIJȦȞ ਝșȘȞĮȓȦȞ (Demosthenes 18. 72); ੕ȜȦȜİȞ Ƞ੝įૃ ਩IJૃ ਩ıIJȚ ȉȡȠȓĮ (Euripides Troad. 1292). See Klowski 1967, p. 139; Kahn 1973, pp. 240-45, 378-79.

2. “Nothing” as “Not-Being”: Some Literary Contexts

41

legerdemain: a linguistic transposition, a metonymy. For that world is one of (existential) non-being not on that world’s own terms and from the perspective of its denizens, but on the terms of the living, and in the perspective of the living. So, the Underworld is the domain of “nothing” (existential) even though it has quite a large population, the members of which, however shadowy in their texture and however reduced they be in their powers, are involved in trials, in acts of expiation or purification, and who relish or loathe, praise or censure, the acts of their survivors in the world above. This metonymy holds only with respect to the existential use of “nonbeing” and “nothing.” In the characterizing use, there is uninterrupted continuity between the two worlds. Indeed, it is the characterizing use that permits us and even encourages us to picture the world of the dead as an extension of the world of the living. The characterizing use mediates, in effect, the metonymous and hyperbolic uses of “non-being” and “nothing.” For while there is no sense to the idea of degrees of existence 10 or degrees of nothingness (in the existential sense), there is no conceptual bar to thinking of degrees of “notbeing” (not of “non-being”) or of “nothing” in the characterizing sense. Thus it is possible for a living person to employ an effect of hyperbole by saying “I am dead already,” “I do not exist,” “I am nothing” (existential) because without hyperbole and with no recourse to metonymy he perceives himself as privatively or negatively characterized to a degree that verges on what is appropriate for someone in the realm of the dead. In a nutshell: the dead can call themselves “nothing” (existential) by metonymy; the living can call themselves “nothing” (existential) by hyperbole. But both the dead and the living can quite properly call themselves “nothing” in the characterizing sense; for in that sense it is simply a matter of more and less. So now let us go back to the Electra passage. Critics, including Moorhouse, have failed to notice an important syntactic ambiguity. The expression to mƝden may well be understood as “the nothing” or “nothingness,” as a synonym of “extinction” or as an abstract name of the Underworld pictured as the domain of non-existence. Preserving parallelism of sense, the meaning of 1165-66 would then be: “receive me, the one who (already) does not exist (WƝQ PƝGHQ) into (the realm of) non-existence (WR PƝGHQ),” a strongly emotional utterance in which the hyperbolic and metonymous uses of “nothing” are juxtaposed. But there is also another way of understanding WRPƝden, by construing the neuter article to not as introducing an abstract noun but simply as referring to the neuter noun that occurs in the preceding line, to stegos, “the shelter,” Electra’s metaphor for the funerary urn. That construction is ready to hand, as is clear from the parallelism “. . . me . . . to this shelter here; . . . the one who is nothing to this (urn) that is nothing.” Both Electra and Orestes, the latter as 10

Cf. Vlastos 1965, pp. 8-9.

42

Alexander P. D. Mourelatos

embodied by the ashes contained within the urn, as well as by the urn itself as container, are a “nothing” in the characterizing sense. The supposed nothingness of Orestes is obvious: he does not speak, he does not move, he cannot come to Electra’s help – and so on and so forth with the many “not-F”s implied in Orestes’ utterly deprived state. It is precisely in that sense that Electra, at the beginning of this scene (1129), calls Orestes – again identified with the ashes in the urn – a “nothing”: “Now I hold you in my hands as a nothing (ouden onta).” 11 In fact, Sophocles does not leave us to imagine Orestes’ “not-F”s; rather, as though to confirm the sense of “nothing” in line 1129 and prepare us for the second half of the “nothing to nothing” of line 1166, he has Electra recite a whole string of Orestes’ deprivations beginning at line 1136: Out of (ektos) your home, an exile on alien (DOOƝV) soil, you died wretchedly (NDNǀVDSǀORX), without your sister near (dicha); nor (koute) did I wash you . . . nor (oute) did I get to pick you up off the blazing pyre . . . but tended by the hands of strangers you have come back a diminutive mass (smikros onkos) in a diminutive receptacle (VPLNUǀLHQN\WHL) . . . [your] nurture unavailing (DQǀSKHOƝWRX). (1136-45)

The nothingness of the urn itself is not simply its diminutive size; the urn is hardly a “shelter” (as it is called at 1165); the oxymoron figure, “a shelter that is not a shelter,” is here very close to the surface. 12 In this same speech Electra gradually identifies her own fate with that of Orestes. At one point the rhetoric goes even beyond identification, picturing the dead Orestes not merely as the one totally deprived but as one who is actively depriving Electra “of everything,” “like a storm” (1150-51). So, we are already inclined to think of Electra as a “nothing” in the characterizing sense before we finally hear her climactic hyperbole, “I am dead” (1152, cf. 1164). Even before this scene, Sophocles has spelled out Electra’s privative nothingness: from the first scenes the play has been studded with Electra’s remarks of self-pity, in which she characterizes her predicament by using negative phrases and privative a- or apo- compounds. 13 To attempt to determine whether at 1166 it is the existential or the characterizing sense of PƝGHQ that constitutes the primary semantic layer would seem 11

The characterizing sense for ouden here is guaranteed by the immediate context: “I did not receive you back in the form in which (KǀVSHU) I sent you off” (1128); “I sent you off resplendent (lampron exepempsa)” (1130). 12 See below, at n. 17. 13 See 164-67 (the prefixes conveying negation are marked here with underlining): ਙIJİțȞȠȢ . . . ਕȞȪȝijİȣIJȠȢ . . . ਕȞȒȞȣIJȠȞ ȠੇIJȠȞ ਩ȤȠȣıĮ țĮț૵Ȟ. 186-92 ȕȓȠIJȠȢ (scil. Electra’s) ਕȞȑȜʌȚıIJȠȢ, Ƞ੝įૃ ਩IJȚ ਕȡț૵· ਚIJȚȢ ਙȞİȣ IJİțȑȦȞ . . . ਜȢ ijȓȜȠȢ Ƞ੡IJȚȢ ਕȞ੽ȡ ਫ਼ʌİȡȓıIJĮIJĮȚ . . . ਕȞĮȟȓĮ . . . ਕİȚțİ૙ ıઃȞ ıIJȠȜઽ, țİȞĮ૙Ȣ (cf. “not full”) įૃਕȝijȓıIJĮȝĮȚ IJȡĮʌȑȗĮȚȢ. 230-32 IJȐįİ Ȗ੹ȡ [Electra’s woes] ਙȜȣIJĮ țİțȜȒıİIJĮȚ· Ƞ੝įȑ ʌȠIJૃਥț țĮȝȐIJȦȞ ਕʌȠʌĮȪıȠȝĮȚ ਕȞȐȡȚșȝȠȢ ੰįİ șȡȒȞȦȞ. 813-22 ਕʌİıIJİȡȘȝȑȞȘ . . . ਙijȚȜȠȢ . . . (in Electra’s life) Ƞ੝įİ੿Ȣ ʌȩșȠȢ. Cf. Orestes’ remarks: ı૵ȝૃਕIJȓȝȦȢ țਕșȑȦȢ ਥijșĮȡȝȑȞȠȞ (1181); ijİ૨ IJોȢ ਕȞȪȝijȠȣ įȣıȝȩȡȠȣ IJİ ıોȢ IJȡȠijોȢ (1183).

2. “Nothing” as “Not-Being”: Some Literary Contexts

43

gratuitous. In Electra’s speech of desolation a line that packs together the characterizing, the hyperbolic (with respect to Electra), and the metonymous (with respect to the Underworld) use of “nothing” is just what we would expect from the pen of a tragic poet. The evidence alluded to and discussed so far is certainly adequate to establish that the characterizing use of PƝGHQRXGHQ and PƝGHLVRXGHLV is at least as viable and familiar in classical Greek as the corresponding uses of “nothing” and “nobody” are in English. But an even stronger conclusion is warranted. Doubtless, in terms of frequency, this use cannot fail but appear minor compared to the existential use. But this comparison belies the importance of the characterizing use in the development of Greek concepts. In non-philosophical sources the use is documented in contexts of major dramatic or rhetorical impact. So, its literary conspicuousness is high. Given this conspicuousness, we should hardly presume – as has generally been the case – that when “nothing” becomes thematic for a philosophical author focus will necessarily be on the existential sense. A good case can be made, I believe, for the exegetic thesis that also in Parmenides Ǻ6.2, Ǻ8.10, Ǻ9.4, the passages in which we have thematically important contrast between the eon, “what-is,” and PƝGHQ, the latter is best understood as conveying the characterizing sense, “not F, for any F one may wish to posit or mention.” 14 With reference specifically to Plato the characterizing use of PƝGHQRXGHQ must have played a significant role in the development of the doctrine of degrees of reality. The latter, as Vlastos has shown, cannot be coherently understood as a doctrine of degrees of existence, but as one that involves the characterizing F-ness and non-F-ness of sensible particulars. 15 The connection with our theme ought to be obvious: the negative pole in a doctrine of degrees of reality corresponds to the characterizing sense of “nothing,” not-F1 and notF2 and . . . not-Fn. If there are to be degrees of reality, there must also be degrees of unreality; the characterizing use both allows this and projects the relevant extreme. The connection can be rendered more faithfully and its significance better appreciated if we remind ourselves of other respects in which Plato’s doctrine draws on the resources of ordinary speech and on the literary tradition. Plato himself called our attention to one such respect. In Republic 479B the notion that there are “ambivalent” or “equivocal” (cf. epamphoterizousin) entities is introduced through humorous allusion to the folk riddle: A man who was not a man (a eunuch), saw (looked at) but did not see (recognize) a bird that was not a bird (a bat) perched on a tree that was not a 14 15

See Mourelatos 1976, which is reprinted as ch. 12 in Mourelatos 2008a. Vlastos 1965, pp. 10-11.

44

Alexander P. D. Mourelatos

tree (a rafter or beam), hit it (hit at it) and did not hit it (successfully hit it) with a stone that was not a stone (a draughtboard piece). 16

Obviously related to this type of folk riddle is one of the most striking forms of the oxymoron figure in Greek literature: “an F-person (or thing) that is not an F-person (or thing).” A good example can be drawn from that same scene of Sophocles’ Electra discussed above: PƝWƝU DPƝWǀU, “unmotherly mother,” or “mother hardly a mother” (1154). 17 As in the examples cited earlier at note 13, this type of oxymoron involves juxtaposition of a noun against a compound FRQVLVWLQJRIWKHVDPHQRXQSUHIL[HGE\Į-negative. 18 An impure variant of this GHYLFHSURGXFHVWKHVDPHHIIHFWRISDWKRVWKURXJKDFFXPXODWLRQRIĮ-negative compounds, negative predications, apo-compounds and similar expressions. We saw such a figure in the description of the supposed ashes of Orestes. 19 %RWK Į-negative alliteration and its impure variant of accumulating negatives are among the most favored devices of Greek authors, from Homer through the dramatists, the orators, and beyond. What is more, predilection for these devices reflects tendencies characteristic of the Greek language itself. For Greek not only has a richer variety of morphological variants of the negative prefix than any other Indo-European language, 20 it also is by far more productive of negative compounds than the two other ancient languages for which we have a comparable body of literature, Sanskrit and Latin. 21 The conception of a person, thing, action, or event that is characterized in negative terms through and through is no invention of philosophical ontologists; it is one of the commonplaces of Greek literature. It is thus no idle historical exercise to point out antecedents or prototypes in literature and in forms of speech for Plato’s doctrine of degrees of reality. Sources for the intermediate degree lie not only in the type of folk riddle Plato alluded to; they also lie in the oxymoron of the type PƝWƝUDPƝWǀU. Correspondingly, the Platonic PƝGDPƝL RQ, “in no way being,” or SDQWǀV PƝ RQ, “altogether not being,” that is equated to PƝGHQ, “nothing,” is prefigured and made familiar in the characterizing use of PƝGHQPƝGHLV etc., especially by the great tragedians – LQWKHILJXUHRIĮ-negative alliteration, and in the figure of accumulating negatives. All of these devices facilitate, reinforce, and promote one another. So, it is properly the whole complex that constitutes an important prephilosophical background for the Platonic doctrine of degrees of reality. Observations made in this paper can be read as providing support, in yet a dif16

See Adam 1963, 1, pp. 342-43 ad loc. I have altered slightly the usual interpretation. For a discussion of the figure, see Hamilton 1899, pp. 48-49; cf. Moorhouse 1959, pp. 66-68. 18 Other examples: Homer Il. 9.63-64 ਕijȡȒIJȦȡ ਕșȑȝȚıIJȠȢ ਕȞȑıIJȚȩȢ ਥıIJȚȞ ਥțİ૙ȞȠȢ/ ੖Ȣ ʌȠȜȑȝȠȣ ਩ȡĮIJĮȚ; Soph., Ant. 876 ਙțȜĮȣIJȠȢ, ਙijȚȜȠȢ, ਕȞȣȝȑȞĮȚȠȢ ਩ȡȤȠȝĮȚ IJ੹Ȟ ʌȪȝĮIJȠȞ ੒įȩȞ. Cf. above, n. 13, and Hamilton 1899, pp. 44-45. 19 Above, at n. 11-12. Cf. description of Electra’s predicament, n. 13. 20 Moorhouse 1959, p. 47. 21 Hamilton 1899, p. 58. 17

2. “Nothing” as “Not-Being”: Some Literary Contexts

45

ferent way, for a thesis advanced by Charles H. Kahn, 22 as well as by others. In a formulation I prefer, the thesis is that the dialectic of Being in early and also in classical Greek philosophical speculation focuses not on “What there is” but on “What it is” or “How it is”; not on existence but on physis, constitution, or form. 23

22 23

See Kahn 1976; cf. Kahn 1973, pp. 394-419. In early drafts, this study received the benefit of criticism by Friedrich Solmsen and by Michael Gagarin. It was also improved by discussions at the following venues: the 1977 meeting of the Society for Ancient Greek Philosophy in Atlanta; and in 1978, at classics seminars at The University of Auckland, and The University of Sydney.

Part II: The Discovery of Form

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

One of the chapters of W. K. C. Guthrie’s The Greek Philosophers, an excellent short book (Guthrie 1950) that has received much use in survey courses in the history of ancient philosophy, bears the title: “Matter and Form,” subtitle “Ionians and Pythagoreans.” The chapter develops the thesis that there are two distinct outlooks at the earliest stages of Presocratic philosophy, in the sixth century B.C.E. On the one side, we have philosophers who focus on the material or constitutive aspect of reality, which they variously conceptualize – first as water, or air, or fire; then, later, as a plurality of stuffs, or as a plurality of atoms. On the other side, we have Pythagoras and his followers, who focus on the formal, structural, or configurational aspect. This thesis of the two outlooks was developed more fully in the first two volumes of Guthrie’s, A History of Greek Philosophy. At one point in vol. 1 we read: Through their mathematical studies, Pythagoras and his school achieved with remarkable suddenness a rational conception of the significance of form, and were much blamed by their successors for ignoring the material side. For them reality lay in form, for others in matter. Both views have persisted . . . (Guthrie 1962-1982, I, p. 467)

There is, of course, venerable ancient precedent for articulating the history of early Greek philosophy along the lines drawn by Guthrie. Plato in the Sophist speaks of an ancient battle between the Giants and the Gods of philosophy. Members of the first group hold that “reality comprises only that which offers resistance and contact to our hand, thus identifying reality with body”; but the Gods maintain that “reality consists in certain intelligible and incorporeal Forms” (246A-B). The thinkers alluded to by Plato as “Gods” may not necessarily have been the Pythagoreans; they may have been the Eleatics, or simply the Platonists. It is, nevertheless, certain that the Platonic precedent has nurtured a tradition of interpreting Presocratic philosophy in terms of a division between friends of matter and friends of form. Guthrie’s division between Ionians and Pythagoreans represents a modern and influential version of a

50

Alexander P. D. Mourelatos

tradition which is long and which can claim venerable standing in ancient sources 1. Might we not fairly object, nonetheless, that the tradition is simplistic? Let us grant that the material or constitutive aspect of reality is inescapably obvious. Even so, are we to suppose that the complementary – yet notably more abstract – aspect of Form should have been discovered merely as a result of simple shift of the mental focus? This query is a serious challenge to the twooutlook thesis generally. But in many versions, and in particular Guthrie’s, a certain ancillary assumption has long served to provide at least the promise of an explanation. For, it was, as Guthrie puts it, “through their mathematical studies” that Pythagoras and his school made the conceptual breakthrough to the recognition of Form. Guthrie could write this with confidence in 1962 because in the preceding half century leading scholars such as John Burnet, A. E. Taylor, Sir Thomas Heath, and F. M. Cornford had endorsed and unhesitatingly promoted the view, widely held also in Hellenistic and later antiquity, that Pythagoras was the first to develop a philosophy inspired by mathematics, and indeed that Greek mathematics was founded by him and remained distinctly Pythagorean through its early stages of development. But the year in which the first volume of Guthrie’s History appeared was also the year of publication of the original German edition of Walter Burkert’s Lore and Science in Ancient Pythagoreanism, 2 a work of brilliant scholarship that radically altered our perception of the role Pythagoras and the early Pythagoreans played in the development of Greek philosophy and Greek science. Crucial for our purposes are the results of Burkert’s analysis of the ancient evidence. I present these here in my own summary: (a) Pythagoras himself was neither a mathematician nor a cosmologist; he was a shaman-like figure who preached a doctrine of immortality through

This Chapter had its origin as the Julius R. Weinberg Memorial Lecture (“The Discovery of Form in Early Greek Philosophy”) presented to a general audience at The University of Wisconsin-Madison, March 3, 1983. The main parts of it were in subsequent years presented to several other academic venues, including presentations in versions not in English as follows: in a German translation, by Christian Wildberg, in Germany, Switzerland, and Denmark; in Italian translation by Iolanda Capriglione in venues in Italy; and in Greece, in a translation by Myrto Dragona-Monachou and Christina Michalakopoulou-Veïkou. It has previously been published, in abridged form, only in its Greek version: “૽Ǿ ਕȞĮțȐȜȣȥȘ IJોȢ ȂȠȡijોȢ ıIJ੽Ȟ ʌȡȫȚȝȘ ਦȜȜȘȞȚț੽ ijȚȜȠıȠijȓĮ,” in ȆĮȜȓȝȥȘıIJȠȞ, 5 (1987), pp. 93-114. Extensively revised, in sections updated, and furnished more amply with notes and with references, it still retains the somewhat dogmatic ex cathedra style and character of its original oral presentation in Wisconsin. 1

The two-line scheme for the successions of the Greek philosophers is famously proclaimed in Diogenes Laertius I.13: “The starting points (archai) of philosophy are two: the one from Anaximander; the other from Pythagoras.” 2 Burkert 1962, English ed. 1972. All references here will be to the English-language edition.

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

51

transmigration and had founded a society of initiates in whom he inculcated a rigorously ritualistic and ascetic mode of life. (b) Even though certain fifth-century figures in the Pythagorean succession – the mathematici of our ancient sources – made contributions to the development of elementary number theory, geometry, and harmonics, the origins of Greek mathematics lie more significantly in Ionian KLVWRULƝ, in Eleatic modes of argument, and in the development of architecture than they do in Pythagoreanism. (c) Of the thinkers Aristotle calls “Pythagorean,” the first to develop a cosmology that rivals the cosmologies of Ionia is Philolaus of Croton, who was philosophically active in the second half of the fifth century, a full century after Pythagoras; and it is Philolaus who comes closest to singling out Form as a cosmological principle.

If the concern with mathematics was not distinctly Pythagorean, if, moreover, a Pythagorean philosophy of Form is a relative latecomer, Guthrie’s version of the two-outlook thesis becomes untenable. Fortunately, Burkert’s analysis made possible a completely new narrative of the development of Presocratic philosophy, and this new narrative also implies a fresh and more convincing thesis concerning the discovery of Form. 3 FOUR PHASES OF DEVELOPMENT IN THE DISCOVERY OF FORM Rather than Guthrie’s “remarkable suddenness [of] rational conception” by a nearly mythical Pythagoras – a legendary SUǀWRV KHXUHWƝV, “first discoverer,” as it were – what we have is development in four phases. In the Ionian cosmologies of the sixth and early fifth century, we have gradual and ad hoc recognition of the relevance and bearing of various aspects of Form in explanation. Then, we have a phase of conceptual crisis. It arises from Parmenides’ strong denial of the possibility not only of absolute genesis and of change generally but also of complexity, diversity, or divisibility. In most versions of the conception of Form, themes of articulation and structure (rather than of unquailfied simplicity) are inevitably implied. And, as Melissus of Samos will argue, the Parmenidean ban on sheer genesis and change must also be applied to the genesis of Form – in the sense of shape, figure, structure. Accordingly, the Parmenidean denial of change must be extended to the possibility of metaNRVPƝVLV, “change in kosmos,” i.e., “structural change” or “trans-figuration,” or “trans-formation.” Without fully coming to terms with Melissus’ critique, Anaxagoras and Empedocles, in the third phase, show stronger awareness of 3

Initial references to a particular Presocratic figure will list chapter numbers in DK, but such numbers will be omitted in subsequent references within the present Chapter where discussion continues to focus on that same figure.

52

Alexander P. D. Mourelatos

Form in their respective cosmologies, precisely after and perhaps because Form has become conceptually problematic. Finally, fourth phase, it is with Philolaus, and then with Atomism (Leucippus and Democritus of Abdera), that we have the first, coherent and fully developed Form-Matter philosophies. In the traditional historiographic contrast between philosophies of Matter and philosophies of Form, the latter term is used quite broadly. So, before I proceed to fleshing out the four phases in this narrative, it is important to make clear the sense of the term that will be thematically relevant in the present Chapter. “Form” and the Form-Matter contrast may well be understood as envisaging the categorial issue of the distinction between (i) things as concrete entities and (ii) the qualities or properties of things. This topic I have discussed in another study, not included in this volume, viz. “Heraclitus, Parmenides, and the Naïve Metaphysics of Things.” 4 A cognate metaphysical topic, is that of the distinction between intrinsic and relational properties, and this is taken up below in Chapter 7 apropos the philosophy of Democritus. Taken up in Chapter 5 is the broader topic of Presocratic stages in the recognition of entities which in later philosophy came to be known as universals. In the present Chapter, the thematic focus will be on Form primarily as “shape,” “figure,” “structure,” and (in the case of multiple entities) as “order,” “ordering,” “array,” “arrangement” or “configuration.” The correspondingly relevant Greek target terms are kosmos, taxis, eidos, idea, and PRUSKƝ; but I do not limit myself to contexts in which just these particular words are found in use. The standard Greek for “world” is, of course, kosmos. But, as is well known, kosmos does not simply and baldly mean “world”; it means “order of things,” even “a well-ordered array of things.” Now if the first philosophers of Greece were, as indeed seems to be the case, cosmologists, is it not rather pointless to speak of the discovery of Form? It would seem that the appreciation of the formal aspect of reality was part of a native Greek intuition of the world as a “world-order.” The history of usage of kosmos does not support this supposition of a native intuition. In its pre-philosophical occurrences, the term kosmos is regularly used with reference to any sort of arrangement – e.g., of herd animals, soldiers, artifacts, words, social practices. In the few early cases in which kosmos occurs in a cosmological context, the word is employed precisely for the purpose of arguing that world-processes are orderly. Thus the transition in the word’s usage from the root meaning “arrangement” to the meaning “world order” is in itself part of the history of the philosophical discovery of Form.

4

See Mourelatos 2008a, ch. 10, which had also appeared earlier, viz. Mourelatos 1973; and in Modern Greek translation, Mourelatos 1981b.

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

53

THE APPEAL TO FORM IN THE “MATERIAL-MONISTS” OF EARLIEST GREEK PHILOSOPHY The term “material monists” has gained some currency for referring to the three philosophers of Miletus in Asia Minor (the “Milesians”), and – perhaps with some strain – also, as fourth, to Heraclitus of Ephesus, whose cosmic Fire is rather more than a material principle. I borrow the term only for its convenience in allowing referring to the four together. About the first of the three, Thales, the sixth-century B.C.E. traditional “father of philosophy,” little is reported in the ancient sources concerning his natural-philosophical or cosmological doctrines; and that little is viewed skeptically by modern scholars. It is perhaps safe to assume that the “first philosopher” ought to count as the first philosopher of Matter, not of Form. For it appears he held that water was, in some sense, the source or principle of all things. For another side, however, of Thales’ intellectual pursuits, we do have fairly reliable reports; and the significance of this fact has not been sufficiently appreciated. Thales does seem to have been preoccupied in a major way with geometry and with computational astronomy (DK11A1). 5 With him, more so than with Pythagoras, we have the possibility of a philosophy growing out of reflection on geometric structure and on number. And yet there is no suggestion in our sources that Thales’ mathematical pursuits bore in some way on his natural philosophy. Ostensibly one can have more than casual brushes with the contexts of structure and of quantity in mathematics without becoming a philosopher of Form. In the case of the next philosopher in the Milesian succession, Anaximander, the situation is exactly the reverse. Our reports do not at all suggest that he was interested in mathematics as such. Nevertheless, themes of geometric structure, symmetry, and proportion are unmistakably implied in his cosmology. He argued, we are told, that the earth requires no support and is stable at the center of the cosmos precisely because it is equidistant from all points of each of the surrounding rings that bear the planets and fixed stars (DK12A26). He gave the shape of the earth as that of a flat cylinder or column drum, and postulated a precise ratio of the depth or height of the cylinder vis-à-vis the diameter of the cylinder’s base: one to three (A10). He drew a map of the earth on which the major land masses were symmetrically arrayed (A1 and A6). And he gave proportions, perhaps in accordance with the arithmetic series 9-18-27, for the sizes of the rings of the stars, the moon, and the sun (A11, A21). The cosmographic theme of symmetry is reinforced by Anaximander’s doctrine of balance and regular alternation between certain opposed stuffs and powers – e.g. the Hot and the Cold – that pervade and constitute the cosmos. In 5

See Burkert 1962/1972, pp. 413-17. Cf. White 2008; Rossetti 2015, pp. 173-207.

54

Alexander P. D. Mourelatos

the sole remaining fragment from Anaximander’s book, this alternation between cosmic opposites is described in quasi-mythical terms: “they make amends and give reparation to one another for trespass according to the taxis of time” (DK12B1). What does taxis mean here? It could well be the “order” of time, i.e., its articulation in terms of measurable recurring stages and periods. If so, Anaximander’s fragment contains the first attested occurrence of a word for “order” in a cosmological context. Another possibility is that taxis here has the juridical sense of “assessment.” Time, accordingly, will have been metaphorically pictured as a judge who settles the claims of injury of the contending cosmic opposites. (The belief is still widespread today that an unusually hot summer will be followed by an unusually cold winter.) The theme of balance and regularity is certainly not lost if we adopt the latter, more picturesque or more concrete, interpretation. Anaximander may not have framed yet the higher-order abstraction of “the order of time,” but his image of the court of personified Time is, as Charles H. Kahn has pointed out, the remote archetype of the Western idea of “the universe governed by law.” 6 In spite of the prominence of the symmetry-regularity theme, the ruling principle in Anaximander’s cosmology is the apeiron, which is both the “unlimited” or “un-traversable” but also the “indeterminate” or “unstructured.” The apeiron is unmistakably a material, rather than formal, principle: an immense reservoir of cosmic stuff and power. Some modern scholars have found here a significant contrast: between an outer or encompassing expanse of indeterminacy, and an encompassed world of heaven and earth that is articulated in spatial proportions and balanced temporal terms. This seems gratuitous. None of our ancient sources hint of a connection between the symmetry-regularity theme in Anaximander and the choice of apeiron as principle. Anaximander was probably the first philosopher to have had some intuitive grasp of the importance of Form in his explanation of such facts as the stability of the earth and meteorological or biological cycles. But his archaic imagination – captivated by what I have elsewhere called the “naive metaphysics of things” 7 – prevented him from elevating structure to the status of coordinate principle. In the case of the third of the Milesian pioneers, Anaximenes, we have no direct attestation either of interest in mathematics or of a doctrine of cosmic symmetry. Nevertheless, the contribution he makes to the development of the theme of Form is by no means negligible. At first blush his philosophy appears as a simple variant of the doctrine usually attributed to Thales: we are told Anaximenes held that the principle of things is air (DK13A7, A5). The crucial difference is that Anaximenes elaborates this global thesis into a theory. Air may become rarefied or again condensed, or – better said, since Anaximenes’ 6 7

Kahn 1960, ch, 3, esp. pp. 183-93. See above, note 4.

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

55

air is living stuff 8 - air dilates and contracts: stones, and ice, and liquids are air in various states of contraction; heat, light, and fire are air in various states of dilation. We do not know what word he might have used – it could have been one of the three synonyms, eidos, or idea, or PRUSKƝ, or some equivalent expression – but it is inevitable that at some point he would have explained that things other than air are variant forms of air. Without explicitly recognizing Form as a coordinate principle, the theory nevertheless suggests that one must do more than posit the stuff of things; one must distinguish and systematically relate the different states of aggregation of that stuff. Historians of ideas have often seen in Anaximenes’ theory an anticipation of the program for the reduction of quality to quantity that is characteristic of modern science. Now given the total absence of the mathematical theme in Anaximenes, this particular connection of Anaximenes to later developments is too loose to be significant. On the other hand, we can say without indulging in overstatement, that Anaximenes does envisage a reduction of the grand phantasmagoria of qualities and powers that constitute the world to the specific set of qualities and powers that are perceived or noted in air. And since the mechanism of that reduction is one of spatial aggregation, Anaximenes’ position is recognizably an intermediate stage to the reduction of all quality to structure and quantity that is later attempted by the Greek Atomists. In speaking of how the dome of stars turns around the earth, Anaximenes uses the analogy of a pilion, a “felt hat,” that can be turned around the head of the wearer (felt being the proper material for a stiff hat); and the same testimonium (A7) gives SLOƝVLV, “felting,” as Anaximenes’ analogue of the process that transforms air first into clouds and then ultimately into stones. The “felting” analogy suggests that Anaximenes was concerned to mitigate some of the conceptual incongruities involved in his transformational physics. How can the light, transparent, and barely palpable stuff that is air be transformed into the heavy, opaque, and impenetrable stuff that is a rock or a stone? The process is made intelligible by the craft of felting, whereby the soft, loose, and fluffy stuff that is wool is transformed into felt: the coarse, densely packed, and stiff material that is suitable for hat-making. Anaximenes’ explanation of the stability of the earth lacks the rationalist vision of Anaximander’s strikingly structural account. Anaximenes’ account is mechanical: not only the earth but the sun, moon, and stars as well “ride” (epocheisthai) on air. But it has often been overlooked by modern scholars that the causal nub of this doctrine is crucially formal: dia platos, “because of flatness” (A20). What makes this riding possible is the flat, leaf-like shape of all these bodies. This is probably just one of many types of phenomena for which

8

Cf. B2 “As our own soul, being air, holds us together (or ‘dominates us,’ synkratei), so too breath (pneuma) and air engirdle (periechei) the entire world.”

56

Alexander P. D. Mourelatos

Anaximenes had given an explanation that tends to point toward what Aristotle later called “the formal cause.” The style of the fourth (and “virtual”) of the so-called Material Monists, Heraclitus of Ephesos, is apothegmatic, and his philosophical message is often elusive or indeed cryptic. But it is Heraclitus who prominently put forward the claim that the world is a kosmos, i.e., a “cohesive order of things” (DK22B30). Moreover, several insights of his are framed in the language of harmonia, “structure” (B8, B51, B54) or of syllapsies, “connections” (B10), or of metron, “measure” (B31) and even of the geometric proportion (A is to B as B is to C; e.g. at B79). 9 In what was probably the opening of his treatise (B1), he arrogates to himself the project of “dividing things (GLDLUHǀQ) in accordance with nature (kata physin) and [then] showing how they hold together” (NDLSKUD]ǀQ KRNǀVHFKHL). In many respects, his contributions reinforce and continue themes introduced by the two Milesians. In particular, the theme of universal law, which was introduced in Anaximander’s philosophy, is richly enhanced and developed in Heraclitus’ doctrine of the logos: a cosmic principle that is allpervading, is sovereign in its sway (B1), and yet is also accessible and manifest to all (xynos, “common,” B2). Human laws are “nurtured (or ‘under the tutelage,’ trephontai) of the one law, the divine” (B114). Most importantly, there is a distinctive and philosophically fecund theme in his choosing the term logos for the purpose of capturing his fundamental principle, viz., that that there is some profound affinity between the nature and structure of reality and the structures and forms that pervade and characterize language. As I have argued elsewhere, it would not be unhistorical to attribute to him the discovery that, apart from taking into account the spatial and temporal relations and structures that are obvious and paramount within a “naïve metaphysics of things,” it is the philosopher’s task and office to take into account logical and conceptual structures. 10 THE SPECIAL CONTRIBUTIONS OF XENOPHANES OF COLOPHON Often skipped in accounts of the early development of natural philosophy is Xenophanes of Colophon. If taken up at all, he is cited primarily for his questioning of traditional beliefs about the gods and for his critique of the Greek cult of the athlete. Much of this neglect has resulted from the tendency, already well-established in antiquity, of assimilating his views to those of Parmenides. But the publication of Lesher 1992, an account that is more ample in scope, more fair to Xenophanes, and more perceptive of the latter’s genius, has inaugurated a period of more sympathetic reading of the evidence for this philosophical pioneer. I have myself sought, over the last thirty years, to contribute 9 10

More examples of the latter in Fränkel 1938. See above, note 4.

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

57

to undoing the “Eleatization” of Xenophanes by putting together a series of studies devoted to Xenophanes’ natural philosophy. 11 Perhaps there may be some excess and much redundncy in the amassing of 348 ancient and medieval texts that bear on Xenophanes, as offered recently in Strobel & Wöhrle 2018. Nonetheless, it is a fact that for Xenophanes we have enough evidence in testimonia – let alone the evidence provided by the fragments from his poetry – to suggest that Xenophanes had worked out a scheme of natural philosophy very much in the mold of his Ionian predecessors, especially of Anaximenes. Relevant to our theme of the discovery of Form are three aspects. First of all, in order for Xenophanes to put forward, and to provide support for, his grand generalization that all PHWHǀUD, all things “suspended above,” are variously different types of coalescence of cloud stuff, he must have looked at the skies with something of a star-gazer’s eyes for taxonomy. The stupendous variety evident in types of ordinary clouds must have provided him with clues for finding affinities with non-ordinary or unusual PHWHǀUD: not only rainbows, parhelia, and lightning, but also shooting-stars, comets, and St. Elmo’s fire. And by further exploiting affinities and transitional types, he felt justified in extending his generalization so as to include stars, constellations, planets, the moon, and the sun. 12 In the Section (earlier in this Chapter) on the “Materialist Monists,” I was hesitant to attribute to Anaximander some philosophical awareness of the conceptually important contrast between things infinite and things finite. But that contrast does come into play in Xenophanes, and this is the second aspect of his doctrine that is worth dwelling on in the story of a four-phase development. It is a fundamental tenet in Xenophanes’ flat-earth cosmology that there are two world masses: earth below; and water above (the latter being perceived also as vapor, and thus as air). Earth extends below “without a limit” (and by implication also laterally without a limit); and air/vapor extends east-west, north-south “without a limit” (and by implication also upwards, in height, without a limit). The single and infinitely extended planar “limit” or “boundary” between the two cosmic masses is the one that is inescapably evident to all of us, the one “by our feet” (DK21B28). 13 Moreover, this doctrine of the “limitlessness” of the two cosmic masses is intimately connected with his doctrine that a qualitative characteristic generally can only mark a position or grade within a limitless continuum. There is no absolutely “sweet” thing, no absolutely “best” thing, and no absolutely true in11

See Mourelatos 2002a and 2002b; also 2008b and 2013. Most of these, both ordinary and non-ordinary types of celestial luminescence, are cited in fragments or testimonia concerning his doctrine. For details, see Mourelatos 2008b. 13 For the evidence concerning the infinity of the two world masses, see now Strobel–Wöhrle 2018. From their long list of Similien on the theme of Unendlichkeit der Erde in their Similienapparat (p. 451), at least a dozen (in accordance with the Strobel–Wöhrle numbering of Xenophanes testimonia) are relevant: Xen 4 and (hereafter omitting the “Xen”), 14, 51, 94, 132, 162, 186a, 234, 235, 256, 317, 330, 331b. – See also. Mourelatos 2020, 2016, and 2008b. 12

58

Alexander P. D. Mourelatos

sight or belief; we only make comparisons, noting that one flavor is “sweeter” than another (B38), one discovery or practice “better” than another (B18), one belief “more plausible” or “more appropriate” than another or others (B34). Even his one God is not described as “great” without qualification but rather “greatest” (megistos) in comparison with human beings and in comparison with the gods worshipped by Greeks or by Barbarians (en te theoisin kai DQWKUǀSRLVLQ, B23). 14 Relevant and helpful in connection with the third aspect of Xenophanes’ contribution to the story of Form is a certain ages-old commonplace about clouds, which is very aptly noted in a famous passage in Shakespeare’s Antony and Cleopatra, Act IV, Scene 14: Sometime we see a cloud that’s dragonish, A vapour sometime, like a bear, or lion, A tower’d citadel, a pendent rock, A forked mountain, or blue promontory With trees upon ‘t, that nod unto the world, And mock our eyes with air. ( . . . ) That which is now a horse, even with a thought The rack dislimns, and makes it indistinct As water is in water. 15

The commonplace is, of course, that clouds are supereminently versatile in their capacity to assume almost any shape, and also to change that shape from moment to moment into any other. Inevitably that commonplace must have been associated with Xenophanes’ cloud-astrophysics. Indeed, we find it already alluded to in Aristophanes’ The Clouds, in what quite probably is a satiric reference to Xenophanes: “. . . centaur . . . leopard . . . wolf . . . bull. Clouds can become whichever of the many things they please to be” (Nub. 346-480). This striking versatility of clouds may well have been seen by Xenophanes as an asset for his theory. But the theme of cloud-changes (otherwise also called “sea changes”) will have brought to the fore and made more mordant conceptually the mystery and non-intelligibility of PHWDNRVPƝVLV, “transformation, transfiguration.” In the third phase in my story of development, in the philosophy of Melissus, not just cloud changes (or “sea changes”) but any sort of PHWDNRVPƝVLV, “alteration of shape,” is conceptually rejected as unintelligible.

14 15

For Xenophanes’ doctrine of epistemological “comparativism,” see especially Mourelatos 2016. Also in Shakespeare we have the exchange between Hamlet and Pollonius at the end of scene 2 in Hamlet act III : “Do you see yonder cloud that’s almost in shape of a camel?”; then in the next five lines, “like a weasel . . . like a whale . . . .”

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

59

After this survey of early Ionian first steps toward the emergence of a philosophy of Form, we can now turn to the second phase, in which Parmenides instigates what has often been referred to as “the Eleatic Crisis.” THE ELEATIC CRISIS AND PARMENIDES’ MULTIVALENT ROLE IN THE RECOGNITION OF FORM The contribution Parmenides made to the discovery of Form is complex: in content it is both positive and negative; in intent it is challenging, and in outcome paradoxical. I shall distinguish four aspects. (A) Parmenides calls attention to Form by, in effect, arguing for its unreality. (B) He further develops – albeit with some expressions of reserve – the Ionian theme of cosmic symmetry and of the relevance of Form in explanation. (C) He deploys a certain dialectical contrast that will come to serve as the prototype for the Form-Matter dualism. (D) His own critique of coming-to-be, perishing, and change challenges the imagination of his contemporaries and successors in ways that ultimately lead to the development of philosophies of Form.

In the fourth aspect, (D), the reader will have recognized reference to the third phase (scil., Empedocles, Anaxagoras) and to the fourth (Philolaus and the Atomists) of my narrative of development. Let me for the moment concentrate on aspects (A) to (C), since these can be discussed independently of reactions to, and the aftermath of, the Eleatic Crisis. The central part of Parmenides’ didactic poem contains a long deduction (DK28B8.5-49) of the conceptual requirements that specify “what-is” (or “what exists” or “the real”). These requirements are: (i) ungenerable and imperishable (DJHQƝWRQ, DQǀOHWKURQ); (ii) indivisible, simple, and coherent (homou pan hen, syneches, adiaireton); (iii) immobile (DNLQƝWRQ); (iv) fully actualized, perfect, and thus changeless (ateleston, RXNDWHOHXWƝWon, tetelesmenon). 16 I speak of a “deduction” in order to suggest that Parmenides does not merely posit these requirements; he argues for them at length, partly on conceptual considerations that seem intuitive, and partly on the basis of epistemological and semantic assumptions that are controversial. 17 16

The argument for (i) is at DK28B8.5-21; for (ii) at B8.22-25; for (iii) at B8.31; for (iv) at B8.32-33 (initially). The argument for (iv) is then amplified and reinforced at B8.42-49, in which the “fully actualized” or “perfect” (tetelesmenon) character of “what-is” is compared in a simile to the perfection of a well-made spherical object. Moreover, the requirements (i) to (iv) are announced programmatically at the start of the deduction, at B8.1-4. 17 See below, Section on “Two Strands in Parmenides’ Argument.”

60

Alexander P. D. Mourelatos

The four requirements deduced may, of course, be taken as specifying either a unique entity or a plurality of entities. Both readings of Parmenides are represented in modern scholarship, and both have the consequence that Form is unreal. Traditional and more familiar is the reading that has Parmenides hold that there exists only a single, both absolutely simple and all-encompassing entity – what Melissus called “the One.” Since Form implies articulation, hence differentiation, hence division, the One cannot possibly have Form. But, as an increeasing number of scholars have urged in recent decades, 18 it is also possible to read Parmenides as arguing for a pluralistic universe of monads. Form fares no better in this monadistic interpretation. Form would have to be either an internal feature of each of these monads or an external feature that organizes sets of monads. If internal, the argument applied to the One applies to each monad with equal force: indivisibility has been compromised. If external, this alternative contradicts Parmenides’ express statement that “there neither is nor will there be anything outside what-is” (B8.36-37), which in the monadist interpretation means “. . . outside each and every monad.” This fundamental uncongeniality of Parmenides’ philosophy to the conception of Form is borne out in Parmenides’ ironic elaboration of Doxa, the doctrine of mortals with which his own doctrine of “Truth” is in significant contrast. Doxa is replete with structural features. The two constituent entities are in themselves powers (dynameis) and bodies (demas); but they are articulated in a dualistic scheme, referred to as a diakosmos, along lines of polarity and affinity (B8.33-61). The universe of Doxa is spherical in shape, and it has outermost limits (B11); there is a system of braided celestial rings that hold the stars (B12, cf. A37); there is a measured alternation of light and night in these rings (B12, note aisa); there is overall radial symmetry, with a cosmic goddess situated at the center and “piloting” the “mixing” of cosmic constituents (B12.3-6). Remarkably, the Doxa does not merely embellish views of the early Ionians. Within its brackets of philosophical irony are propounded Parmenides’ own geographical and astronomical discoveries. For it was Parmenides himself who first argued that the earth was a sphere (A44), who distinguished the five zones marked by the two tropics and the two arctic circles (A44a), explained the phases of the moon on the hypothesis that the moon derives its light from the sun (B14 and B15), and posited the identity of Morning Star and Evening Star (A1 § 23). In all these discoveries it is Form that provides the principle of explanation. This veritable celebration of Form is, nevertheless, embedded in a scheme Parmenides approaches with express diffidence, declaring at the start that it offers “no assurance of truth” (B1.30), and carefully segregating its exposition. Since his official doctrine in the “Truth” part of his poem rules the concept of 18

For example, Barnes 1979, Curd 1998.

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

61

Form out, it is natural that words directly expressive of the concept appear in poignantly critical contexts. The word kosmos, “arrangement,” occurs twice, first in an isolated fragment, which rather mysteriously argues as follows: The what-is (to eon) will not sever (DSRWPƝ[HL) its connection with what-is, either by virtue of scattering itself (skidnamenon – alternate translations: “spreading,” “expanding”; or “being scattered”) in every direction (SDQWƝL), every which way (SDQWǀV), kata kosmon (possible translations: “with respect to shape”[?], “across the world’ [?]), or by virtue of its being gathered together (synistamenon; which may be alternatively translated “gathering itself together,” “contracting”[?], “collapsing”[?]). (B4.2-3)

The word kosmos occurs again at B8.52, with reference to “the deceptive arrangement of words” that expound the Doxa, the cosmological and naturalphilosophical second part of his poem. Indeed, the doctrine of the Doxa is compendiously referred to in the phrase diakosmon eoikota panta, “a scheme that is “in its entirety plausible” (or “appropriate,” or “seeming,” or “apparent”). Here the prefix dia- not only conveys the sense of structural elaboration but also hints at the violation of the requirement of indivisibility (B8.60). Even one of the three Greek words we regularly translate with “form” appears in a compromising context: it is conspicuously the two constituent entities of the Doxa that are spoken of as morphai, “forms” (B8.53); and in the line in which morphai occurs, Parmenides speaks of the mortals’ willful posit of these “forms”; then in the following line he explains the mortals’ fateful act of “going astray” (B8.54 SHSODQƝPHQRLHLVLQ). I have now presented the first two of the four aspects of Parmenides contribution to the discovery of Form that I distinguished earlier. I turn now to aspect (C), and begin by providing some relevant background. Readers of Parmenides’ poem are often struck by the prominence of the imagery of “bounds” (B8.42 peiras, B8.49 peirasi) and “binding” (B8.14 SHGƝLVLQ, B8.26 GHVPǀQ, B8.31 desmoisin, B8.37 HSHGƝVHQ) in the poem’s main argument. The subject, the what-is, is pictured as totally confined and held immobile in shackles or straps that “constrict it all around” (B8.31). A trio of personifications, Justice, Constraint, and Fate, are presented as responsible for this metaphysical incarceration and as keeping vigil to ensure that the bonds remain tight. This image of centripetal confinement and stricture is then poetically modulated into Parmenides’ famous simile: the what-is in its perfection may be compared (enalinkion, B8.43) to a ball of perfect curvature (B8.42-49). Some readers, already in antiquity, mistook the simile as a literal deduction of physical shape for what-is. But if the what-is actually and literally had spherecal shape, the requirement of indivisibility would be all too obviously violated. Surely the sphericity of what-is is as much a part of the metaphor as is the trio of divine wardens with their shackles.

62

Alexander P. D. Mourelatos

But the more abstract language of the peirata, the “bounds” or “limits” of what-is, conveys a property that does belong to the envisaged subject quite properly and literally: what-is is “delimited,” “definite,” or “determinate.” In our text the two allusions to peirata are glossed immediately by Parmenides’ alternate expressions of ontological determinacy: ateleston, “not (something expected) to be realized,” RXNDWHOHXWƝWRQ, “not unfulfilled” or “not unrealized” (B8.32); and (in pointedly positive terms) tetelesmenon, “accomplished,” or “consummated” (B8.42). Against this background the third aspect, the one marked earlier as (C), now comes into view. The what-is does not and cannot have Form. If, nevertheless, its nature is that of determinacy, then what-is-not must correspondingly be the “indeterminate.” Indeed, ouk anyston, “not realizable,” is part of Parmenides’ official indictment of what-is-not (B2.7). In this connection, another – the most important – of Parmenides’ metaphors deserves attention. He speaks of a “quest” and a “journey” to what-is, thus implicitly representing the latter as our fixed destination, our telos. He also distinguishes two “routes of inquiry”: one that provides answers in terms of “is”; and another, in terms of “is not.” Those who travel the route of “is not” wander helplessly on a trail of “no tidings” (B2.6 panapeuthea atarpon). Semantically and epistemologically, the route of “is” has a determinate goal and “accomplishes” its journey; the route of “is not” is one of semantic vagueness and cognitive vagrancy. The ontological corollary is that what-is-not is something utterly unsettled, loose, unfocused, and disunified. By contrast, the what-is is something fixed, focused and self-contained so tightly as to be absolutely simple. Earlier, in connection with the first phase of my narrative, I called attention to the contrast Xenophanes drew between the single cosmic “limit” (the earth’s surface) and the two “limitless” cosmic stuffs. Here then, drawn large in the images that shape Parmenides’ thought, is the contrast between determinacy and indeterminacy. Unlike Xenophanes, Parmenides is no dualist; his ontology countenances only what-is; the opposition between what-is and what-is-not is purely dialectical. But, as we shall shortly see, the first Form-Matter ontology, the philosophy of Philolaus, is couched in the terms of a dualism of “limiting things” and “unlimited things.” Since Philolaus’ philosophy is one of the fifthcentury answers to Elea, it is fascinating to consider that this ontological dualism is a transposition of the dialectical opposition of determinacy vs. indeterminacy in Parmenides, and even before him, of the opposition between “limits” and “unlimited” or infinite entities in Xenophanes. The fourth and most important aspect of Parmenides’ contribution to the discovery of Form cannot be presented directly. It will emerge gradually as I discuss the aftermath of the Eleatic Crisis, apropos Empedocles, Anaxagoras, and Melissus. But it is now relevant to examine more closely the rationale of Parmenides’ deduction of the fundamental properties of “what-is.”

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

63

TWO STRANDS IN PARMENIDES’ ARGUMENT Discernible in Parmenides’ deduction are two argument strands: the first one leads out from the intuition that there is no coming-to-be from nothing and no perishing into nothing. This is the strand that is retraced and reformulated by Melissus, who expressly bases his own deduction of the properties of “what-is” on the principle, “in no way could anything come to be out of nothing” (DK30B1). The intuition Parmenides and Melissus draw on may have already played a role in the development of pre-Parmenidean philosophy, and the force of the Eleatic argument in this respect lies in articulating that intuition and exhibiting its implications. 19 The second strand in Parmenides’ argument ties on to certain epistemological-semantic considerations concerning the admissibility of statements of the “is not” locutionary scheme, i.e., either negative existence statements, or negative descriptions, or denials of the actuality of certain states-of-affairs. This second strand was not picked up by Melissus, and the radical argument it guides in Parmenides was not in any way answered by the later Presocratics. It was, notoriously, exploited by the Sophists, who extended the strand toward such paradoxical conclusions as the impossibility of false discourse, the impossibility of disagreement, the impossibility of discursive communication, 20 even the non-existence of everything. Then, in the fourth century, the Parmenidean second strand led Plato to some of his most profound reflections on the relation between language and reality, in the dialogue he named The Sophist. When I speak here of answers to Elea by the immediate philosophical successors of Parmenides, I refer not to this second strand, the one that leads to the Sophists and to The Sophist, but to the first, the one that has a reprise in Melissus. FORMS AND FORMATION IN THE NATURAL PHILOSOPHY OF EMPEDOCLES Empedocles, too, cast his philosophy in the medium of didactic poetry. 21 His answer to Parmenides can be summarized simply enough, drawing mainly on DK31B6-B9 and B17: There are four elements – fire, air, water, and earth – and two cosmic forces, Love and Strife. Each of the elements and each of the two forces is unborn and imperishable, and each is in itself changeless. Plants 19

See Chapter 1 above, on “Presocratic Origins of the Principle that There Are No Origins from Nothing.” 20 Otherwise put, the impossibility of language that operates under the constraints of truth-telling. This Sophistic argument, by Gorgias of Leontini, is discussed below in Chapter 8. 21 By approximate and likely dates of life, Anaxagoras (500-428) is, to be sure, older than Empedocles (484-423). But the two are generationally close enough to count as contemporaries. Modern accounts often find it thematically more helpful to take up Empedocles first. In Laks–Most (2016), Empedocles is in vol. 5 and Anaxagoras in vol. 6.

64

Alexander P. D. Mourelatos

and animals are compounds of these elements produced by the creative action of Love. The coming-to-be of such compounds is the coming-together of the constituent elements; the perishing of compounds is the falling apart of the constituents under the influence of Strife. Intermediate processes of qualitative and quantitative change involve a balance between Love’s work of bringing the elements together and Strife’s work of pulling them apart. What is true of individual compounds is also true of the world as a whole, which undergoes a cycle of integration and disruption. Periodically Love succeeds in bringing the elements together into the shape of a great cosmic Sphere. Without actually ceasing to exist, or changing their identity, the four so totally interpenetrate one another in the Sphere that they become, for that stage of the cycle, indistinguishable. In the other stage or stages, 22 Strife keeps things in more or less the arrangement familiar to us: most of the earth in the middle; most of the water on the earth’s surface; most of the air above the water; and most of the fire in the heavens. Two of the requirements for what-is posited by Parmenides seem (at least prima facie) to have been met: no generation or perishing; no intrinsic change. Empedocles presumably intends to meet the two other requirements, but it is obvious that he does so only equivocally or partially. It is significant that he has made the two moving causes external to the four elements. For he can thus plead that, at least with reference to the four elements, the requirement of immobility has been met as well: they do not move intrinsically; they are merely shuffled around by the two forces. As for the requirement of oneness or simplicity, he can plead that each of the elements and each of the forces is uniform and homogeneous, though he quite obviously must concede their divisibility in space. And, as though to further assuage his Eleatic conscience, he allows for the regular recurrence of that entity he pointedly calls “The One” and Sphairos, the immense spherical to pan, “the whole,” within which the maximum of possible integration of the four elements is achieved. Within this scheme of eternally stable elements, eternally active motive forces, and ephemeral compounds, Form is naturally given full play. The word eidos occurs eight times in the fragments; the cognate term idea also occurs, once; and there are two occurrences of the third of the Greek synonyms for “form,” PRUSKƝ. More significantly, we find in Empedocles’ fragments the first attested instances of explicit opposition of structure to quality. Employing the paired conjunctions te-te, Empedocles notes at B71 that mortal compounds have both an eidos, “shape, figure” and a chroia, “quality” or “texture.” Then 22

Some interpreters of the fragments reconstruct the cyclical cosmogony/cosmology of Empedocles as involving two stages: dominance of Love; followed by an era of disruption (whether maximal, moderate, or minimal) by Hate. Other interpreters see evidence of additional stages: a stage of total separation of the four elements into zones or regions, allowing for no possibility of compounds in that stage (thus as balancing counterpart to the era of total dominance by Love); and two intermediate stages between the extremes of total fusion and total separation, living things and other compounds emerging only in these two intermediate stages.

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

65

again, at B35 he uses cheito, “were poured,” to express the qualitative blending that takes place as the elements mix; whereas in the second half of the sentence he uses the phrase SDQWRLDLV LGHƝLVLQ DUƝURWD, “fitted together in all sorts of shapes,” to express Love’s work of structuring (B35, lines 15 and 16). In addition to conferring shape, Love’s act of structuring also selects the proportions of the constituent elements. For it is also Empedocles’ doctrine that qualitative effects of blending depend on certain ratios: e.g., one part earth, one part water, two parts fire for bone (B96); nearly equal parts of all the elements for blood (B98). As it has often been pointed out, these Empedoclean ratios are primitive versions of the formulae of modern quantitative chemistry. Love’s activity is conveyed through such expressions as harmozein or synarmozein, “to fit together” (B107.l, B71.4) and SODGƝ, “molding” (B75.2). She is herself referred to as harmonia, or perhaps in personification Harmonia (B27.2, B96.4, B122.2), and she is unmistakably pictured as a craftsman – the first cosmic craftsman in Greek philosophy. There is also a telling contract between the state of maximum organization represented by the Sphairos and the state of chaotic disintegration at Strife’s moment of triumph. Aristotle was uncharacteristically charitable to Empedocles with respect to the latter’s concern for Form. In Parts of animals he takes note – somewhat condescendingly, to be sure – of Empedocles’ introduction of the concept of logos for compounds, their numerical formula (I.1.642a17 ff.). In the Physics he recognizes – again grudgingly – that Empedocles, along with Democritus, “touched rather slightly” on Aristotle’s own conception of Form as WR WL ƝQ einai, “essence” (II.2.194a18-21). But we must not be misled by these comments, even as they come from a philosopher who is often a sharp detractor of Empedocles. The truth is that Form has not assumed the status of a principle in Empedocles’ philosophy. To develop the point just made I must draw a distinction between Form as pattern-of-relations and Form as resulting-structure (and it matters not at all that Empedocles may not himself have drawn that distinction): the elements must assume certain relations vis-à-vis one another in order for a particular structure, say, that of a plant, to arise. In regard to Form as pattern of relations, Empedocles seems utterly unconcerned or even dismissive. It is undoubtedly the thrust of his metaphysics that the coming to be or perishing of compounds is nothing but the rearrangement of the elements vis-à-vis one another. So, whereas he finds something fundamentally wrong in the notion of the coming-to-be or the perishing of a plant, he finds the underlying process of relational change unproblematic and not calling for philosophical comment. In regard to Form as resulting structure, Empedocles’ attitude is not dismissive but decidedly hostile. He does not, as we saw, seek to enforce the Parmenidean requirement of indivisibility; it is not on that basis that he must rule Form out of his ontology. Rather he must do so because it violates the two Parmenidean requirements that he accepts without demur: no generation or

66

Alexander P. D. Mourelatos

perishing; and no inherent change. In Empedocles’ exposition, the term eidos, “form,” is often a way of referring to a compound. Indeed, I see no way he could distinguish between “form” and “compound.” And compounds are for Empedocles “mortal” entities. He openly acknowledges that the only entities that count as “what-is” are the four elements and the two forces. In a sentence that recurs formulaically through the poem (B17.34, B21.13, B26.3), he emphatically states, “these things alone [the four and the two] exist [or ‘are real’].” Interestingly, Empedocles must have sensed the paradox of his introducing entities that are officially outside his ontology: viz., compounds generally; and biological forms in particular. For he speaks of the emergence of compounds in a quasi-religious idiom appropriate for describing miracles and effects of epiphany: “fitted together numinously (thespesiethen, B96.4); “a marvel to behold” (thauma idesthai, B35.17); “unspeakably wondrous in their source” (B23.10). This lingering ontological mystery of the emergence of Form – and, of course, the corresponding mystery of Form’s dissipation – is something that can only be devoutly acquiesced in. The philosophically crucial point is that the emergence or dissipation of Form as structure should be viewed as the byproduct or after-effect of a (supposedly) unexceptionable underlying process of “merely” relational transposition. 23 ANAXAGORAS: HOMOEOMEROUS VS. NON-HOMOEOMEROUS ENTITIES, AND COSMIC ORDERING Another early answer to Elea came from Anaxagoras of Clazomenae. His philosophy is subtle; but it is inadequately represented in our sources, with the result that major problems of exposition obtrude. Fortunately, we can, without seriously begging contested issues of interpretation, abstract just the points relevant for our theme; and these also happen to be points on which Anaxagoras is in fundamental agreement with Empedocles. Anaxagoras studiously avoids postulating a definite set of elements. For him, every stuff, every power, and every feature that we ever encounter has always existed and will always exist. Everything is always there, but we notice just the things that have become temporarily conspicuous as a result of a process of decomposition and recomposition. Motive force for this process is a separate principle he calls nous, “mind” or “intelligence.” Ostensibly this cosmic Mind does no more than agitate things so as to bring off decomposition and recomposition; but Anaxagoras nonetheless speaks of its activity as diakosmein, “ordering” (DK59B12), and also speaks of the global product of 23

This major theme of Empedocles’ almost reverential attitude toward effects of qualitative “emergence” (or “supervenience,” or epigenesis) is taken up at greater length below in the next Chapter, “Quality, Structure, Emergence: in Anaxagoras, Philolaus, Democritus, and in Empedocles”.

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

67

that activity as a kosmos, which in Anaxagoras undoubtedly bears the sense of “world order” (BK59B8). An important property of Anaxagorean stuffs is conveyed by a term that may in fact have been coined not by him but by later philosophers: homoeomery, “sameness of parts.” Water, for example, is homoeomerous in the sense that, just as the whole of a certain aqueous mass is water, every part of that mass, no matter how small, is also water. It is, as one might say, water through and through. Plato was later to contrast homoeomerous entities with ones that are nonhomoeomerous, giving such examples of the latter as a hand or a face (the parts of a hand are not a hand; the parts of a face, not a face). Let me prepare the statement of a significant contrast in Anaxagoras by interposing a systematic comment. It is not clear that forms generally are nonhomoeomerous: straight line is surely a form, and it can be thought of as made of straight lines. But shapes certainly are nonhomoeomerous: a circle is not made of circles, a triangle need not be thought of as made of triangles only. Thus, by both emphasizing the homoeomery of stuffs and declaring that the activity of Mind produces spatial arrangements, Anaxagoras in effect draws attention to the contrast between constituent stuffs and shapes. Now whereas the constituent stuffs may be phenomenologically recessive at one time and dominant at another, this cannot be said of the shapes and structures that emerge as Mind orders things. The emergence and disappearance of the forms of plants, animals, and other complexes are not an effect of illusion; we cannot blame the effect on the “weakness of our senses” (B21). Anaxagoras cannot acknowledge that these effects are real without violating his Eleatic conscience. But it would also be wrong of him to blink these effects as mere by-products of the metaphysics of composition-decomposition. In the evidence for Anaxagoras, we find references to spermata, “seeds,” not only in his biological doctrine but in cosmogonic contexts as well (notably in B4). So, perhaps he was beginning to feel the need for extending mechanisms of preformation from embryology to cosmology and cosmogony. But the references do not amount to more than hints. So, for the purposes of our narrative on Form, it would not be unfair to observe that Anaxagoras failed to notice the violation of Parmenidean-Eleatic scruples that apply to effects of compositionrecomposition-decomposition. Even anything like the Empedoclean attitude of reverential awe in pondering such effects is not to be found in the evidence for Anaxagoras. MELISSUS AND THE ELEATIC REJECTION OF TRANSFORMATION In connection with Parmenides, I offered earlier a translation of two lines of the isolated (preserved out of context) fragment DK28B4. I reproduce the

68

Alexander P. D. Mourelatos

translation of the two lines here, without listing the alternative possible translations: The what-is will not sever its connection with what-is, either by virtue of scattering in every direction, every which way across the world (kata kosmon), or by virtue of its being gathered together. (B4.2-3).

Appearing outside the main body of the deductions of what-is, these lines may not have drawn the attention they merit. The implications were driven home by Melissus in a passage of his treatise, “On Nature or On What-Is.” It is no exaggeration to say that this Melissus passage represents a major turning point in the history of Western philosophy: It is not possible that (what-is) should come to be rearranged (metaNRVPƝWKƝQDL); for (we cannot suppose that) the arrangement (kosmos) that existed previously perishes, or that an arrangement that does not previously exist comes into being. (DK30B7[3])

In the section that immediately precedes, Melissus argues that any sort of change from a given state to another involves the perishing of the first state and the coming into being of the second. In the passage just quoted Melissus points out that this applies with full force if the change at issue is one of arrangement. Of course, what Melissus is concerned to deny, is, in the first instance, the possibility of an internal rearrangement of “the One” (B8, to hen, which is how Melissus eventually and characteristically refers to the subject of his deductions, after initially adhering to Parmenides in referring to it as “what-is”). But the argument fastens on rearrangement as such; so, the argument also goes through if the rearrangement should be one of a plurality of externally related entities. It is worth noting that Melissus could have used a more direct logical strategy. He could have argued that the One cannot undergo rearrangement for the simple reason that it can have no arrangement: for, quite obviously, an arrangement involves distinctions and limits, and it is one of the earliest deductions of Melissus that the One involves no limits. Had he used this logical strategy he would have simply fomented the Eleatic paradox of the unreality of Form. In fact he argues by conceding – if only provisionally or dialectically – the reality of Form. He is saying in effect: You cannot disregard kosmos, “arrangement”; it, too, is a something, a what-is; and so PHWDNRVPƝVLV, “rearrangement,” amounts to a perishing and a coming-into-being. By having opted for this indirect strategy, Melissus became the first philosopher to entertain the proposition that Form may be fully as much a reality as a thing, or a stuff, or a power. We have no way of knowing that Melissus was criticizing Empedocles or Anaxagoras; he may simply have been anticipating their general strategy of answering Elea. But it is clear that we have a dialectical responsion here, and it

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

69

can fairly be said that Melissus’ rejoinder is devastating. Recall the distinction between Form as pattern of relations and Form as resulting structure. Empedocles cannot avoid holding that the elements frequently – even regularly – change their relations vis-à-vis one another. Unlike the pseudo-change (“pseudo-” in Empedocles’ eyes) involved when mortal forms come into being, the relational transposition of the elements themselves is indisputably real change – underived and fundamental. And even though Anaxagoras avoids the postulation of elements, corresponding points could be made against him. Julius Weinberg dwells on this great moment of the Melissan rejoinder in his essay, “The Concept of Relation: Some Observations on Its History.” 24 But because the focus of Weinberg’s investigation was strictly on relation rather than on Form, Weinberg did not point out that the Melissan rejoinder also proclaimed a fresh and intriguing challenge and inspired a new program for answering Elea. Consider this line of reflection. We could think of Empedocles and Anaxagoras as having worked out principles of pre-existence and conservation of the material aspect of reality. If we could similarly work out principles that provide for the pre-existence and conservation of the structural, configurational, or relational aspect of reality, the Melissan rejoinder would be disarmed. My thesis is that this, in essence, is the line that was pursued, in different ways, by Philolaus and by the two Atomists. PHILOLAUS OF CROTON: THE SCHEME OF “UNLIMITED THINGS” AND “LIMITING THINGS” Before the publication of Burkert’s Lore and Science, the prevailing opinion among students of Presocratic philosophy was that we have no genuine fragments from any work by Philolaus of Croton. The twenty or so bits of text attributed to him in various ancient sources were all thought to be post-Aristotelian forgeries. But Burkert showed that, in the case of at least nine of these fragments, the suspicions of modern scholars were not justified: in both content and style, these nine give every sign of having been composed in the latter half of the fifth century. 25 Burkert offered his own reconstruction of the essentials of Philolaus’ cosmology, and this has encouraged other scholars in subsequent years to attempt variant interpretations. 26 In due course, a full chapter that was devoted to Philolaus appeared in a survey of Presocratic philosophy – in volume 2 of The Presocratic Philosophers by Jonathan Barnes – and that chapter bears the inviting title “Philolaus and the Formal Cause.” Most importantly, since 1993 we have a complete critical edition of the Philolaus fragments and

24

A chapter in a collection of his essays, Weinberg 1965, pp. 61-119. See Burkert 1972, pp. 238-77, with summary at 276-77. 26 For a very attractive early such account see Nussbaum 1979. 25

70

Alexander P. D. Mourelatos

testimonia in Carl Huffman’s magisterial work, Philolaus of Croton: Pythagorean and Presocratic. This is a happy turn of events, because, out of the nine fragments whose authenticity Burkert had defended, six contain fascinating material. Fragment 1 states: Nature (physis) in the world-order (kosmos) was fitted together (harmoFKWKƝ) both out of things which are unlimited (DSHLUǀQ) and out of things which are limiting (NDLSHUDLQRQWǀQ), both the world as a whole and all the things in it. (Huffman 1993, p. 93. – In the rest of the present section on Philolaus, subsequent references by pages alone are, of course, to Huffman 1993.) 27

This is again stated in fragment 2, as the conclusion of an argument in which the alternatives, that the world consists only of limiting things or only of unlimited things, are eliminated (pp. 101-113). Huffman, drawing his examples from the actual sources for Philolaus, shows that the “unlimited things” are entities that can be viewed as inherently constituting a “continuum”: scil., such entities as fire, air, breath (indeed stuffs generally) along with the quasi-substantival opposites (hot-cold, light-dark); but also entities conceived more abstractly, viz., the void, time, sound (pp. 4153). Correspondingly, the “limiting things,” which are also referred to by Huffman as “limiters,” are such entities as shapes, numbers, specifications of place (for instance “middle” or “center”) or specifications of pitch. We already saw that Melissus (in DK30B7[3]) had acknowledged – if only indirectly and dialectically – metaphysical parity between, on the one side, shape/figure/arrangement and, on the other, Being as material constituent or stuff. In Philolaus that metaphysical parity is recognized unequivocally and is made fundamental in his philosophy. As Huffman puts it, “Philolaus is approaching something akin to a distinction between form and matter” (p. 52). But Philolaus goes even beyond parity; epistemologically, he gives priority to Form. Fragment 4 reads: And indeed all the things that are known have number. For it is not possible that anything whatsoever be understood or known without this. (p. 172)

Fragment 3 reinforces this by making the complementary point that the act of knowing is itself a determining of something, a limiting: For there will not even be anything that is going to know (WRJQǀVRXPHQRQ) at all, if everything is unlimited. (p. 115)

It is now relevant to ask: Beyond agreeing with Melissus in treating Matter and Form on a par, does Philolaus have an answer to Melissus’ challenge to the 27

And for Huffman’s full comment on the fragment, see pp. 94-101.

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

71

possibility of transformation or transfiguration, to the genesis of arrangement (kosmos) “X” from the prior existence of arrangement (again kosmos) “W”? At first sight, he appears to be more of a mind to defy the Eleatics than to answer them. For not only does he use the language of genesis freely, at one point in fragment 6 he speaks – shockingly to Parmenides or to a Parmenidean – of a “having come-into-being” (JHJHQƝVWKDL) of “things-that-are.” And yet, the latter starkly un-Eleatic collocation is embedded in a long sentence in which the voice of Philolaus’ Eleatic conscience is unmistakably heard: [For] it was impossible for any of the things that are (HRQWǀQ) . . . to have come to be (JHJHQƝVWKDL), if the being (HVWǀ) of things from which the world-order came together (synesta) did not preexist (PƝ K\SDUFKRXVDV). (pp. 123-124)

Still, an Eleatic-minded critic might protest that positing eternally stable underlying realities is not sufficient. How has Philolaus come to terms – if he has – with the Melissan rejoinder? The crucial problem, you will recall, is that of the emergence of patterns of arrangement of the stable underlying realities. Philolaus can plausibly argue that at the original cosmogonic stage no emergence of Form is involved. The unlimited things, taken in themselves, cannot be said to be in any sort of pattern of relations vis-á-vis one another. Whatever organization, articulation, or configuration they ever possess will be conferred upon them by the relevant advent of likewise pre-existing limiting things. Suppose that the first compound that comes to be in the world is a snowflake: neither the end product nor the generative process (however the latter is conceived) involves the cominginto-being of a structure that is not already prefigured in that special pre-existing limiting thing, viz., the snowflake-shape. But how are we to understand the metaphysics of events of generation that are subsequent to the initial stage of cosmogony? That Philolaus did address this problem seems certain. For in fragment 2 he specifically recognizes that things compounded out of the limiting ones and the unlimited ones (IJ੹ ਥț ʌİȡĮȚȞȩȞIJȦȞ IJİ țĮ੿ ਕʌİȓȡȦȞ) may in turn provide either the stuff or the form for further combinations (ʌİȡĮȓȞȠȞIJȚ țĮ੿ Ƞ੝ ʌİȡĮȓȞȠȞIJȚ). His argument would then be that when, say, a handful of snowflakes are compacted into a snowball, nothing new comes into the world; rather, a certain other pre-existing limiting thing, the spherical shape, comes to be superposed. PHILOLAUS’ RECOURSE TO THE PRINCIPLE OF HARMONIA Such effects of post-cosmogonic advent of Form are not, however, free of conceptual difficulty. Unlike the original water stuff that was constituted into a snowflake, the individual snowflakes already have a history of relational ar-

72

Alexander P. D. Mourelatos

rangements that are independent of their eventual compaction into a snowball. Are we back to where things stood in Empedocles? Or will Philolaus posit distinct advents of limiting things for each moment of rearrangement or reconfiguration? The proliferation of pre-existing limiting things we would then have to posit appears mind-boggling. Besides, there is the difficulty of the crushing of all those snowflake shapes in the compaction. Philolaus would at best be answering only half of the Melissan rejoinder (ignoring the half concerning the perishing of the previous arrangement), and he would be violating his own tenet that “limiting things” last forever. These difficulties arise because we have not yet taken into account the third principle of Philolaus’ metaphysics, that of harmonia, “harmonious structure” or “attunement,” which is introduced in a later passage of fragment 6: But since these beginnings preexisted and were neither alike nor even kindred to one another (homophyla), it would have been impossible for them to be ordered, if a harmony (harmonia) had not come upon them (epegeneto), in whatever way (KǀWLQLǀQDQWURSǀL) it came to be. Like things (homoia) and things that are kindred (homophyla) did not in addition require any harmonia, but [in the case of] things that are unlike . . . it is necessary that such things be bonded together (synkekleisthai) by harmonia. (pp. 123-124) 28

A harmonia is the unification of the many. When harmonia supervenes, the component parts lose their distinct identity; and yet they are preserved in the whole, which is necessarily a one that is also a many. It is at this point that it is right to bring into the narrative the themes of musical theory, number theory, the Pythagoreans, and Greek mathematics. 29 Philolaus belongs to the younger of the two lines that claim Pythagoras as their ideological ancestor, the mathematici. The harmonizing of Pythagorean mathematics with speculative cosmology took place not in the sixth century but in the latter half of the fifth; and the dissonance it sought to resolve was precisely the Eleatic Crisis. The point has been made well by Burkert, so I shall simply quote him: [H]ow was Parmenides to be overcome? Faced by this problem, the Pythagoreans were able to enlist the help of the technique of calculation . . . . Out of a few fundamental figures, the numbers from 1 to 10, there develops an inextricably complicated system. The “generative” character of mathematics broke through the inflexibility of the Eleatic system: plurality and becoming are not unthinkable and unsayable after all, but they can be thought and expressed in the form of numerical operations. 30 28

I have left harmonia without translation. Moreover, instead of Huffman’s translation of homophyla as “related things,” I have substituted “things that are kindred.” 29 See Huffman pp. 54-56 and 64-77. 30 Burkert 1972, p. 49.

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

73

We might perhaps object that Burkert (or Philolaus and his fellow Pythagoreans) are equivocating. The sense in which numbers are “generated” in accordance either with the rule of a series or the rule of a proportion is not the same as the sense in which a plant or an animal is generated. In all likelihood, this distinction was not as obvious in the fifth century as it certainly came to be in the fourth. Be that as it may, the issue that has concerned us here is that of the emergence or generation of shape, structure, configuration; and these are precisely the sort of entities with respect to which mathematical “generation” is quite appropriate. Even so, speaking on behalf of the Eleatics, one might further protest, that the requirement of no generation and no perishing cannot be met by providing for qualified (“in a sense”) pre-existence and conservation. But once again it is relevant to cite the distinction I drew earlier between two strands in Parmenides’ argument. 31 The strand that was picked up by Melissus and by postParmenidean cosmologists was the first one, the one that led out from the intuition that there is no coming-to-be out of nothing and no perishing into nothing. If Philolaus is to have a good Eleatic conscience, he need not be expected to enforce the Parmenidean requirement of no-generation and no-perishing precisely as that requirement was conceived by Parmenides; he is entitled to reach behind it, as it were, to the original intuition and to its source. That source is a demand for total intelligibility: there must be no mysteries, no capricious happenings, no miracles, no brute facts. It is also true of Empedocles and Anaxagoras, of course, that they harked back to the original demand for intelligibility. But for all the richness and ingenuity of their respective schemes, the Philolaic program holds much greater promise of dispelling mysteries. Because the constitutive entities of these other two systems are plastic and inherently amorphous stuffs – “unlimited things” only – there is a total absence of formational and transformational constraints. At any one time any formation is possible, and over a stretch of time any sequence of formations is possible. Here we are being brought back to the issue that was touched upon in connection with Xenophanes’ cloud astrophysics. 32 Think of a cloud as it presents itself to an observer. Putting to the side super-sophisticated explanations offered by twenty-first century meteorology and chaos theory, if over a period of, say, five minutes a cloud should successively assume the shapes first of a centaur, then of a leopard, then wolf, then bull, 33 this would pose no epistemic surprise or need of explanation. In the cosmology of Xenophanes, that is exactly how things stood with clouds, whether ordinary or astrophysical. And this is still how they stand with respect to the constitutive stuffs in the systems 31

See above, Section titled “Two Strands in Parmenides’ Argument.” Section titled “The Special Contributions of Xenophanes of Colophon.” 33 Retaining, of course, the examples from Aristophanes’ Clouds 346 ff. 32

74

Alexander P. D. Mourelatos

of Empedocles and Anaxagoras. No patterns either of precedence or of continuity are posited or expected. Inevitably then, all transitions from one state of the system to another are brute mysteries. True, there is, in Empedocles, a pair of cosmic forces; and of the two, Love often appears to play the role of a cosmic craftsman. In Anaxagoras, we have a cosmic Mind and (though this one detail was not dwelled upon earlier), we have references to “seeds” that may serve to provide some recourse to pre-formation. 34 But these devices only push the mystery to another level. As Aristotle repeatedly complained, why do these forces act in one way at one time but differently at another? FORM AS THE SOVEREIGN PRINCIPLE IN THE PHILOSOPHY OF DEMOCRITUS Before I turn to the fourth and final phase in my narrative, I need to introduce a simplification in my references to the two figures at issue. Our ancient sources often do not distinguish between the views of Leucippus and Democritus. But it is not right to refer to them simply as “the Atomists,” 35 since that designation also takes in Epicurus in the fourth century, as well as the Roman philosopher Lucretius, and other Epicureans. Of the two founders of Atomism, it was Democritus who was the prolific author, and it is certainly he that was the philosopher of stature. So, instead of “fifth-century Atomists,” I shall simply henceforth say “Democritus.” If we should think of Democritus as a critic of Empedocles and Anaxagoras, his approach parallels that of Philolaus. We saw that Philolaus’ aim was to introduce into the mixing bowl of amorphous Empedoclean-Anaxagorean stuffs the factors of “limiting things” and of harmonia so as to provide mathematical intelligibility by organizing and articulating the stuffs into discrete shapes and enumerable or measurable quantities. Democritus’ theory has the great advantage of ontological simplicity over that of Philolaus. Instead of postulating a separate class of entities as the bearers of Form, Democritus conceives of stuff as fundamentally parceled out into tiny bits that intrinsically have shape. 36 But since not only the stuff of these corpuscles but also their individual shapes must be unborn, imperishable, and changeless, the corpuscles – unlike the familiar stuffs of earlier philosophy – can have no plasticity: they must be absolutely full, rigid, and impenetrable. Next in the system’s logical agenda comes the concern to do justice to the evident facts 34

These components in the theories of Empedocles and Anaxagoras are discussed more fully in the next Chapter. 35 Which is the solution adopted in the Loeb set on Early Greek Philosophy, Laks–Most 2016, vol. 7. 36 The ancient evidence for doctrines of the “Atomists” (my Democritus) is perspicuoously presented in well-organized thematic units in Laks–Most 2016, vol. 7, pp. 7-12, and then laid out in the “D” (for “Doctrine”) section, pp. 60-345. – My own discussion of Democritus here is supplemented in Chapters 6 and 7 below.

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

75

that there is both motion in the world and a great variety of textures – ranging from very hard to very soft. These facts indicate that the corpuscles are not packed tightly against one another but are dispersed in some sort of accommodating medium. Now if that medium were another sort of stuff, its accommodating nature merely a function of its softness or looseness, then the medium’s intrinsic shape would constantly undergo change as the corpuscles swim through it. (And there would also be some unanswerable questions: why is this medium uniformly loose? if there can be a loose sort of stuff, why is there no variation of texture among corpuscles?) The only reasonable assumption is that what accounts for the dispersal of the corpuscles is an entity that is no medium and no stuff at all – sheer emptiness, the void. And so Democritus, emphaticcally disassociating himself from the second strand of Parmenides’ argument – the one that leads from the semantics of “what-is-not” – speaks of the principle of dispersal alternately as “the void” (kenon), but also and more poignantly as “nothing,” and as “what is not.” I have spoken of corpuscles, but it is now time to refer to them by their original designation of atoma or atomoi, “indivisibles.” I must simply skirt the great scholarly controversy as to whether Democritean atoms are only physically unsplittable or whether they are simples in some strong metaphysical sense. The point I want to stress is that the atoms are units not only of stuff but also of Form. The feminine atomos and the neuter atomon are adjectives, and they are syntactically construed in our sources either (i) with nouns they modify or (ii) as nominalized adjectives, with the article prefixed. The nominalized feminine occurs more frequently than the nominalized neuter, and there is good reason to think that the original Democritean collocation was atomos idea, “indivisible form” (cf. DK68A57). Significantly, one of the works of Democritus bore the title 3HUL LGHǀQ, “On Forms” (B6), and the ancient dictionary of Hesychius, in a transparent allusion to Democritus, explains that idea is sometimes used with reference to “the smallest body” (B141). 37 In one of his discussions of Atomism, Aristotle says that “Democritus and Leucippus having set the shapes (SRLƝVDQWHV WD VFKƝPDWD) produce (poiousi) generation and alteration from these” (De gener. et corr. I.2.315b5-6), where “setting the shapes” obviously refers to the postulation of atoms. The ontological economy achieved by introducing Matter and Form together, at the most fundamental level of reality, not through separate prinicples (as in Philolaus) but in the unified and simple conception of the atom, is impressive indeed. But even more impressive are the resources the conception provides toward handling of the problem of relational arrangement and re-configuration. When one first thinks about this it may well seem that Democritus had no answer to the Melissan rejoinder. Consider the case of three atoms falling 37

In Laks–Most 2016, vol. 7, at D34, p. 96.

76

Alexander P. D. Mourelatos

through the void in a triangular formation; then, one is knocked off its path by a careening fourth atom, with the result that the original three now fall in a straight row formation. Surely the configuration of the triangle has been destroyed and that of the rectilinear moving row has come into being. But a Melissan reading of this simple example overlooks the rich fabric of causal continuities Democritus posits. Besides shape and absolute rigidity, Democritean atoms have one other intrinsic property, size. There is an infinite number of atoms, and there is also infinite variation of atomic shapes and atomic sizes; but, of course, each atom has eternally the same determinate shape and size. Now atoms also have the properties of motion, spatial position (GLDWKLJƝ or taxis), and spatial orientation (WURSƝ or thesis). But although they eternally possess these properties as determinables, the determinates of motion (including speed and direction) position, and orientation will vary from time to time for each atom. For these determinates are all properties an atom acquires as a result of previous encounters or entanglements with other atoms. Thus, the motion of a Democritean atom at a given time is, in modern terms, the resultant of two component motions: the motion imparted to it by the most recent “blow” (SOƝJƝ); and its own antecedent motion, which in turn is a resultant of two components, and so on. (The infinite regress involved is strictly one into an infinitely receding time. Accordingly, given Democritus’ doctrine of the infinity of time, the regress need not be regarded as vicious.) Similarly, the atom’s spatial position and orientation at a given time is the kinematic upshot of a history of encounters and entanglements with other atoms. This chain of motions and transpositions is strictly mechanistic and deterministic; and so, Democritus is in a position to argue that the motion, position, and orientation of an atom at a given time is prefigured in its own kinematic history taken together with the kinematic histories of all atoms with which it has come in contact. But there is also a prefiguring in a stronger and literal sense. Since there is no Love or Strife or Anaxagorean Mind in Democritus’ system, the only determinants of the motion of Democritean atoms – apart from the general metaphysical principle that atoms should have some motion – are shape and size. The motility, as we might put it, of atoms is strictly a function of their particular shape and their size. The smallest, spherical ones will be the most mobile; atoms with faces that are either smooth and flat or smooth and convex will smartly rebound upon collision; elongated ones will spin; hooked atoms will have long intervals of rest as they become trapped into compounds by atoms with deep cavities; barbed ones will tend to get tangled and will move sluggishly; cubical atoms will stack with various other polyhedral atoms – and so on. Generally, how two atoms will respond upon contact is determined by three factors: shape, size, and force of collision. But the force of collision is itself largely determined by shape and size.

3. Form as Shape, Array, Order: Earliest Approaches and Eleatic Scruples

77

Whereas Democritus used the term idea, “form,” to refer to the atoms directly, he had an unusual way of discussing differences of shape among atoms. He spoke of their rhysmos (DK67A6), 38 a word that is the Ionic equivalent of rhythmos, the latter (of course) being the Greek from which we get the English “rhythm.” A cognate of the word is UKHǀ, “to flow,” which seems to be the focal meaning of rhythmos or rhysmos (and our “rhythm” attests to the enduring pull of that semantic focus). This striking usage makes, in effect, the point developed in my preceding remarks: differences of shape determine the kinematics of atoms, their “flow.” 39 But I must face an objection. It is by no means certain that Democritean atoms move only by “blow” (SOƝJƝ, Latin plaga). While some of our sources draw or imply a contrast between Democritus and Epicurus (the latter of whom definitely posited weight for atoms), other sources attribute weight to Democritean atoms as well. 40 I am inclined to believe that it is the former who have given us the authentic Democritean doctrine. But if I am wrong, it can still be said that shape and size are the main determinants – rather than the only determinants – of the kinematics of atoms. We may now go back to the example of the three atoms that were aligned into a straight row after one of them collided with a passing fourth. Had we taken into account the trajectory and speed of the fourth atom, also the shapes and sizes of the two which collided, we could have both predicted and fully understood the new alignment. And what holds in that simple case carries upwards to formations of larger and larger scale. Form, pre-formation, and transformation – that is the essence of Democritean metaphysics. The solid interior of the atoms is merely a prop that holds the outer form rigid; and the void is merely the stage on which the drama of transformation is enacted. The atoms themselves provide not just a given or limited repertoire of forms; they are infinite in number and come in an infinite variety of shapes. 41 So they are units of Form in a very strong sense: they are paradigms of all possible closed structures; and they are also modules for the production of all possible configurations. Appropriately enough, we find not only rhysmos in Democritus’ vocabulary but also the verb PHWDUUK\VPHǀ, “to change figure,” “to transform,” and the abstract noun DPHLSVLUK\VPLƝ. The latter appears in the title of a whole book, 3HUL DPHLSVLUK\VPLǀQ “On Exchanges of Form.” 42 Two other titles show Democritus’ conscious intent to present his doctrine as a philosophy of Form: Megas diakosmos, which we 38

In Laks–Most 2016, vol. 7, at D31 and D32, pp. 90-96. See also vol. 1, at the set’s Glossary, p. 252. See also below, Chapter 7, “Intrinsic and Relational Properties of Atoms in the Democritean Ontology.” 40 Laks–Most 2016, vol. 7, very helpfully presents the evidence for each alternative under separate subheads: pp. 8, 106-09. 41 Laks–Most 2016, vol. 7, pp. 106-09. 42 Laks–Most 2016, vol. 7, p. 64 (at D2bV) 39

78

Alexander P. D. Mourelatos

might translate, “An Account at Large [i.e., in general terms] of the Structure of Things”; and Mikros diakosmos, “A Detailed Account of the Structure of Things.” (I merely note that some of our sources attribute the first work to Leucippus. 43 The point I am making, that fifth-century Atomism is first and foremost a philosophy of Form, still holds.) THE TWO CLASSICAL PHILOSOPHIES OF FORM: PLATO, ARISTOTLE My narrative now begs for an epilogue. The Philolaic-Democritean program of answering Elea by emphasizing Form and then adopting a pre-formationist solution to puzzles of generation and perishing was famously continued by Plato and by Aristotle. Plato is closer to Philolaus. Plato’s Forms, like Philolaus’ “limiting things,” are ontologically separate. But they are no longer just another sort of concrete thing; they are abstract, purely intelligible, and transcendent principles of structure, order, and meaning. Still, this way of characterizing Platonic Forms is probably true only of Plato’s most mature thinking. In early and some middle Plato, Forms are often conceived of as things or quasi-thing 44 – thus, in the terms of the present account, close to the way in which Philolaus had conceived of his “Limiting Things.” The counterpart to Democritus is Aristotle. Stuff and Form are separated only conceptually. Form very definitely is not just another sort of thing; nonetheless, Form is always (except in the special case of the Unmoved Mover) immanent in a concrete thing, in a substance. But Aristotle diverges from Democritus in two important respects: The Aristotelian answer to Elea does not ever call for the pre-existence and preservation of some physical tokens of, say, square shape; it suffices that there should be continuity of Form-type across transformations, that there be the specific identity of square shape, not the numerical identity of a square piece. Most importantly, Aristotle emphasizes that Form is not just shape, not even shape in the kinematically pregnant sense of Democritean rhysmos. Form is preeminently ergon, “function,” and physis, “nature” – the latter defined as an “innate principle of motion and rest” (Phys. II.1.192b). 45 And so the Aristotelian conception of preformation finds its most appropriate model in the development of biological organisms and the continuity of biological species: “It is a Human Being that begets a Human Being” (Metaph. XII.3.1070a).

43

Loc. cit. at D2bIII. As Prauss 1966 has argued passim and at length. 45 Nussbaum 1987 has rightly emphasized this in her Respondent’s comment on Sections I-VII of what in the present collection appears as Chapter 4. 44

4. Quality, Structure, Emergence: in Anaxagoras, Philolaus, Democritus, and in Empedocles

The development of cosmology or natural philosophy in the Greek fifth century B.C.E. used to be given a Hegelian narrative: Heraclitus had proclaimed a doctrine of flux and diversity; the Eleatic philosophers (Parmenides and Zeno of Elea, plus Melissus of Samos) had countered with a doctrine of eternal changelessness and monism; most of the cosmologists of the middle and late fifth century – Anaxagoras, Empedocles, Philolaus, Democritus, Diogenes of Apollonia – sought a compromise between the speculative extremes represented by Heraclitus and by the Eleatics. Students of Presocratic philosophy have known for several decades now that this elegant triadic narrative of thesis, antithesis, and synthesis is both simplistic and incorrect. Heraclitus is not purely a philosopher of flux; he gives equal emphasis to the constancy and stability found in flux, to the unity found in diversity. (Even so, textbooks and encyclopedias continue – down to our own day – recounting the triadic narrative.) Among cognoscenti, the approach that has been favored articulates the development on a dyadic scheme. It leaves Heraclitus out, grouping him with the Ionians of the sixth century, and concentrates on canvassing the logic of exchanges (both actual and putative) between, on the one side, the Eleatics and, on the other, the later cosmologists, aptly called “Neo-Ionians” by Jonathan Bames. 1 In this modern canvassing of filth-century philosophical dialectic, the Eleatics have been lionized. They are perceived as having issued a logically exacting “challenge” and as having precipitated a methodological and episteOriginally published under the title, “Quality, Structure, and Emergence in Later Pre-Socratic Philosophy,” in John J. Cleary, ed., Proceedings of the Boston Area Colloquium in Ancient Philosophy, vol. 2, New York, University Press of America, 1987, pp. 127-94. At the time of that first publication, this study was the ad interim outcome of research that was supported by grants awarded by the National Endowment for the Humanities (for 1982-83), and by the University Research Institute of The University of Texas at Austin (for spring 1978 and spring 1983). I gratefully acknowledge these awards. Successive ancestral versions of Sections I-VI were nurtured with helpful and challenging comments offered by academic audiences in the U.S.A., in Australia, and in England. Parts of these same Sections, and with the addition of Section VII, were read in 1986 at Boston University, under auspices of The Boston Colloquium in Ancient Philosophy, and received the benefit of excellent discussion by Martha Nussbaum, who served as Respondent on that occasion (see Nussbaum 1987). 1

See Barnes 1979a, passim.

80

Alexander P. D. Mourelatos

mological “crisis,” every subsequent philosopher in this period thus being portrayed as struggling mightily to come up with an adequate answer or appropriate resolution. This approach to the development of later Presocratic philosophy has been most conspicuously promoted by scholars in the AngloAmerican tradition. 2 But the approach has also been favored outside this tradition and indeed earlier. 3 Just as the thesis-antithesis-synthesis approach betrays the influence of Hegel on the historiography of philosophy, the approach in terms of Eleatic challenge and Neo-Ionian answer reflects the fascination of twentieth-century philosophers – and indeed of twentieth-century academics generally – with what might be called the paradigm of a “debaters’ forum.” This dyadic twentieth-century approach has produced a rich fund of insights into the dialectic of post-Parmenidean philosophy. It is, indeed, a conceptually fruitful approach, one that deserves to continue to be pursued. But, as we become more sophisticated in our appreciation of details and nuances in the postParmenidean systems, we also need to be on guard against some of the shortcomings and some of the excesses of the dyadic approach. To the extent it employs the principle of interpreter’s charity, this approach tends to do so primarily to the benefit of the Eleatics. For when Neo-Ionian cosmologists are judged strictly by the standard of the Eleatic elenchus, they are inevitably found inept. 4 The Eleatics did not argue merely against a generation ex nihilo and a reduction in nihilum; they argued against any sort of generation, any sort of perishing, and any sort of change. Moreover, Parmenides argued – at the very least – against making any distinctions that involve contrariety, and this notably includes the distinctions implied in locomotion; Melissus and Zeno argued – unmistakably – against countenancing any sort of plurality. The NeoIonians, positing as they do, without so much as a whisper of a counterargument, ontologically fundamental pluralities, and positing the reality of locomotion, are made to appear guilty of a gross ignoratio elenchi. To hold the Neo-Ionians accountable to the full Eleatic elenchus strikes me as more than mildly unhistorical. We recognize that some of the semantic and 2

In the original edition of The Presocratic Philosophers (1957), G. S. Kirk and J. E. Raven wrote: “Both Empedocles and Anaxagoras repeatedly and clearly reveal, not only by their thought but also by the language in which it is expressed, an almost servile observance of the Parmenidean demands” (p. 319). The corresponding paragraph in the second edition of 1983 omits this statement and adds several qualifying remarks. But it does make the same announcement made in the original edition, of an exegetic program of “showing, where possible in the philosophers’ own words, how these post-Parmenidean systems are deliberately designed to take account of the findings of the Truth [Parmenides’ poem],” and makes only minor changes to this line of interpretation (Kirk–Raven– Schofield 1983, p. 351). 3 It would seem that the path was opened as early as 1916, with the appearance of one of last century’s great works in the field, Reinhardt 1916, a work which – in spite of some extremes of revisionism, such as placing Heraclitus and Xenophanes after Parmenides – has inspired hundreds of classicists and philosophers in several national schools and traditions. 4 Michael C. Stokes 1971, p. 35, speaks of “the complete logical failure of these systems [those of the Nen-Ionians] to tackle the fundamental problem of Eleaticism”. Barnes 1979a, II, p. 140, correspondingly uses the term “flop” (cf. pp. 11-15, 125-45, and I, p. 216).

4. Quality, Structure, Emergence

81

epistemological assumptions that generate the Eleatic elenchus are faulty and naive. Indeed, it seems that those assumptions are so deeply embedded conceptually that only a sophisticated analysis, one that lies beyond the intellectual horizon of the fifth century, could ferret them out. 5 By contrast, Neo-Ionian cosmologies have the virtue of a “robust sense of reality”; some even give us the rudiments of a world view that has been vindicated by the development of modern science since the Renaissance. Are we not being rather schoolmarmish when we insist, as it were, that the Neo-Ionians should have fulfilled the relevant curricular pre-requisite of the Eleatic etenchus before they could have been entitled to graduate to what, after all, is a more tenable position than that of the Eleatics? This hermeneutic parti pris in favor of Elea goes against much that we have learned in the twentieth century about the process of reading philosophical texts. We freely grant that a creative philosopher – be he Aristotle, or Augustine, or Russell – reads the classics from his own theoretical perspective. We similarly concede that all observation is “theory-laden,” and that even historians of philosophy cannot achieve, and indeed should not even attempt, a reading of classical texts that is free of all theoretical commitment. Why, than, should we insist that the Neo-Ionians should have shut their eyes to a vision of the cosmos that was authentically theirs and should have attended instead with scrupulous care to the precise terms of the Eleatic elenchus? It might be thought that this exercise, of trying to gauge the degree to which any one of the Neo-Ionians succeeded in meeting the Eleatic challenge, unrealistic though it may be as history, is nevertheless instructive as philosophy. This is certainly true. But too much of this regimen of exercise may impair our ability to appreciate certain highly creative conceptual moves made by the Neo-Ionians. These moves show, when studied in their own proper context, deep insight, whereas, if judged against the unrefracted Eleatic standard, they tend to appear evasive or irrelevant or fatuous. My theme in this study is to examine the dialectically resourceful moves made by those Neo-Ionians traditionally called Pluralists (i.e., Neo-Ionians minus Diogenes) as these philosophers aim either to articulate or to resolve a problem peculiar to pluralist cosmologies. Here is a general and preliminary statement of the development I perceive. At the constitutive level of reality – i.e., at the level of elements or ingredients – all qualities or shapes are to be considered, as the Eleatics had demanded and the Pluralists freely conceded, eternal and unchanging. But how can the Pluralists make intelligible the genesis of quality and structure at the constituted level, the level of everyday objects? To take an example from Empedocles: blood may be a certain combination of the four elements, each of them eternal and unchanging. But when blood is formed through mixture, where does the 5

For some of these modern readers, that sophisticated analysis is found in the later Plato, for others in Aristotle, and for still others it is not forthcoming before Frege or Wittgenstein or Heidegger.

82

Alexander P. D. Mourelatos

peculiarly life-giving power of blood, or even simply its red color, come from? When the elements articulate themselves further into the form of a tree, or fish, or bird, or beast, or human being, where does the shape and structure of these compound entities come from? For convenience I shall refer to this effect of genesis, in either of its two aspects – whether quality or structure – by the modern term “emergence.” Eleatic arguments did much to bring this problem to the surface. And yet the special urgency of the problem is felt only if Eleatic arguments have been read against the background of a pre-Eleatic demand for intelligibility. And it is precisely insofar as the Eleatic elenchus has come to be informed by that more traditional demand that the Pluralists are challenged to devise their variously resourceful solutions to the problem of emergence. I. THE RATIONALIST QUEST FOR INTELLIGIBILITY In the fragment 6 now standardly referred to as B8, the central argument of his “Truth,” Parmenides deduces four requirements for “what-is” or “the real” (to eon): it should be unborn and imperishable; indivisible and unitary; immobile; fully actualized (i.e., changeless). The Pluralists focused approvingly on the first of these, the requirement that it should not be liable to generation or perishing. The Parmenidean requirement that what-is should be totally immune to change they interpreted as a requirement that ontologically fundamental entities, taken severally or individually, should be immune to any sort of internal change. To the rest of the requirements, as well as to other Eleatic arguments, they responded either much more obliquely or not all. Characterizing the thrust of the Pluralist response to Elea this bluntly may seem to convey the very effect of that “uncharitable” approach to the NeoIonians of which I spoke at the start of this essay. We do, however, come to appreciate that the stance of the Pluralists vis-à-vis the Eleatics was neither evasive nor insincere as we take in the wider context. Contrary to what is sometimes said, the principles ex nihilo nihil and in nihilum nil are not Eleatic in origin. Historians of philosophy have often overlooked that the first of these two principles is stated in a one-line fragment of the lyric poet Alcaeus, 7 who was a contemporary of the very first of the Ionian philosophers, Thales of Miletus. The context of Alcaeus’ utterance has been 6

In references to the Presocratics, numbers preceded by the letter A refer to testimonia, numbers preceded by the letter B, to fragments, as in Diels–Kranz 1952. I henceforth refer to this edition by the initials DK. 7 țĮ઀ ț’ Ƞ੝į੻Ȟ ਥț įİȞઁȢ Ȗ੼ȞȠȚIJȠ: Alcaeus 76 (Bergk) = 23 (Diehl) = 320 (Lobel and Page). One of the manuscripts reads Ƞ੝įİȞંȢ, which is unmetrical. See Lobel–Page 1955, p. 263. The phrase ਥț įİȞંȢ may mean (a) “from something” or (b) “from nothing” or (c) “from any old thing.” Possibility (a) yields for the whole fragment the patently false and pointless sense, “And nothing could come to be from something.” Possibility (b) yields the ex nihilo nihil. Possibility (c) yields the stronger proposition, “And nothing could come to be from any old thing.” If the latter was what Alcaeus had expressed, then it stands to reason that he would have considered the ex nihilo nihil no less obvious.

4. Quality, Structure, Emergence

83

lost. Most probably it was not philosophical, in which case it is noteworthy that he counted on his audience to receive the principle as a commonplace. We do not know enough about Thales to be able to determine whether either of these two principles is implied in his cosmology. But the endeavors of other preParmenidean philosophers to establish some stuff or agency as one that either uniquely constitutes or generates or directs the cosmos are hardly intelligible except as responses to at least tacit acceptance of these principles. The point can be made of the apeiron of Anaximander, of the air of Anaximenes, of the one God of Xenophanes, and both of the “ever-living fire” and of the logos of Heraclitus. The intuitions of continuity and permanence that are captured in statements of these two principles are also implied in the doctrines of transmitgration and eternal recurrence that are well attested for Pythagoras in the sixth century. 8 What, in turn, lies behind these principles and intuitions may well be some deeply felt personal or mystical yearning for continuity and stability; but almost certainly, and in major part, what lies behind them is the (broadly speaking) rationalist quest for intelligibility. An entrenched Greek tradition, expressed in poetry and ritual, had pictured the world’s processes as routines in the behavior of quarrelsome and capricious gods, and so as subject to reversals and alterations at those gods’ whim. The philosophers challenged that tradition with their ideal of a “universe governed by law” – a law that was not only constant and inexorable in its application but also inherently intelligible. 9 Mysterious epiphanies and magical disappearances – sea changes, like those of the marine god Glaucus – were not credible. All change must have causally relevant antecedents and traceable consequences. It is, I believe, against that background and with that frame of mind that all the Neo-Ionians – not just the Pluralists – approached Parmenides and Melissus. They may well have felt, not at all unreasonably, that the latter were formulating reflectively, critically, and explicitly assumptions and principles that had been tacitly present in earlier speculation. Refusing to accept the Eleatics’ conclusions, they nonetheless judged that Parmenides and Melissus had done an impressive job of coming up with a rigorous deduction of the defining criteria of fundamental or ultimate reality. Zeno they approached quite differently. Democritus, for one, does seem to have been concerned to make his system immune to the Zenonian paradoxes of plurality, 10 but the rest of the NeoIonians must have seen Zeno, as Plato sometimes saw him, as “the Eleatic Palamedes,” an eristic type, a sophist. 11 This is not to say that they were wholly 8

See above, Chapter 1, “Presocratic Origins of the Principle That There Are No Origins from Nothing.” 9 See Vlastos 1975, ch. 1 “The Greeks Discover the Cosmos.” The phrase “universe governed by law” I borrow from ch. 3 of Kahn 1960, on which I also draw for the observations made above. 10 See Furley 1967, pp. 79-103 (repr. in Mouretatos 1974, pp. 504-26). 11 Cf. Solmsen 1971.

84

Alexander P. D. Mourelatos

unresponsive to Zeno. For there is at least one Zenonian paradox that holds special poignancy for the Pluralists, as I shall show in the next Section. All this said, I now wish to point up one crucial respect in which the Eleatic elenchus, unrefracted and at its most sweping range, had a profound and highly constructive effect on the Pluralists. Precisely because the Eleatics were prepared to argue against all forms of change, the Pluralists became progresssively more sensitized to the problem I introduced earlier under the rubric of “emergence.” Eleatic paradoxes may not have moved the Pluralists quite as much as twentieth-century interpreters have supposed. But the elenchus did have the effect of forcing the Pluralists to scrutinize their own systems with a view to detecting types and levels of change or transition that might count as violations, not just of Eleatic strictures but, more seriously, of that older and paramount philosophical demand for intelligibility. The next Section will attempt to spell out exactly what is involved in the problem of emergence. Sections III and IV will examine texts that indicate awareness of the problem among fifth-century philosophers. The remaining eight Sections will lay out and analyze the evidence of solutions that were offered primarily by two philosophers, Anaxagoras and Empedocles, with a brief notice, for purposes of comparison, in Sections VII and XII, of the solutions advocated by Philolaus and by Democritus. II. THE ISSUES OF EMERGENCE FOR PRESOCRATIC PLURALISTS Characteristic of all fifth-century Pluralist systems is the distinction between constituent parts and constituted wholes. At a fundamental level, constituent parts are eternally existent, and have only occurrent or actual properties, which are preserved unalterably. (It is important to note that when the Pluralists speak of a dynamis or “power” present in a fundamental constituent they must be taken to imply a power that is continuously, even eternally, manifest; nothing like the Aristotelian scheme of potencies that are triggered into actualities can be presupposed; nothing like Aristotelian qualitative alteration or qualitative interaction can be envisaged.) The only change allowed at this level is transposition or redistribution – either a mutual displacement, in the case of plenum Pluralists, or motion in the void, in the case of the atomists. With transposition we get formations of aggregates. In the case of plenum theorists, this comes about as a result of interpenetration of the constituents; in the case of the atomists, it comes about as atoms happen to become mutually entangled. With the formation of each aggregate, something ostensibly new comes to be in the world, in two ways. There are qualities that do not seem to be identical with qualities of any one of the constituents; and there is the shape, figure, or form of the aggregate that is not intrinsically that of any one of the constituents. How can the Pluralists account for any such new quality or any such new

4. Quality, Structure, Emergence

85

structure that comes into being? It is important for us to contemplate the problem in a formulation that borders on being simplistic. (A formulation that gives too much detail may already beg a particular solution.) In the formulation just given, I assume there is total absence of relevant antecedents for the new entities generated. The Pluralists, even as we allow that they should not to be expected to answer the Eleatics on the Eleatics’ specific terms, are nonetheless confronted here with an effect that their own Ionian heritage cannot sanction, a genesis ex nihilo. The problem could be given many names. It might be called the problem of secondary genesis, or of epigenesis; it might be called a Many-One or a OneMany problem; it might be called a problem of supervenient entities. I have chosen the term “emergence” because of its resonance with related problems in twentieth-century philosophy. 12 But we should note immediately that the term is in two respects misleading – and this would also be true of the terms “epigenesis” and “supervenience.” In the context of twentieth-century philosophy, “emergence” is associated more often with a certain solution rather than with a certain problem. To be sure, there have been advocates of physicalism or of the thesis of the unity of science who have used the term to mark effects that pose a problem for the program of reduction this group of philosophers favors. Typically, however, the term has been used by proponents of “emergent evolution” and cognate theories, who make positively the assertion that there are discontinuities or categorial leaps in the order of nature – e.g., that psychology is not reducible to biology, biology not reducible to physics, and so forth. My use of the term will resemble the use among the former rather than that among the latter group of philosophers. We also need to notice that, in the context of Presocratic cosmology, emergence is properly one of four facets of the problem involved. As aggregates are formed, it is possible that qualities and structures present at the level of the constituents may become obliterated in the constituted things. A formative process would result, ostensibly, in a reduction in nihilum. We might call this an effect of “submergence.” In processes of decomposition, there would correspondingly be a reduction in nihilum whenever an effect of emergence is reversed, and of generation ex nihilo whenever an effect of submergence is reversed. At least one of the Pluralists, Empedocles, has taken explicit note of these complexities (B17.3 “double is the genesis of mortal things and double their perishing”), but the problem is typically broached from its first facet only. I shall, accordingly, use the term “emergence” to stand by synecdoche for all four facets of the problem.

12

For a helpful general account of the concept of emergence, see two articles by Thomas A. Goudge: “Emergent Evolutionism” (1967) and “Evolutionism” (1973), esp. pp. 183-84.

86

Alexander P. D. Mourelatos

III. TWO ELEATIC ARGUMENTS THAT BEAR ON EMERGENCE By our best estimates concerning the chronology of fifth-century philosophers, Anaxagoras, Empedocles, and Zeno were contemporaries, whereas Melissus probably was ten-fifteen years their junior. It is, accordingly, misguided to base inferences as to who was responding to whom on considerations of chronology. It is probably safe to assume that Pluralist systems and post-Parmenidean Eleatic philosophy developed side by side, and that oral reports of doctrines propounded in the philosophical circles of Southern Italy, Sicily, Ionia, and Athens often preceded the circulation of manuscripts. (Plato’s dramatic convention or device of the “guest speaker from abroad,” or of “the narrator of an entire dialogue,” probably reflects a historical pattern.) The full force of the problem of emergence did not strike the Pluralists immediately. Two Eleatic arguments, one of which can be read as bearing on quality emergence, the other as bearing on the emergence of shape, may have functioned to hone the Pluralists’ sensitivity to the two aspects of the problem. The first argument is the least discussed of Zeno’s paradoxes, the Millet Seed (Simplicius In Phys. 1108.18; DK29A29). It has been preserved in the form of a dialogue, supposedly between Zeno and Protagoras. Here is the gist of the dialogue: A single millet seed does not make a sound when it falls; nor does a onethousandth part of a millet seed make a sound. But when ten thousand seeds of millet, a whole bushel, fall, they do make a sound. The ratio of parts to whole is, nevertheless, the same in the two cases. How can it be that, in the first case, from no sound we get no sound but we do get sound from no sound in the second?

Interpreters of Zeno have often appeared baffled by this particular paradox. 13 It does not have the form of the destructive dilemma, which is characteristic of Zenonian paradoxes; and it seems to be concerned with a problem of perception, whereas Zeno’s other paradoxes are robustly ontological. What it does resemble are puzzles involving open-textured concepts, such as the Megarian puzzles of the Heap and the Bald Man. In any event, it admits, as Aristotle saw (Phys. VII.5.250a9-28) of an easy solution, viz., to posit a threshold of audibility. In line with what I have said earlier concerning the dialectical relationship between Pluralists and Eleatics, we need not concern ourselves with the question of how Zeno would have intended the paradox. But it is clear enough that the Millet Seed must have had great poignancy for the Pluralists. For them it could have served as an ironic emblem of the effect of quality emergence: the constituent parts do not make a sound; the constituted whole does. How can that be? 13

Cf. Barnes 1979a, I, pp. 258-60.

87

4. Quality, Structure, Emergence

The argument that calls attention to the emergence of structure comes from Melissus: Nor is it possible that it [scil., to eon, “what-is”] should come to be rearranged (PHWDNRVPƝWKƝQDL); for (we cannot suppose that) the arrangement (kosmos) that existed previously perishes, or that an arrangement that does not previously exist comes into being. (B7[3])

In its Melissan context, the intention of the argument is unmistakable. Melissus aims to undercut any system that would envisage either an internal transposition of the parts of a single substance or a transposition of diverse constituents. Given their unswerving commitment to the reality of locomotion, none of the Pluralists would have felt threatened by the argument taken broadly. But in a philosophically strict and pointed construal, the argument does pose for them a serious problem of intelligibility. When an aggregate is formed, how are we to evade the implication that the previous arrangement of the constituent parts has perished in nihilum and that a new arrangement has come to be ex nihilo? (See also the discussion of this issue above, Chapter 3, Section on “Melissus and the Eleatic Rejection of Transformation.”) IV. EMPEDOCLES ON THE MARVEL OF EMERGENCE Among the Pluralists themeselves, the clearest and most eloquent recognition of the problem of emergence in both its aspects, quality and structure, came from Empedocles. He describes the formation of bone in these words: And gracious earth received in its broad-bosomed receptacles, out of the eight parts, the two of gleaming Nestis, and four of Hephaestus. And these parts came to be white bones, having been fitted together out of a divine marvel (WKHVSHVLƝWKHQ) through the cohesive bonds of harmonia. (B96)

The names of some of the elements are given – as is often the case in Empedocles – by quasi-mythological personifications: Nestis is water; Hephaestus, fire. The noun harmonia may refer either to the state of cohesive structure, or it may be an alternate designation of the agent force of attraction – in which case that noun, too, is a personification and should be capitalized. Reducing the language of epic poetry to the language of modern chemistry we have the equation: 2 Earth + 2 Water + 4 Fire

Love

Bone

It would be wrong to discount the adverb WKHVSHVLƝWKHQ as poetic decoration. Elsewhere Empedocles emphasizes that each of the elements has its own proper character (B17.28 ʌ੺ȡĮ į’ ਷șȠȢ ਦț੺ıIJ૳). But when bone comes to be, an

88

Alexander P. D. Mourelatos

ostensibly new ƝWKRV, the special texture and look of bone, comes into the world. The unusual suffix -then for the adverb, which conveys the ablatival sense “out of,” is aptly expressive. Empedocles’ own theory posits that bone comes to be “out of” the elements. But so far as our eyes can tell, bone comes to be “out of a divine miracle.” My suggestion is that the adverb WKHVSHVLƝWKHQ is intended to convey this prima facie mystery of the advent of an ostensibly emergent quality. It is not only in this fragment that we find Empedocles speaking of the generation of compounds as wondrous; he does so repeatedly and by using a variety of devices. In these other contexts the two aspects of quality-emergence and structure-emergence are often alluded to in conjunction. In B35 he speaks of the formation of complex living things as “a marvel to behold” (thauma idesthai). And in the two verses that end with the clausula just cited, he seems to distinguish between effects of the formation of quality and effects of the formation of structure by using two different expressions: Ȥİ૙IJ’ ਩șȞİĮ ȝȣȡ઀Į, “countless tribes were blended” (B35.16), and ʌĮȞIJȠ઀ĮȚȢ ੁį੼ૉıȚȞ ਕȡȘȡંIJĮ, “articulated in all manner of shapes” (B35.17), respectively. In B71 he acknowledges his listener’s incredulousness concerning the manner in which “mortal things” might be thought to arise out of the four elements (İੁ į੼ IJ઀ ıȠȚ ʌİȡ੿ IJ૵Ȟįİ ȜȚʌંȟȣȜȠȢ ਩ʌȜİIJȠ ʌ઀ıIJȚȢ, / ʌ૵Ȣ . . .). The two major aspects of the effect that strains the listener’s credulity are distinguished explicitly in line 3 of the fragment: the “shapes” or “forms” (HLGƝ) of things; and their color (chroia). In B23 he speaks of mortal things as “unspeakably wondrous in their source” (DVSHWDSƝJƝQ). It is worth dwelling on this fragment, which compares cosmogonic processes with certain effects in painting. To begin with, the translation of DVSHWDSƝJƝQ I just gave construes the two words together, taking SƝJƝQ not as the subject of einai (in what would be a stretched anacoluthon) but rather as an accusative of respect, and giving the phrase its very likely function of a hexameter clausula. Here are the relevant two lines and my translation: Ƞ੢IJȦ ȝ੾ ı’ ਕʌ੺IJȘ ijȡ੼ȞĮ țĮȚȞ઄IJȦ ਙȜȜȠșİȞ İੇȞĮȚ șȞȘIJ૵Ȟ ੖ııĮ Ȗİ įોȜĮ ȖİȖ੺țĮıȚȞ, ਙıʌİIJĮ ʌȘȖ੾Ȟ. Let not deception overcome you into thinking that those of mortal things that have come to be evident, unspeakably wondrous in their source, are from something else. (B23.9-11) 14

14

Translations of these two lines in standard works on the Presocratics or on Empedocles generally overlook this clausula-function of ਙıʌİIJĮ ʌȘȖ੾Ȟ. They construe ʌȘȖ੾Ȟ as subject of İੇȞĮȚ – taking ʌȘȖ੾Ȟ with șȞȘIJ૵Ȟ in a hyperbaton that breaks up the natural four-cola articulation of the hexameter line. This construction leads them to give ਙıʌİIJĮ the rather colorless rendering “countless.” Thus Wright 1981, p. 15: “. . . that there is any other source for the countless perishables that are seen.” This pattern of construction appears to have been established by DK; cf. Bollack 1969, II, pp. 36-37; Guthrie 1962-1982, II, p. 148; Barnes 1979a, II, p. 8; Kirk–Raven–Schofield 1983, p. 294.

4. Quality, Structure, Emergence

89

The PƝDOORWKHQ of this fragment corresponds to the WKHVSHVLƝWKHQ of B96. “Not from something else” means, of course, from the four elements through a variety of combinations. Empedocles invites us to view this effect as analogous to a painter’s depiction of a variety of objects through skillful mixing of a few basic paints: As when two painters are artfully producing votive tablets (ਕȞĮș੾ȝĮIJĮ ʌȠȚț઀ȜȜȦıȚȞ) . . . – first they select (ȝ੺ȡȥȦıȚ Ȥİȡı઀Ȟ) paints capable of producing many colors (ʌȠȜ઄ȤȡȠĮ ij੺ȡȝĮțĮ); then, after mixing more of some and less of others in due proportion (ਖȡȝȠȞ઀ૉ ȝİ઀ȟĮȞIJİ IJ੹ ȝ੻Ȟ ʌȜ੼Ȧ, ਙȜȜĮ į’ ਥȜ੺ııȦ), from these they produce forms (İ੅įİĮ) that resemble all manner of things – constructing trees, and men, and women, and beasts, and birds, and fish nurtured in water, and long-lived gods, highest in honor . . . (B23.1-8)

Once again two “marvels” are involved: the production of a variety of colors from the few basic ones; the production of filled-out shapes. But the simile draws our attention rather more insistently to the first of the marvels. Empedocles dwells on the act of mixing the paints and on the variety of ratios employed. 15 Note that the painters mentioned in the fragment are not producing an abstract composition; we are explicitly told that they are reproducing likenesses of objects. Since, with respect to shapes, the process is merely one of imitation, the marvel of shape-composition involves no novelty; it is no more than the marvel of the artist’s imitative skill. Indeed, the simile works better if we focus on the effects of color mixture than it does if we focus on composition of shapes. For there is no analogue in the cosmogonic process to the painter’s originals or prototypes – the real plants, etc. Unlike Plato’s Demiurge, Love, the cosmic craftsman in Empedocles’ system, does not copy pre-existing patterns. In the case of color mixture, on the other hand, there is not only the marvel of the artist’s skill; there is also the marvel of an emergent quality. How can red and yellow produce orange? red and black, brown? blue and yellow, green? 15

I cannot agree with Wright 1981, that “the mixing of colors of which Empedocles is speaking in the simile is almost certainly not a blending to produce further shades but the setting of paints of one color side by side with those of another in an arrangement to give the effect of a familiar object” (p. 38, cf. p. 180). This interpretation commits her to assuming that “the ij੺ȡȝĮțĮ are the appropriate colors ready before the painters start on their picture” (p. 180). But this gives a poor sense to ਖȡȝȠȞ઀ૉ ȝİ઀ȟĮȞIJİ and to the adjective ʌȠȜ઄ȤȡȠĮ. It also goes against the rather emphatic staging of the painter’s acts conveyed by the verb forms: (1) ਥʌİ઀ . . . ȝ੺ȡȥȦıȚ, (2) ȝİ઀ȟĮȞIJİ, (3) ʌȠȡı઄ȞȠȣıȚ. Vincent J. Bruno, who is one of the authorities she cites in support of the assumption that paint mixing (as distinct from the superposition of color washes) was rare before the Hellenistic period, actually contradicts her in his own discussion of Empedocles B23: see Bruno 1977, pp. 56-58; see also Keuls 1978, pp. 66-70. I shall myself, in Sections IX-X, defend Wright’s view that Empedocles cannot countenance in his system any case of real blending or fusion. But we must distinguish between Empedocles’ simple statement of ordinary perceptions concerning painters’ practice from his theoretical explanation of the phenomenon of intermediate shades obtained by mixing.

90

Alexander P. D. Mourelatos

One of the technical challenges of fifth-century painting is the production, by suitable mixing, of that “carnation” color, the andreikelon, the tint of human flesh. 16 Underlying this technical challenge – and similar challenges of color mixture – is the metaphysical marvel of the emergence of an astonishing variety of hues each of which has its own irreducibly unique quality, its own ƝWKRV. 17 Early twentieth-century emergentist philosophers had spoken of “natural piety” (a phrase Samuel Alexander borrowed from Wordsworth) as the appropriate and ultimate stance for a philosopher of this persuasion. Empedocles was perhaps enough of a mystic to have adopted precisely such an attitude of reverent acquiescence. At one point in my study of this subject, I had myslf reached the conclusion that Empedocles was the first emergentist. But for reasons I shall give in due course, I am now convinced that he, like any other Neo-Ionian, exercised the utmost of ingenuity and imagination toward dispelling the mystery involved in effects of emergence. As my own earlier quandary might suggest, Empedocles’ solution is by no means obvious. The evidence gives prima facie support to a hypothesis that Empedocles had recourse to several quite different mechanisms. I shall, accordingly, defer the work of analyzing and interpreting that evidence until we have a clear sense of the solutions offered by other Pluralists. 18 V. ANAXAGORAS’ STRONG DENIAL OF QUALITY EMERGENCE No other Presocratic philosopher, and probably no other philosopher in the entire history of philosophy, has been more determined and more drastic in explaining away effects of emergence than Anaxagoras. This comment applies with full force to his handling of effects of quality emergence. But we shall find evidence that, even in the case of effects of structure, Anaxagoras seeks to deny emergence by propounding a doctrine of preformation. 16

See Keuls 1978, p. 68. 7KHUHLVDQLQWULJXLQJSDUDOOHOIURP,QGLDQSKLORVRSK\7KHVFKRRORIWKH&ƗUYƗNDWKHKRPRQ\PRXV founder of which is supposed to have been a contemporary of Empedocles, offered the production of a particular shade of red color “from the combination [by chewing] of betel, areca nut, and lime” as a simile of the emergence of consciousness or intelligence from material elements. See Riepe 1961, p. 68. I am indebted for this reference to Ali Hossaini. 18 It is not crucial for my purposes to resolve the vexed issue whether Anaxagoras influenced Empedocles or vice versa. The definitive study is O’Brien 1968, which closes with the statement, “I conclude that Empedocles wrote later than Anaxagoras, and was influenced by him.” Cf. Kahn 1960, pp. 163-64; Schofield 1980, pp. 33-35; Sider 1981, pp. 1-11. These authors favor the translation “inferior” for the notoriously ambiguous ੢ıIJİȡȠȢ in Aristotle’s remark (Metaph. I.3.984a11), ਝȞĮȟĮȖંȡĮȢ į੻ ੒ ȀȜĮȗȠȝ੼ȞȚȠȢ IJૌ ȝ੻Ȟ ਲȜȚț઀઺ ʌȡંIJİȡȠȢ ੫Ȟ IJȠ઄IJȠȣ (scil., of Empedocles) IJȠ૙Ȣ į’ ਩ȡȖȠȚȢ ੢ıIJİȡȠȢ. But Jaap Mansfeld, at the end of an exhaustive analysis of the sources for Anaxagoras’ life, defends the “natural” translation: “Anaxagoras, though earlier in age than [Empedocles], was later in his works”: Mansfeld 1979-1980, esp. part II, pp. 90-91. Anaxagoras’ philosophical activity is also placed after Empedocles’ in Teodorsson 1982, pp. 70-71. 17

4. Quality, Structure, Emergence

91

The distinctive thesis of Anaxagoras’ metaphysics is the ontological principle of universal mixture (UM, for short): “In everything there is a portion of everything” (B6, B11, B12 ad init.). The “everything” of the principle excludes, as we are explicitly told in B12, nous or “mind,” which stands “alone itself by itself.” This ontological principle is supplemented by a principle of dominance and latency: “each thing is most manifestly those things of which there is most in it” (B12 ad fin.). As the term “most manifestly” (HQGƝORWDWD) suggests, this principle makes a transition from ontology to epistemology. Anaxagoras also posits several other principles, of which I only need to mention here one that is purely epistemological in import. I shall call it the principle of sensory limitations: “Because of the feebleness of our senses we are not able to discern the real” (B21). It will suffice if we give this epistemological principle the minimal interpretation of which it admits: Our senses are feeble because those minority portions of other things that are present in any one thing fall below our sensory threshold. Armed with this trio of principles, Anaxagoras undertakes to eliminate any putative instance of quality-emergence. Suppose we had a thing of character X, and out of it there came to be a thing of character Y. No emergence is involved here: the character Y was already actually present in the X thing, but it was a minority ingredient, and so its presence was not felt. As it separated out of the X thing, character Y became the majority ingredient in the post-separation extract, and has thus bestowed its appearance upon the whole to which it now belongs. Correspondingly, if things of character V and of character W combine to form a blend of character X, this is because each of the V and W things contained enough of character X as a minority ingredient so that when the V and W things combined, character X became the majority ingredient in the new compound. The full gamut of natural phenomena, from cosmogony and meteorology to nutrition and growth, may thus be explained in terms of the intuitively clear process of locomotion and by appeal to the fundamental insight that the more prevails over the less. 19 Needless to say, UM requires some refinements in its formulation. With nearly all recent interpreters, I assume that Anaxagoras did not hold a particulate theory of matter. The mixture of any two ingredients in Anaxagoras involves total reciprocal permeation. The model is inevitably implicated in our explication of the sense in which the hot, the white, and the liquid are compatibly and pervasively present in an object that displays these qualities (say, a cup of warm milk). UM, accordingly, is not to be interpreted as the thesis that a given lump of X stuff contains distinct particles of every other kind of stuff; it means, rather, that X contains portions of every other kind of stuff in uniform or homogeneous (or, to use the more precise ancient term, homoeomerous) mixture. I have, in effect, already introduced this refinement by distinguishing 19

Cf. Furley 1976.

92

Alexander P. D. Mourelatos

between “things of character X” and “the character X.” This is essentially the distinction between c-substances (substances proper) and e-substances (character-importing ingredients) introduced by Colin Strang in a classical artlcle. 20 UM would, in effect, be given this formulation: In everything there is a portion not only of its own manifest ingredient(s) but also of each of the ingredients manifest in each of the rest of things. So, in bone there is not only that which makes bone bony but also latent portions of what makes salt salty, milk milky, water watery, the mist misty, the hot hot, the heavy heavy – and so forth. What I have just given is the traditional and more widely held interpretation, in accordance with which the pronoun “everything” in UM has the widest possible scope. There is, however, an alternative interpretation, advocated in its crucial respects by Malcolm Schofield, which gives UM a narrower scope, viz.: In everything there is a portion of each of the opposites. By “opposite” we are to understand any of the familiar reified qualities and powers of Presocratic philosophy – e.g., the rare or the dense, the dry or the wet, the hot or the cold, the bright or the dark, and so forth. 21 This controversy concerning the scope of “everything” in UM has important repercussions on our theme of quality emergence. According to the narrow scope interpretation, substances and stuffs other than the opposites must be conceived of, in Schofield’s words, as “logical constructions out of opposites.” 22 Schofield explains: “water, for example, is in essence nothing but stuff in which the dense, the wet, the cold and the murky predominate – to a greater extent than they do in clouds, but less than in earth or stones.” 23 This, in effect, attributes to Anaxagoras a two-stage strategy for eliminating quality emergence. First, as with similar programs in the tradition of modem empiricism, there would be a purely logical reduction of complexes to their constituent features. Second, if, in the course of the formation of a given complex, any one of its features should seem to have come into being ex nihilo, the expunging of that effect is given by the principle of dominance and latency. Schofield is quite right in observing that this narrow-scope interpretation has the attractiveness of the perennial view that “the external world presents itself directly and fundamentally only in the form of sensible qualities.” 24 20

Strang 1963. Schofield 1980, correspondingly, distinguishes between “substance stuff and ingredient stuff” (p. 112, cf. pp. 108-09). That ingredients could not exist in the form of elementary particles follows from an additional Anaxagorean principle of infinite divisibility. Strang’s distinction (and its equivalent in other authors) applies to portions that recede to infinity in smallness (B6, cf. Strang 1963, p. 366; Schofield 1980, p. 109). 21 The alternative interpretation goes back to Paul Tannery and John Burnet and has been defended, among others by Gregory Vlastos. For references, see Schofield 1980 or Sider 1981 or Teodorsson 1982, the last of whom gives an extended account of the different interpretations of the scope of “everything” that have been proposed (pp. 25-64). 22 Schofield 1980, p. 116. 23 Ibid. 24 Ibid.

4. Quality, Structure, Emergence

93

There is, however, one phenomenon that poses difficulties for the narrow scope interpretation. How are we to explain changes that result in the production of characters that are intermediate between opposites? When the hot and the cold mix, how does the tepid or the lukewarm come into being? Obviously something more than “logical construction” is involved. For we cannot define the tepid as that which has simultaneously the (pure) characters of the hot and the cold. 25 What underlying mechanism for the production of intermediate effects might Anaxagoras have envisaged? By considering how the narrowscope interpretation can handle this phenomenon of intermediates we stand to gain deeper insight into Anaxagoras’ handling of effects of quality-emergence generally. The phenomenon is one that Anaxagoras did definitely discuss. A scholium on a passage of Gregory of Nazianzus contains what may or may not be (in its exact wording) a fragment of Anaxagoras: “How, says [Anaxagoras], could hair come to be from what is not hair and flesh from what is not flesh?” (B10). 26 The scholiast then goes on to observe that Anaxagoras extends his doctrine of latent inherence beyond “bodies” (VǀPDWǀQ): “For there is presence, in the white, of the black; and in the black of the white. And he posits the same in the case of tendencies of motion (UKRSǀQ), holding that the light [i.e, the ‘lightweight’] is mixed with the heavy; and vice versa” (Gregor. XXXVI 911 Migne). A report from Sextus not only bears out this testimony but even suggests that Anaxagoras had done some experimenting with the generation of intermediate shades from the mixing of black and white. This Sextus passage which is the one from which we get fragment B21, is worth full translation and comment: Anaxagoras, the natural philosopher par excellence (SK\VLNǀWDWRV), denigrating the senses as infirm, says: “we are unable to discern the real (krinein WR  DOƝWKHV) because of their feebleness” (B21). As proof of their untrustworthiness he adduces gradual color shifts. If we should take two colors, black and white, and then pour some of the one into the other drop by drop, our eyes will be unable to discern the small changes, even though they do really occur (țĮ઀ʌİȡ ʌȡઁȢ IJ੽Ȟ ij઄ıȚȞ ਫ਼ʌȠțİȚȝ੼ȞĮȢ). (Adv. math. VII.90)

In accordance with a theory widely held by Greek philosophers – both contemporaries of Anaxagoras and philosophers of the classical and Hellenistic periods – all colors (not just shades of grey) can be produced from the combination of black and white. Some of the perplexity we are bound to feel in 25

26

It would seem that Schofield 1980 overlooks this difficulty. For he speaks of Anaxagorean manifest objects interchangably as “epiphenomena” and “logical constructions out of opposites” (p. 116 and passim). Schofield 1980 is inclined to doubt that B10 is a fragment (see pp. 133-44); Sider 1981 concludes that “in the absence of evidence to the contrary, it deserves to keep its place as an Anaxagorean fragment” (p. 90).

94

Alexander P. D. Mourelatos

considering such a theory is mitigated by evidence that leukos, which we usually translate “white,” also means “bright” or “clear” or “shiny,” and melas, which we translate “black,” also means “very dark” (of any markedly darkshade color). 27 Sextus either had evidence that Anaxagoras subscribed to this theory, or simply assumed he did. It is, in any event, quite believable that Anaxagoras did subscribe to the theory, and that the experiment he adduced was one whereby a bright color can be continuously changed into each of a wide variety of color shades by gradual admixture of a darkening pigment. The experiment shows that our eyes fail to detect the continuous change; we perceive the emergence of a new color only after certain quantum jumps in the addition of black – e. g., at first lingering white, then suddenly yellow, then after a while red, and so forth. But what is the underlying Anaxagorean mechanism of change? If we opt for the wide-scope interpretation of the pronoun “everything” in UM, then the natural answer to our question would seem to be that each of all possible color shades is latent in any color, and, as drops of the darkening pigment are added, new majority ingredients are formed, which then become manifest. If we opt for the narrow-scope interpretation of “everything,” we have, in principle, four additional choices. Intermediates will count either as illusory, or as phenomenal only, or as fully and properly emergent, or as resulting from some sort of interactive process of tempering. We today have, to be sure, some difficulty imagining that out of black and white we could generate some color other than grey, in some shade or other of the latter. So, it would be better to return to the intuitively more obvious example of the generation of the tepid or the lukewarm from the hot and the cold. I believe the following set of five mechanisms covers the full range of relevant possibilities: (a) There is no such thing as the tepid; this is just a blur of distinct stretches or bits in which the hot and the cold alternately dominate. A sharpened tactile sensitivity would show the tepid to be a mere illusion. (b) There is no such thing as the tepid; this is merely a state (pathos) of our sense of touch in the presence of distinct stretches or bits in which the hot and the cold alternately dominate. The tepid is not an illusion, but it is an appearance that has no correspondence to reality. (c) The tepid is a genuine, irreducible emergent. (d) The tepid arises from a process whereby the hot and the cold become reciprocally tempered. This would imply an interaction of the two opposites involving qualitative alteration. (e) There is enough of the tepid in each of the hot and the cold so that, when the latter two mix, the tepid becomes dominant. 27

See, s. vv., Liddell–Scott 1940; henceforth referred to as LSJ. Cf. Bruno 1977, pp. 84-85 and 91-93.

4. Quality, Structure, Emergence

95

One of these possibilities we can dispose of immediately. Mechanism (c) seems to go against the general thrust of Anaxagoras’ anti-emergentist metaphysics. The remaining four mechanisms require lengthier examination. Mechanism (d) is recognizably the solution to the problem of mixis offered by Aristotle in I.10 of De generatione et corruptione. 28 In the formulation I have given it above, it certainly seems out of place in the fifth century. Moreover, it would be open to the same objection raised against (c): interactive qualitative alteration is surely the sort of coming-to-be that Anaxagoras is bent on ruling out. But perhaps there is a way to reformulate (d) so that it neither begs peculiarly Aristotelian concepts nor openly defies Anaxagoras’ own ban on coming-to-be. For is it not true that fifth-century thinkers – Alcmaeon, Heraclitus, Philolaus and other Pythagoreans, the Hippocratics – spoke of a tempering of opposites in krasis, “blending” or “fusion,” and of harmonia, “fitting together”? 29 Might not Anaxagoras, too, have thought that ostensible emergence can nonetheless be mitigated by some principle of reconciliation of opposites? And yet this very comparison with other fifth-century views serves to discredit (d) as a relevant option for Anaxagoras. A striking fact about Anaxagoras’ vocabulary is that neither words of the har- root (harmonia, harmozein) nor words of the kera-root (krasis, NHUDLǀ, kerannymi) have any occurrences in the fragments. The contrast, especially with Empedocles and Philolaus, is quite telling. It would seem that Anaxagoras is avoiding this genre of words precisely because they import the connotation of a reciprocal adjustment or reconciliation of the ingredients. 30 His central conception was significantly different: not conflict and reconciliation; rather, latency and manifestation. In any event, in order for us to take mechanism (d) seriously, we would need to posit an Anaxagorean principle of tempered union that supplements or modifies the principle of dominance. And it would be gratuitous to posit this merely because of our interest in accounting for the effect of intermediates. Mechanism (b) is ruled out by the general consideration that we have no evidence that Anaxagoras postulated any sort of a “veil of appearances”; nothing that comes, as it were, to be superimposed on the real in the process of perception. To the contrary, the famous fragment B21a, ੕ȥȚȢ IJ૵Ȟ ਕį੾ȜȦȞ IJ੹ ijĮȚȞંȝİȞĮ, is best understood as stating that “phenomena afford us a glimpse into what lies unseen.” There is, for Anaxagoras, no appearances/reality split; 28

See Mourelatos 1984a, esp. pp. 8-14. See Schwabe 1980, pp. 15-39. 30 Anaxagoras uses words of the mig- root (ȝ઀ȖȞȣȝȚ, ȝ઀ıȖȦ, ȝ઀ȟȚȢ), which, in general usage, differ from words of the kera- root, roughly, as “scramble” and “mingle” differ from “blend” and “fuse,” as conturbatio differs from temperatio. See Schwabe 1980, pp. 24-25, 31-33. This pattern of usages does not, of course, amount to evidence that Anaxagoras has a particulate theory of matter; it merely serves to emphasize that the ingredients preserve their identity unaltered in the mixture. 29

96

Alexander P. D. Mourelatos

even though phenomena are a small part (a specimen or limited sample) of reality, they are reality nonetheless. Mechanism (a) gives an odd twist to the principle of sensory limitations: not only do we fail to perceive the minority portions in things; even majority portions of a given opposite can fail to make themselves manifest if present in close juxtaposition with majority portions of the corresponding opposite. Given that twist, mechanism (a) can be made compatible with fragment B20. But is it also compatible with fragment B21a (which I quoted immediately above)? Granted the distinction between an illusion and an appearance, it would nevertheless seem that mechanism (a) contradicts fragment B21a no less forcefully than mechanism (b) does. We are not speaking here of some isolated and rare effect; the fact is that most of the qualities we perceive, through any of our senses, are intermediates. It would be hard to imagine that such an effect would not fall within the scope of ta phainomena. These abstract considerations can now be reinforced with evidence drawn from the Sextus passage on the mixing of colors. Note the clause: “though these [changes] do really occur.” It is only (c), (d), and (e) that could possibly be counted as mechanisms of real change. Sextus could not have said what he says if he had evidence that Anaxagoras favored mechanism (a) or mechanism (b). With (c) a non-starter and (d) quite unlikely, the only possbility remaining is (e). The latter is, as we have seen, the only mechanism postulated by the broad scope interpretation of UM. If, now, the narrow-scope interpretation is explicated along the lines suggested by (e), the distinction between the two lines of interpretation seems less sharp. Within the scope of the pronoun “everything” in the statement of UM, the narrow scope interpretation would, by virtue of mechanism (e), allow not just the polar opposites but also the infinity of intermediate grades that fall between the two extremes. Thus, any advantage of explanatory economy, which the narrow scope interpretation might offer over the broad scope interpretation, is largely forfeited. There remains the difference that the narrow-scope interpretation provides, at a first stage, a logical reduction of manifest complex objects to manifest opposites, including manifest intermediate grades of opposites, 31 before it gives full play to UM and to the principles of dominance/latency and sensory limitations, at a second stage. In any event, Anaxagoras comes across on either interpretation as the determined foe of quality-emergence. The things that come to show themselves are in every case and in every respect identical with things that had previously lurked unperceived; and the things that are not perceived are identical with things that eventually come to be perceived. We are now in position to appreci31

Schofield 1980 does acknowledge this addition: “to the . . . view that the external world presents itself directly and fundamentally only in the form of sensible qualities . . . [Anaxagoras] adds the idea that sensible qualities always fall within quality ranges which are specified in terms of opposites” (p. 116).

4. Quality, Structure, Emergence

97

ate the pointedly realist and anti-emergentist message of fragment B21a: ੕ȥȚȢ IJ૵Ȟ ਕį੾ȜȦȞ IJ੹ ijĮȚȞંȝİȞĮ, “Phenomena afford us a glimpse into what lies unseen.” VI. THE PREFORMATION OF STRUCTURE IN ANAXAGORAS As we turn to consider Anaxagoras’ handling of effects of the emergence of structure, it may well seem that we should next investigate Anaxagoras’ nous, “mind,” inasmuch as the latter is explicitly said to “structure” things. The results of this approach will disappoint us – as they disappointed Plato. But a different and more viable approach will fortunately be forthcoming. In compensation, as it were, for the initial disappointment, I propose the alternative of exploring a fascinating connection between Anaxagoras’ doctrine of nous and his doctrine, which we have just finished studying, of qualitative change. The connection has often been missed. Jonathan Barnes writes that the doctrine of nous “is readily separated from Anaxagoras’ physical theories,” 32 a remark that expresses the opinion generally held. What brings the connection into view is the focus I have sought to sustain on the Neo-Ionian demand for intelligibility and on the problems posed for that demand by effects of quality emergence. In a famous passage in the Phaedo (97B ff.), Plato criticizes Anaxagoras for introducing nous but then giving it the trivially mechanical role of a motive force. Modern interpreters, taking their cue from Plato, have commented that the action of agitation and separation imparted by Anaxagoras’ nous is no different than that of Empedocles’ personified Hate or Strife. But can this be right? Could it be that Anaxagoras would have introduced this strikingly novel conception without any intent to exploit systematically the attributes and themes of awareness, knowledge, rationality, and intelligibility that are standardly associated with nous? I think it is far more likely to suppose that there was such a connection, even if we thus incur the debt of explaining how or why Plato should have suppressed it in the Phaedo passage. The relevant passage in Anaxagoras is the middle of fragment B12: And nous came to recognize (HJQǀ) all the things that were mingling together (symmisgomena), and the things that were separating off (apokrinomena), and the things that were being distinguished (diakrinomena). And of what sort things were meant to be, and of what sort they were, what now is not, and what now is, and of what sort they will be – nous structured (GLHNRVPƝVH) all these things . . .

I shall discuss shortly the second part of this statement, the one that speaks of the role of nous in structuring things. For the moment I want to concentrate on the first part. My suggestion is that Anaxagoras is here stating his anti32

Barnes 1979a, II, p. 16.

98

Alexander P. D. Mourelatos

emergentist doctrine in epistemological terms. No qualitative change will over surprise nous. No character, or power, or stuff is a supervenient novelty. Rather it is something nous already conceives of as a latent ingredient something that nous comes to “recognize” when conditions are right for the ingredient to bestow its proper character upon the whole of which the ingredient becomes an augmented part. Nor is there anything arcane about these conditions: the insight that the more prevails over the less is simple and compelling. Schofield has argued very convincingly that when Anaxagoras speaks of nous we must not necessarily assume that he is referring to a unique supreme Mind. 33 There are, to be sure, contexts in which Anaxagoras may be referring in the first instance to a supreme mind; but, prima facie or by default, occurrences of the term nous are quite naturally interpreted as references to mind in general. The nous that “recognizes” all the ingredients and all the states of mixture is not only the cosmic nous; it is also your nous and mine. The choice of nous over some personified force of agitation is well motivated after all. It proclaims that the universe holds no mysteries impermeable to reason. If all qualitative change is intelligible, it is fitting that the agent that initiates change should itself be intelligence. It would seem, then, that Plato’s account of Anaxagorean nous in the Phaedo is, to a certain extent, misleading or tendentious. Still, on a crucial point Plato was justified in his critique. For in the context in which the critique develops, Plato’s focus was not on the aspect of quality emergence but on the aspect of structure emergence. You will recall that fragment B12, after speaking of the insight that nous has into mixing and unmixing, goes on to speak of nous’ work of “structuring” (diHNRVPƝVH) things past, present, and future. Part of what is meant here has been anticipated in my explication of the immediately preceding passages: nous physically produces the mixings and unmixings that it understands. The point made is a realist inversion, so to speak, of the conceptualist point made in a famous sentence from the second preface of Kant’s first Critique. Kant wrote: “Reason has insight only into that which it produces after a plan of its own.” 34 Anaxagoras might be paraphrased as saying: “Nous produces, after a plan of its own, only that into which it has insight.” But clearly the language of structuring in B12 brings in structural as well as qualitative change; and every change that involves transposition, shaping, reshaping constitutes an instance of emergence of new structure. Is there anything in Anaxagoras’ scheme that could make intelligible how nous could impart to the mixing-unmixing ingredients the form of a plant, or of an animal, or of a human being? What Anaxagoras needs are principles of latency or preformation of structure to supplement his principles of latency of quality. But 33 34

Schofield 1980, pp. 12-22. “. . . die Vernunft nur das einsieht, was sie selbst nach ihrem Entwurfe hervorbringt” (BXIII).

4. Quality, Structure, Emergence

99

if we concentrate on the role of nous, no such principle is forthcoming. On that approach, Plato’s complaint is perfectly understandable. What nous would need, but Anaxagoras does not supply, is something like a repository of eternally existing patterns that nous might envisage and imitate as it goes about its business of structuring things. Of this conception there is certainly not the slightest hint in Anaxagoras. But perhaps Plato’s disappointment, and ours, comes from looking in the wrong quarter. For we have some intriguing hints of a different, un-Platonic, principle of the latency of structure. In B4 Anaxagoras speaks of “seeds of all things (spHUPDWD SDQWǀQ FKUƝPDWǀQ),” and goes on to say that these seeds “have both all kinds of patterns (ideas) and all kinds of colors and flavors.” The juxtaposition in this remark of the two aspects, structure and quality, is worthy of note. In another segment of the same fragment we get a similarly noteworthy pairing the set of qualitative ingredients (“the opposites and earth”); and, “seeds infinite in multitude and in no respect like one another.” What exactly are these seeds, each of which is unique? One line of interpretation is that the seeds are simply discrete particles of pure (unmixed) stuff; another, that they are distinct aggregates of fully mixed stuff. According to yet another line, the seeds are growth points (as in the seeding of crystals) for the processes of separation and accretion that produce the manifest stuffs of our world. 35 But the opinion is increasingly taking hold, in literature, that the key to understanding Anaxagoras’ seeds is in the obvious and standard biological sense of the term “seed.” Thus Furley proposes that the term refers, quite straightforwardly, to “the seeds of all plants and animals,” and goes on to make this comment: “if the no-coming-to-be rule is applied strictly, there are no emergent characteristics; all the present differences – between species, subspecies, varieties, perhaps even individuals – must have been latent in the original mixture.” 36 Schofield gives a similar interpretation, pointing out that “cosmic rotation . . . is plainly inadequate to account for the generation of plants and animals,” given their obvious complexity of structure and individuality, and also pointing out that “there is a known mechanism whereby plants and animals are produced, viz. the propagation of seeds.” 37 Teodorsson resists narrowing the reference strictly to the biological realm, and argues that the “seeds of all things” are “prefigurations, or matrices, of individual perceptible things, inorganic as well as organic, with one sperma for each individual thing of the universe.” 38 Accordingly, not only will there not be hair from what is not hair, but also no man from what is not man, no cubical crystal from what is not a cube, and even no house from what is not a house. In other characterizations of the seeds, Teodorsson also uses the modern cyber35

For review of the literature, see Teodorsson 1982, pp. 45-64. Furley 1976, p. 72. 37 Schofield 1980, p. 124. 38 Teodorsson 1982, p. 82. 36

100

Alexander P. D. Mourelatos

netic metaphors of “governing devices, already programmed in detail” and, more simply, of “programs”; 39 but – as he would doubtless recognize – in the ancient context, a generalized conception of “seed” would necessarily have to be based on the biological model. (The perfect modern analogue for the Anaxagorean doctrine of seeds would seem to be in the evolutionary doctrine of emboitement, “encapsulation,” propounded by the eighteenth-century biologist Charles Bonnet.) 40 This newest conception of the Anaxagorean seeds has the advantage that it mediates yet another link between the doctrine of nous and the physical doctrine. Indeed, the link mediated by Teodorsson’s conception is even stronger than the one for which I argued above. A personified force of agitation is not something that intrinsically has knowledge. A nous is. And by speaking of a nous that has complete foreknowledge of future events, Anaxagoras is, in effect, stating in epistemological terms what the doctrine of seeds implies in ontological terms, viz., that the unfolding of all changes – and this preeminently includes changes of structure – is prefigured and predetermined in the seeds. It is important to appreciate that, on this conception of the seeds, the role of nous can be quite plausibly limited to agitation and foreknowledge. The verb GLHNRVPƝVH, “structured in detail,” of B12 need not refer to the sort of shaping and molding performed by the Platonic Demiurge; it can express nous’s activity of stirring and dispersal (cf. SHULFKǀUƝVLV) that provides the occasion for the seeds to germinate. Thus, Plato’s complaint that the nous has no teleologically intelligible plan is robbed of its force. 41 Attractive as this view is, it is not without difficulties. The principle of preformation of structure implied in the doctrine of the seeds would have to count as an independent additional principle in Anaxagoras’ system. Indeed, it must somehow be completely unaffected by the application of UM. In the state of precosmic mixture, the seeds have not germinated, and their individual features lie completely suppressed under the blanketing presence of the predominant ingredients, which are “mist and shiny air” (DƝU WH NDL DLWKƝU, B1). But we cannot say, even though UM, taken broadly, would seem to require it, that in every part and region of the precosmic mixture there are specimens of every 39

Teodorsson 1982, p. 85. See Goudge 1973, p. 178. 41 Toodorsson 1982 fails, curiously, to exploit this obvious advantage of his own view. He supposes that, prior to the initiation of SHULFKǀUƝVLV, “nous had disposed of the coming development . . . had programmed the emergence of every individual thing” (p. 85). He bases this on the assumption that the aorists ਩ȖȞȦ and įȚİțȩıȝȘıİ must refer to activities that precede the initiation of motion. But clearly these two aorists are parts of a series of aorists that articulates the entire second half of B12 (ਥțȡȐIJȘıİȞ. . . ਵȡȟĮIJȠ. . . ਩ȖȞȦ. . . įȚİțȩıȝȘıİ . . . ਥʌȠȓȘıİȞ), and in the absence in our text of any lexical, grammatical, or syntactical devices of phasing, all these aorists must be taken as coincident (in the grammatical sense). If, in the midst of this series of aorists, Anaxagoras had meant to convey that the knowing and the disposing had preceded the rotation, the appropriate tense would have been the perfect or the pluperfect (cf. pluperfects in Teodorsson’s own statement). 40

4. Quality, Structure, Emergence

101

type of seed; for we are explicitly told that no two seeds are alike (B4). Outside the primordial mixture, a given seed may have been realized in an indefinite plurality of structured bodies. But even then we cannot say of every part and region of the developed world that there is something of every type of structured body in it. We might perhaps say of a spherical triangle that it contains something of the circular as well as of the triangular; but this cannot be generalized at all plausibly. There is not the tiniest bit of the circular in a flat triangle, nor vice versa. Moreover, unlike stuffs, structures are not homoeomerous. The parts of a man are not smaller men, but such things as hands, feet, a face, etc. A face, in turn, is not made up of smaller faces, but of eyes, nose, mouth, etc. A corollary to this is that structures do not grow in the way stuffs grow: a man does not grow by the addition of man, whereas a bone grows by the addition of bone. 42 Nor is the principle of dominance applicable to structures. For in order for us to say that a feature F is dominant in a thing x, we should also be able to say that x is more F than y is. But though we might perhaps say that the flat triangle is more of a triangle than the spherical triangle, we cannot say that something is “more man,” or “more finger.” But as we pursue this line of exploration, we realize that these are not difficulties that would have to be answered by this particular interpretation of Anaxagoras. Rather, they are consequences, that merely need to be acknowledged, of attributing to his system any degree of recognition of the distinction between the aspect of quality and the aspect of structure. And it is hardly believable that a philosopher who discussed the gradual “separating out” of an elaborately structured cosmos from a shapeless and nearly featureless nebula of “mist and shiny air” should have failed to recognize that distinction. VII. TWO PREFIGURATIONIST SOLUTIONS: PHILOLAUS AND DEMOCRITUS It would now be relevant for me to investigate in detail two other types of Pluralist solution to the problem posed by effects of emergence: those of Philolaus of Croton, and of the two Atomist philosophers, Leucippus and Democritus of Abdera. 43 The Pythagorean Philolaus posits two sorts of entities as fundamental in his universe: “unlimited things,” which are probably stuffs; and “limiting things.” The latter are probably shapes and numbers, conceived of not as abstract entities but as more or less corporeal molds or modules. Both sorts of fundamental entities are unborn and imperishable and give rise to the whole cosmos and all things in it through a process of coalescence. Involved in such coalescence may 42 43

Cf. Furley 1976, p. 81. This Section and Sections VIII-XI of the present Chapter supplement and develop what (with the exception of much shorter treatment of Philolaus here) is presented compendiously in the final four Sections of the preceding Chapter 3.

102

Alexander P. D. Mourelatos

be either unlimited things only, or limiting things only, or one unlimited thing and one limiting thing. There is even provision for a second-level coalescence of things that are already compounded of an unlimited thing and a limiting thing. In those cases in which like things coalesce (bone to bone, moisture to moisture, a cube stacked on a cube) more juxtaposition suffices. But if the coalescing things are different, then, to quote B6, “it is impossible for them to be arrayed together unless harmony should supervene, in whatever manner it comes to do so (ਕįȪȞĮIJȠȞ. . . țȠıȝȘșોȞĮȚ, İੁ ȝ੽ ਖȡȝȠȞȓĮ ਥʌİȖȑȞİIJȠ નIJȚȞȚઅȞ ਚįİ IJȡȩʌ૳ ਥȖȑȞİIJȠ).” The sort of harmony Philolaus envisages is suggested by his interests in harmonics, the theory of mean ratios, and number theory. Especially remarkable about the sentence just quoted is the characterization of harmony as “supervenient” and the note of amazement in the “whatever manner” phrase. The supervenient status of harmony receives special emphasis in the text, because of a rhetorical contrast with the “underlying and preexisting principles” (ਕȡȤĮ੿ ਫ਼ʌ઼ȡȤȠȞ). The note of amazement corresponds, no doubt, to the philosophical wonderment at effects of the emergence of quality or structure variously expressed by Empedocles. Also striking is the parallelism with Anaxagoras. The “unlimited things” correspond to Anaxagoras’ stuffs; the “limiting things” to Anaxagoras’ seeds. But the two philosophers differ sharply in their handling of effects of emergence. Whereas Anaxagoras seeks to explain away emergence by positing latency and preformation, Philolaus is content to render emergence intelligible by keeping it within the constraints of certain mathematical or quasi-mathematical laws of harmony. The advent of compounds and configurations are, to be sure, the genesis of something new in the universe. But if their formation is in accordance with a mathematically intelligible procedure – e.g., the derivation of a quartet of entities from a trio and a sextet, in accordance with the rule of the harmonic mean – the product is appropriately tempered, and the effect of emergence is mitigated. My supposition is that Philolaus was greatly impressed by the generative character of mathematics: in a proportion, the extremes converge on the mean in accordance with an intelligible rule; in a mathematical series, the ordinally later elements can be thought of as anticipated by the earlier ones; in a geometrical construction, the outcome is prefigured in the properties of the figure from which we start; in a mathematical proof, it is intuitively easy to grasp the conclusion as prefigured in the premises. Like other Pythagoreans, Philolaus seems lax in observing a distinction between things and numbers. So, it was a simple matter for him to project onto the realm of physical change generative rules which, to us, are strictly logicomathematical in nature. To distinguish between the Anaxagorean and Philolaic solutions to the problem of emergence I shall from this point on cease employing the terms “preformation” and “prefiguration” interchangably. I shall reserve the former term for the Anaxagorean conception alone.

4. Quality, Structure, Emergence

103

As I summarize the relevant doctrines of the two fifth-century atomists, I shall use the name of Democritus, who is far better represented in our sources, to refer to both of the Abderites. The system of Democritus is classified rather poorly by the term “materialism.” Democritus gives considerable prominence to the aspect of structure, form, configuration. It is no accident that the term he uses for his atoms is the same term which Plato uses for his transcendent Forms: ideai. Atoms are, to be sure, units of stuff. But this aspect of stuff amounts to nothing more than their absolute rigidity and absolute hardness – it is simply their defining difference from the void. What is far more interesting is the sense in which atoms constitute units of structure, both in the sense of shape and in the sense of configuration. The atoms have an infinity of shapes – which inevitably entails that all possible solid figures, and thus all possible flat shapes and all possible lines and curved edges, are represented at the atomic level. Moreover, the pattern in which two atoms will behave when they come in contact is determined solely by two factors: the motions they have acquired after collisions with other atoms; and their shapes. But those motions previously acquired are in turn determined in part by the shapes of the atoms involved in the collisions. Thus, the pattern of motion of a given atom is, in the long run, a function of its shape. Some shapes bestow greater motility, others tend to make atoms sticky and sluggish; some shapes would promote a smoothly rectilinear motion, others a tumbling or gyrating motion – and so forth. Shape, one might say, is an atom’s destiny. That Democritus’ system is mechanistic and deterministic is a familar point. But we should also note that the system fulfills requirements of prefiguration. The generative rules are not those of number theory or harmonics, as they seem to be in Philolaus. But they are, broadly speaking, mathematical: they involve the domains of arithmetic, geometry, and an extremely simple and highly idealized kinematics. The prefigurationist nature of the system seems all the more striking as we ponder what was said immediately above, about an atom’s shape constituting its destiny. If we take a diachronic approach, the state of a group of atoms at time t is generatively prefigured in the states of these and certain other atoms at a time t - d. If we take a synchronic approach and confine our attention to the structural aspect, manifest objects are nothing but geometric summations of the atoms and the gaps that separate them as the atoms are variously spread out and disposed vis-à-vis one another in the void. (On the prefigurationist aspect of Democritean metaphysics, see also Mourelatos 1984b – in Greek.) Modern students of Democritus have focused a little too impatiently on the failure of his system to account for biological structures and for the so-called “secondary qualities.” What is remarkable to me is how very far the system succeeds in reducing manifest qualitative features to underlying atomic structures. A texture that is perceived as hard, stiff, tight, and as resisting penetration, corresponds to a structure of closely packed atoms. One that is

104

Alexander P. D. Mourelatos

perceived as soft, loose, and yielding corresponds to an expanse mostly of void, with atoms moving freely through it. The entire spectrum from hard to soft corresponds to varying degrees of atomic congestion. Indeed, once we combine the factor of congestion with factors of geometric shape, the compass of reduction becomes truly immense. We can give intelligible accounts of all the possible grades involved in such contrasts as fine/coarse, smooth/rough, brittle/hard, flexible/rigid, permeable/impermeable, elastic/inelastic, runny/viscous – and so on through the vast domain of mechanical qualities and dispositions. Even in the case of some of the secondary qualities, notably flavors and such visual contrasts as bright/dark and transparent/opaque, Democritus manages to establish at least a partial correspondence between the manifest qualities and the underlying atomic structures. And this is a correspondence we would find intelligible – or, at any rate, intuitively compelling. Even more remarkably, Democritus acknowledges that there will inevitably be some unreduced residues – “raw feels,” as some of today’s philosophers of mind would call them. And so, in a famous fragment he propounds a version of the “disappearance view” (or “eliminative materialism”): 44 “sweet exists by convention, bitter by convention, in truth only atoms and the void” (B9 and B125). VIII. EMPEDOCLES AN EMERGENTIST? Let us now return to Empedocles. His cosmology posits the familiar four elements and two motive forces, Love and Strife. Wherever there is mixture, or cohesive structure, or the formation of compounds, Love is at work. What we call “birth” is just this process of aggregation. The work of Strife is to separate things and break them apart. What we call “death” is just this process of dispersal. With the passage of aeons of time, Love succeeds in bringing things together to a state of total union in the form of a perfectly spherical divine creature, the Sphairos. Soon thereafter – perhaps the very next moment – Strife reasserts its cosmic rights by shattering this unity. Love again resumes her work of gradually putting the severed elements together, against the constant, albeit inevitably diminishing, opposition and interference of Strife. The cycle is repeated. It is compounds – including the gods of traditional belief and the Sphairos – that are “born,” that “perish,” and that “change.” The elements and the two forces are eternal and unchanging in their character. In Section IV of the present Chapter, Empedocles was portrayed as in the grip of the marvel of emergence. Might he have been the first emergentist philosopher, after all? Might he have held that compounds (both structures and 44

For a clear statement of that position and of the philosophical issues that frame it, see Churchland 1984, pp. 43-49.

4. Quality, Structure, Emergence

105

smoothly homogeneous qualitative blends) are not reducible, logically and geometrically, to their elements, and should therefore count in his ontology as real, albeit supervenient, entities? The suggestion is by no means far-fetched. Evidence in support of it can be drawn from Empedocles’ handling of the language of genesis and from consideration of the ontological status of the Sphairos. True to his principles, Empedocles does not ever use gignesthai, “to come to be,” in an absolute construction (which would yield the sense “to be born”) with the elements or forces as subject; but he does use it in predicative constructions of the form, “the elements came to be F.” 45 Actual qualitative alteration of the elements is out of the question, of course. Though they become “at certain times different,” they “continue to be always the same” (ਙȜȜȠIJİ ਙȜȜĮ țĮ੿ ਱Ȟİț੻Ȣ Įੁ੻Ȟ ੒ȝȠ૙Į, B17.35). To speak of a transformation of the elements is to use the language of nomos, “convention” (B9.5). But, as Solmsen and Furley have pointed out, 46 Empedocles’ forbearance of linguistic convention is in itself significant. Anaxagoras said that the Greeks were wrong to speak of coming-to-be and of perishing, and, consistent with this ruling, he appears to have avoided the censured terms in statements that expound his own doctrine. Empedocles may or may not have made a similar remark – it all depends on how we reconstruct the corrupt first half of B9.5. 47 But it is striking that he uses freely both gignesthai and its synonyms – not just in polemical contexts but even in statements that express his own views concerning cosmological and biological processes. This would seem to suggest that effects and products of genesis have some ontological status of their own, that they are no mere ghosts of linguistic convention. Moreover, there would seem to be at least one compound in Empedocles’ system that has so superlative a degree of unity as to make a reduction to underlying pluralities seem quite implausible. That compound is, of course, the Sphairos, Empedocles’ epochal equivalent of the Eleatic One. If Empedocles’ position were such as to have implied that the sharp-eyed Lynceus of Greek legend could make out the grain of each of the four elements in the Sphairos, then Empedocles’ celebration of the unity of his quasi-Eleatic One would appear either insincere or misplaced. At least the Sphairos, in its periodic epiphanies, should count as an emergent. But if the Sphairos has this ontological 45

See the passages referred to in the relevant entry of Index Verborum in Wright 1981. As Solmsen 1975 (esp. pp. 126 ff. and 131 ff.) has shown, the gignontai of B17.11 has WKQƝWRL as subject. Moreover, even though Solmsen resists the corresponding construction for B26.10, it seems to me that the analysis he offers of the context of B17.11 dictates that the subject in B26.10 is not the elements but the DQWKUǀSRLand WKƝUHVof B26.4. 46 Solmsen 1975, pp. 125 f., 132f.; Furley 1976, pp. 62-64. 47 The attested readings are ਲ șȑȝȚȢ țĮȜȑȠȣıȚ and İੇȞĮȚ țĮȜȑȠȣıȚ, both of which are unmetrical and hardly intelligible in context. The emendation that has been most widely adopted is rather tendentious; for it assimilates Empedocles’ sentiment to that of Anaxagoras by inserting a negative particle: ઞ șȑȝȚȢ țĮȜȑȠȣıȚ.

106

Alexander P. D. Mourelatos

status, why might there not also be some isolated patches of homoeomerous brown color here and there that could claim, in their own small way, the status of emergents? All the above considerations notwithstanding, decisive evidence against an emergentist interpretation can be found in two Empedoclean remarks of major import. In fragment B17, which is the central statement of Empedocles’ ontology of four elements and two forces, we read at line 30: țĮ੿ ʌȡઁȢ IJȠ૙Ȣ Ƞ੡IJૃ ਙȡૃIJȚ ਥʌȚȖȓȖȞİIJĮȚ Ƞ੡IJૃ ਕʌȠȜȒȖİȚ, “and neither does anything come to be over and above these [the four elements and the two forces] nor does it [= any part of them] fail to be.” 48 Doubtless, epigignetai does not have the technical sense of the modern philosophical noun “epigenesis” or of the modern philosophical verb “supervenes.” But, in the context in which it appears, the statement has unqualified scope. If a product of mixture is to count as a genuinely new third entity, the appropriate way to describe the situation would surely be the precise contradictory of B17.30: “something did come to be over and above the ingredients, and something of the original ingredients was lost.” Indeed, Empedocles betrays some concern that he might be misinterpreted as introducing either entities that “come to be over and above” the four (or six) he officially acknowledges or entities that amount to “decaying” or “denatured” versions of the same four (or six). In a statement that recurs formulaically at crucial junctures of his exposition of metaphysical fundamentals, he reinforces the remark of B17.30 with its logical equivalent: auta gar esti tauta, “these things alone [the four and the two] exist” (B21.13, B26.3, and in the slight variant all(a) aut(a) esti tauta at B17.34). The actual observations that suggested an emergentist reading should not, however, be ignored. We shall, in due course, look for ways to do justice to them along different lines. What we should now consider is the evidence that Empedocles sought to dispell the mystery of emergence by joining in the NeoIonian enterprise of seeking relevant antecedents and intelligible transformations. That evidence is quite ample. Indeed, at first blush it would appear that – either by way of anticipation or by way of imitation, and so in a mood of eclecticism – he availed himself of several different mechanisms, each corresponding to one of the major Neo-Ionian systems I have discussed in the preceding Sections. In the Section that follows I study the conceptual options available to Empedocles for solving problems of quality emergence.

48

Wright 1981 is being excessively cautious in placing ਙȡૃ IJȚ between daggers: see pp. 97-98 and 17172. Though there may be some doubt about the relevant parallels, the metrical liberty at issue (elision of the iota, or synaloephe with the epsilon of ਥʌȚȖȓȖȞİIJĮȚ) is mild, and transparent enough.

4. Quality, Structure, Emergence

107

IX. THE MECHANISM OF QUALITATIVE MIXTURE IN EMPEDOCLES To begin with, there are hints in Empedocles of the Anaxagorean mechanism of dominance and latency. Testimony from Aristotle attributes to Empedocles the view that water, which is in itself tasteless, “holds within it the varieties of all flavors, which are imperceptible because of their minuteness” (De sensu 4.441a3 = DK31A94); and there is related testimony from other sources. 49 It is also a priori unlikely that Empedocles would have resisted the supposition that the four elements lend something of their respective characters to every compound in which they are present in sufficient quantity. Indeed, he specifically tells us in B21.6 that things composed of earth have some of the manifest characteristics of earth itself. 50 It is also likely that Empedocles assumes that there is some fire in wine, some air in translucent marble. In the formula for bone at B96, the large portion of fire is doubtless intended to capture the “white” color that is the shared epithet of bones and of the sun (B96.3 ostea leuka, B21.3 ƝHOLRQPHQOHXNRQKRUDQ). It may also be intended to capture the charcoal-like brittleness of that form of tissue. The darker, flexible, and fleshy sinews (neura, a global term for tendons, blood vessels, and nerves), by contrast, have a high content of water (A78). There are also more than a few hints of a mechanism of harmonia, “cohesive structure,” the sort of mechanism that reconciles unlikes in Philolaus. “From these [from the four elements] have all things been molded, by being cohesively fitted together (SHSƝJDVLQ KDUPRVWKHQWD),” we read in B107 (cf. B122.2, B96.4). The mathematical implications of harmonia are reinforced by the specification of distinct logoi, “ratios,” that make each compound of the elements possible. The ratio for bone, it will be recalled, is 2 Earth + 2 Water + 4 Fire. In the formula for sinews, the quantities specified for Water and Fire are transposed: 2 Earth + 2 Fire + 4 Water. A mixture of almost exactly equal measures of the four elements constitutes blood; and somewhat larger departures from the norm of perfect equality constitute the various “other forms of flesh” (B98). 51 Absolutely perfect equality in the measures of the four elements is found only in the tissue that constitutes the Sphairos. 52 Occasionally, Empedocles goes even beyond harmonia to suggest reciprocal assimilation (cf. B22.5 DOOƝORLVKRPRLǀWKHQWD). And it is worth repeating the observation that

49

See Beare 1906, pp. 161-62. For discussion of this fragment, see below p. 131 and n. 12. Also relevant is B76.3, which speaks of the evidence of the earth component on the shells of turtles and mollusks. 51 For excellent comment on this fragment, see Solmsen 1950, pp. 435-41 [= Solmsen 1968, I, pp. 50208]. 52 It is also likely that the transmigrating GDLPǀQ of Empedocles’ mystical poem, “Purifications,” is a superlatively stable compound of equal portions of the four elements, a surviving part of the Sphairos. See Wright 1981, pp. 69-76. 50

108

Alexander P. D. Mourelatos

Empedocles differs strikingly from Anaxagoras in making use of the vocabulary of krasis, “tempered blending.” 53 Pointing in an entirely different direction are hints of a mechanism that foreshadows the Democritean reduction of manifest qualities to underlying structures. Unlike Anaxagorean mixtures, which involve total interpermeation, it would appear that Empedoclean mixtures are made possible by a mechanism of distinct poroi or “passages” of various shapes and sizes (or gauges). Empedocles deployed the theory of poroi to account for effects not only of mixture, but of growth, respiration, perception, translucency, transparency, and magnetism. The evidence is mainly in Aristotle De generatione et corruptione I.8 and in Theophrastus De Sensibus 7-24, 54 but the theme of poroi is heard discernibly in the fragments, too. A convincing analysis of this conception of “passages” has been offered by C. J. F. Williams: “[The poroi] might be interstices between horizontal layers of a substance, and in this case, if the interstices between the layers of substance A exactly fitted the layers of substance B, it would be possible for A and B to interpenetrate.” 55 We need not suppose that the poroi are rigid. The theory probably envisages wide variation from fairly supple to quite stiff. The conception of mixture as some sort of articulated structure is implied in several of the fragments. Water, we are told in B91, “lends itself more to articulation (mallon enarthmion) with wine, but refuses to do so with oil.” In other words, there is something about the inherent grain or texture of wine that promotes the admixture of water, whereas there is something about the grain or texture of oil that prevents the admixture of water. Since wine and oil are presumably compound stuffs, we cannot, from this fragment alone, draw the inference that the mixture of the elements themselves is effected by “passages.” But that the elements taken severally have a characteristic grain is suggested by B22: “all of these [the elements] are in articulation (arthmia) with their own parts.” Significantly, that fragment draws a parallel between the manner in which (KǀV G¶ DXWǀV, B22.4) one of the elements articulates itself with its own proper parts and the manner in which a mixture (NUƝVLQ, B22.4) is produced from diverse ingredients – which could be either different elements or different stuffs. There is, in any event, ample testimony in Aristotle, in Theophrastus, and in other sources 56 to support the assumption that at least three of the four elements have characteristic “passages.” The exception may have been air, which, given that Empedocles is a plenum theorist, could well have been given the role of keeping the passages full (and thus distended) when they are not 53

See Wright 1981, pp. 332-33 (Index verborum, s. vv. țİȡȐȞȞȣȞĮȚ, țȡ઼ıȚȢ). For these and other relevant testimonia, see DK31A86-87, A89, A92. Cf. Bollack 1969, II, pp. 13657, 256-61; and corresponding commentary in III, pp. 136-57 and 256-61. 55 Williams 1982, p. 125, ad I.8.324b25. 56 See above, note 54. 54

4. Quality, Structure, Emergence

109

otherwise penetrated. 57 Though much of the relevant testimony is concerned with Empedocles’ recourse to poroi in explaining the process of perception, it does support the more general assumption that the mixture of the elements themselves is made possible by certain patterns of symmetria, “reciprocity” or “congruence” or “meshing,” between the poroi of different elements. That Empedocles should have availed himself of the Anaxagorean mechanism of dominance/latency is quite understandable. Once outside the context of Anaxagoras’ theory, and deprived of the extraordinary implications that UM confers upon it, the principle – that the majority component suppresses the features of any number of other components that may be present in a mixture – is a commonplace of Greek cosmology from the fifth century forward. 58 Clearly this particular mechanism can supplement a variety of different cosmological schemes. But the two other mechanisms are different enough so that, had Empedocles used both of them, he could not escape the indictment of being rather carelessly eclectic. We need to ask two questions: Which of these mechanisms has a firmer footing on the fundamentals of Empedocles’ cosmology? Which of these can more easily be adjusted or reinterpreted so as to make it compatible with the other? Theophrastus appears to imply that Empedocles gave “numerical specifications of the mixtures that characterize” a variety of compounds. 59 Modern scholars suspect that Empedocles did this only for the biological tissues. 60 What is crucial, in any event, for purposes of the present discussion is that we do not have the slightest evidence that Empedocles worked out any sort of mathematical (or even numerological) theory to explain why some logoi yield one sort of substance and others another. In fact, it is arguable that Empedocles was not at all inclined to develop a theory of the Philolaic type. For he seems to allow for potentially infinite variance or deviation from the standard or prima facie ratios he cites in whole-number formulae: “either a little more, or, where there is more, a little less” (B83.4, cf. B23.4). 61 This has the implication that some of the ratios will be incommensurable – hardly a Pythagorean approach.

57

See Philoponus excerpt in DK31A87. Owen Goldin has pointed out to me that air, too, may have poroi: according to Gener. et corr. I.8.324b29-30 (DK31A87), the transparency of air is explained by the allignment of its poroi. There is, in fact, no requirement that air should have no poroi; the poroi of air could, in turn, be filled by microscopically thin filaments of the other elements, in which case these other elements would be said to have much wider poroi. Aristotle’s objection, that if all things had poroi then things would consist of nothing but gaps (325b7), is captious. 58 In Democritus, too, and then in the Epicureans, we find the principle in both an ontological and an epistemological version: natural substances contain numerous minority components the properties of which are suppressed by those of majority components; from the infinity of emanations from atomic structures, our senses pick out only those sets of similar emanations that have an unusually high number of members. 59 Cf. IJ੹ į੻ ਙȜȜĮ ਫ਼Ȗȡ੹ țĮ੿ ʌİȡ੿ ੖ıȦȞ țĮIJĮȡȚșȝİ૙IJĮȚ IJ੹Ȣ ੁį઀ĮȢ țȡ੺ıİȚȢ, De Sensibus 12. 60 See Solmsen 1950, p. 437 [= Solmsen 1968, p. 504]; cf. Wright 1981, pp. 209-10. 61 For the text and translation of B83.4, see Wright 1981, pp. 123 and 238.

110

Alexander P. D. Mourelatos

But perhaps it is unreasonable even to entertain the hypothesis that Empedocles should have taken the same approach Philolaus took. Might we not find support for a more relaxed hypothesis, viz., that Empedocles sought at least a qualitative analogue to the Philolaic model of harmonious convergence? Might he not have exploited the simple conceptual logic of contrariety? Thus, for example, with reference to the problem of the emergence of the tepid, his solution might have been the one Anaxagoras did not (and, as we saw, properly could not) exploit: We intuitively understand the tepid to be something between the hot and the cold; so when the tepid is produced from the mixture, not only is the result anticipated and appropriate, we are also in a position to argue that the hot and the cold have survived in this intermediate product. This, after all, may be the harmonia involved in those cases of smooth homogeneous mixture that are aptly described with the adjective ]ǀURV (B35.15) and with words of the kera- root, viz., kerannymi (B35.8, 35.15), kirnamai (B71.3), NUƝVLV(B21.14, 22.4, 22.7), akratos (B128.6, 128.8). I know of no evidence in the fragments or the testimonia that would contradict this alternative hypothesis. Nonetheless, from the standpoint of interpretation, this hypothesis is inherently unsatisfactory. For it entails a reading of Empedocles that is less attractive than would be otherwise possible. The evidence for a theory of poroi is strong; there can be little doubt that Empedocles had recourse to that particular theory. So, the most that can be at issue is whether Empedocles deployed only the theory of poroi, or both that theory and, by way of supplement, the model of qualitative convergence. Interestingly, the theory of poroi is able to accommodate any evidence suggesting that Empedocles had discussed cases of qualitative convergence. It simply takes such cases as phenomenal effects, as explananda rather than as explanantia. The reverse is not true. There is no point in explaining effects of congruence of poroi in terms of a model of underlying qualitative convergence. Indeed, there is no point in postulating poroi if one has a model of qualitative convergence. For one who has the latter model, a theory of poroi becomes otiose. 62 The ingenuity of the Empedoclean theory of poroi and its centrality within Empedocles’ metaphysical scheme has not been adequately appreciated by modern interpreters. So, let us take a moment to explore the theory’s full explanatory import. To begin with, the assumption of poroi protects Empedocles’ system from an obvious charge of incoherence. Pairs of the Empedoclean elements have opposite features. In making this simple observation we are not, of course, foisting upon Empedocles the qualitative contrarieties of Aristotle’s scheme of the four simple bodies of the sublunary realm. Surely, Empedoclean fire is hot, and the three other elements cold; water is dark, but fire and air are 62

And that is why Aristotle, who does have the qualitative model, and unhistorically assumes that Empedocles had it too, objects that “it is altogether otiose (KROǀV SHULHJURQ) to postulate poroi” (326b21; cf. comment of Williams 1982 ad loc., p. 124).

4. Quality, Structure, Emergence

111

bright; earth sticky, but the other three runny; earth and water heavy, but the other two light – and so forth. Let us recall here that what makes it possible for Anaxagorean opposites to combine directly is the plethora of latent intermediates each contains. Now Empedocles does not make Anaxagoras’ assumption of universal mixture. If Empedoclean elements combined directly, without the benefit of mediation provided by the poroi, it is hard to see how he could escape the charge that his compounds, especially those biological tissues that come closest to the 1-1-1-1 formula, are congeries of contradictory features. Empedocles’ system would, in effect, be indistinguishable in essentials from Parmenides’ cosmology of combinations of Light and Night – the system Parmenides expounds with some diffidence in his Doxa. Empedocles corrects Parmenides by having the elements mix not directly but through the mediation of certain channels, the poroi. Moreover, the theory of poroi makes it possible for Empedocles to determine and to explain, by a sort of mechanical calculus, salient facts concerning events of mixture, viz., the ease or the difficulty of mixing; the character of the outcome; the stability of the outcome. Here is what that calculus might look like: (a) If the poroi of substances A and B have the sort of smooth texture that would permit both easy interlacing and easy extrication, then mixing will be quite easy. The product C of this type of mixture will have a homogeneous character; but C will also be relatively unstable. (b) If the poroi of substances D and E are ridged or notched, so that interlacing is relatively easy but extrication difficult, then mixing will again be relatively easy. Product F will again be of homogeneous character; and it will be quite stable. (c) If the poroi of G and H are such as to allow interlacing only under certain favorable conditions of approach (e.g., if a structure of dovetail joints is involved), then mixing will be laborious. The product J will, nonetheless, be of homogeneous character; and it will be fairly stable. (d) If the poroi of K and L are such as not to allow any interlacing, then no mixing wlll take place. The stirring of K and L together will produce a confused appearance.

Naturally, Empedocles is also in a position to exploit a wide array of possible combinations of (d) with each of (a) or (b) or (c); and additional explanatory scope is gained by taking into account the degree of rigidity or suppleness of the poroi. Two substances both of which have overly supple poroi will obviously have difficulty mixing. I would now like to cite, from the body of fragments and testimonia, an additional selection of features that gain significance when they are framed by the theory of poroi. I shall draw on some of the same ancient texts that have

112

Alexander P. D. Mourelatos

persuaded one interpreter of Empedocles, Rosemary Wright, that Empedoclean mixture is, in all cases, a mechanical juxtaposition or assembly, not a real fusion. 63 Love’s work of mixture is often described in pointedly mechanical terms. In her identity of Aphrodite she “expertly joins” (DVNƝVDVD) things with “pegs” or “dowels” or “bolts” (gomphois, B87). The corresponding sockets are, almost certainly, the poroi, “passages” of our testimonia. The cognate verb JRPSKRǀ, “to fasten with bolts,” is used at B33 with reference to the action of juice in curdling milk. That passage is especially instructive because it is a transparent imitation of Iliad 5.902 ff. What we have in Homer is a simile between the swiftness with which juice makes milk solidify and the swiftness with which a certain treatment of a wound made the blood coagulate. 64 We have only the first limb of Empedocles’ but it is clear that the focus of his comparison is no longer on the swiftness of coagulation but rather on the “fastening” and “binding” that is involved: ੪Ȣ įૃ ੖IJૃ ੑʌઁȢ Ȗ੺ȜĮ ȜİȣțઁȞ ਥȖંȝijȦıİȞ țĮ੿ ਪįȘıİ . . ., “as when juice bolts tight and binds white milk . . .” 65 In the curdling of milk, Empedocles seems to be saying, we have visual and palpable evidence that even in the mixing of liquids there is an underlying process of fitting of bolt and socket. The process is not obvious, of course, in the mixing of wine and water. But the term enarthmios, which Empedocles uses to describe the readiness with which water combines with wine in a seemingly homogeneous blend, conveys the sense of the meshing of discrete, articulated parts. Love is not only a carpenter; she is also pictured as a potter, kneading, modeling, and baking the clay (B73, B75, B95). In its underlying texture, clay is no more a homogeneous blend than tempered wine is. An obvious analogue to the potter’s kneading of clay is the kneading of barley meal and water, which is described thus at B34: ਙȜijȚIJȠȞ੢įĮIJȚțȠȜȜ੾ıĮȢ,QWKLVIUDJPHQWNROODǀ surely cannot mean “to glue,” or “to cement,” since no third bonding material is involved. In Homer, NROOƝHLV and NROOƝWRV are used of structures that derive their strength from close overlapping; and NROOƝWRV is specifically an epithet of chariots – which are hardly “glued.” 66 Empedocles B34, accordingly, should be translated “having fastened barley meal tightly on to water.” This translation is borne out by the plural kollai at B96.4, which rules out mass-noun translations, such as “glue” or “cement”; the reference must be to mechanical “joins,” even to the tenon and mortise structure of dovetailing joins.

63

Wright 1981, pp. 34-40. ੪Ȣ įૃ ੖IJૃ ੑʌઁȢ Ȗ੺ȜĮ ȜİȣțઁȞ ਥʌİȚȖંȝİȞȠȞ ıȣȞ੼ʌȘȟİȞ / ਫ਼ȖȡઁȞ ਥંȞ, ȝ੺ȜĮ įૃ ੯țĮ ʌİȡȚIJȡ੼ijİIJĮȚ țȣțંȦȞIJȚ, / ੬Ȣ ਙȡĮ țĮȡʌĮȜ઀ȝȦȢ ੁ੾ıĮIJȠ șȠ૨ȡȠȞ ૓ǹȡȘĮ. (Iliad 5.902-904) 65 Bollack is certainly right in his observation that the context of this fragment in Plutarch’s De amicorum multitudine gives us no clue of the missing second limb of the simile: Bollack 1969, II, p. 133, and III, pp. 310-12. 66 See LSJ s. vv. The verb NROODǀ does not occur in Homer. 64

4. Quality, Structure, Emergence

113

B96 is, of course, the fragment concerning that wondrous mixture which results in the production of bone. The whole fragment is worth a second look: ਲ į੻ ȤșઅȞਥʌ઀ȘȡȠȢਥȞİ੝ıIJ੼ȡȞȠȚȢȤȠ੺ȞȠȚıȚ IJઅ į઄ȦIJ૵ȞੑțIJઅ ȝİȡ੼ȦȞȜ੺Ȥİȃ੾ıIJȚįȠȢĮ੅ȖȜȘȢ IJ੼ııĮȡĮįૃ ૽ǾijĮ઀ıIJȠȚȠǜIJ੹ įૃ ੑıIJ੼ĮȜİȣț੹ Ȗ੼ȞȠȞIJȠ ਖȡȝȠȞ઀ȘȢțંȜȜૉıȚȞਕȡȘȡંIJĮșİıʌİı઀ȘșİȞ And gracious earth received in its broad-bosomed choanes, out of the eight parts, the two of gleaming Nestis, and four of Hephaestus. And these parts came to be white bones, having been fitted together out of a divine marvel (WKHVSHVLƝWKHQ) through the cohesive bonds of harmonia. (B96)

We are now in position to do justice to a detail that was passed over in my earlier discussion of the fragment. The curious phrase in the second half of the first line, earlier translated “in her broad-bosomed receptacles,” has strong sexual coloring, to be sure. But the word choanes has the literal meaning “funnels.” Empedocles, who not only is aware of the semantic phenomenon of ambiguity but also exploits ambiguity for expressive purposes, 67 is playing off the transparent sexual imagery against a no less transparent allusion to his mechanical theory of poroi. A similar play is involved in B98, the fragment which correspondingly gives the formula of the composition of “blood and other forms of flesh”: ਲ į੻ ȤșઅȞIJȠ઄IJȠȚıȚȞ੅ıȘıȣȞ੼țȣȡıİȝ੺ȜȚıIJĮ ૽ǾijĮ઀ıIJ૳ IJૃ ੕ȝȕȡ૳ IJİțĮ੿ Įੁș੼ȡȚʌĮȝijĮȞંȦȞIJȚ Ȁ઄ʌȡȚįȠȢ੒ȡȝȚıșİ૙ıĮIJİȜİ઀ȠȚȢਥȞȜȚȝ੼ȞİııȚȞ And earth joins up with these in parts as close as possible to being equal, with Hephaestus [=fire] and rain [=water] and all-revealing air, as she takes refuge into the perfect harbors of Cypris [Love] . . . (B98)

The adjective teleios is expressively polysemous. It evokes in the first instance the “ultimate” protection a ship finds in the enclosed bay of a harbor; it also evokes the “fulfillment” of sexual intercourse; and it certainly conveys the “perfection” of the 1-1-1-1 ratio. But the theme of ingress and enclosure implied in the two images of the ship and of sexual intercourse mediate a further allusion, to the fitting of earth stuff into the poroi of the other elements. B96 and B98 show how the intuitive appeal of the mechanical theory is reinforced by the theory’s obvious psychological association with forms of sexual intimacy that involve penetration – witness the naturalness of our own idiom of “male” and “female” elements in mechanical assembly.

67

See Mansfeld 1972, esp. 24-35. Cf. Wright 1981, p. 188.

114

Alexander P. D. Mourelatos

One other detail of B98 serves to emphasize that even in that most thorough mixing that results in the production of blood, the four elements do not achieve a fusion: the verb used for the process of mixing is synekyrsen. Normally this would have the sense of “happened to meet with,” which is not admissible here, since the production of blood is almost certainly no accident but results from Love’s single-minded and precisely measured work. The only relevant sense of the verb, then, is purely that of “encounter,” the mutual approach and joining of entities that preserve their distinct identity. Nor is this encounter the first or tentative approach in a process that should ultimately lead to a fusion; the phrase LVƝ V\QHN\UVH PDOLVWD expresses the final state of mixture that is achieved. Remarkably, even the merging of separated parts of the same liquid is described not as a smooth fusion but as the successive overlaying of part on part: “so did the sweet lay hold of the sweet (glyky marpte), and bitter sprang upon bitter (epi pikron orousen), and sour climbed on top of sour (HS¶R[\HEƝ) . . .” (B90.1-2). 68 The centrality of the theory of poroi also helps us understand a certain theme in the testimonia that has often been regarded – undeservingly – with suspicion, on the grounds that it must have been inspired by a later assimilation of Empedocles’ system to that of the atomlsts. Aëtius reports that the four elements of Empedocles were constituted of “smaller bodies” (HN PLNURWHUǀQ RQNǀQ), or of “minima (elachista), and, as it were, elements of the elements” (A43). Under another heading, Aëtius reports that Empedocles posits “before (pro) the four elements, certain smallest smithereens (thrausmata elachista), homoeomerous (KRPRLRPHUƝ) elements earlier (prin) than the elements, as it were” (A43). There is nothing inherently unbelievable in these reports. The language of “minima” is used loosely; “smaller” and “tiny” is clearly what is meant. The characterization of them as “homoeomerous” shows that they are not pictured as having a firm shape; they are not atoms. It is quite plausible to suppose that these “smaller bodies” are tiny strands and filaments that are the beginnings of the texture of poroi. Here is the scenario I would propose: the “smithereens” are what results from the shattering of the limbs of the Sphairos by Strife (B31). Love immediately exploits the presence of these to form a mesh of strands and filaments; the very cracks of dismemberment that Strife’s violence had inserted are transformed in the hands of Love into “passages” (poroi) to promote reunification. Once enough of the smithereens have been reassembled into the various textures of poroi, mortal compounds of the elements become possible, and Empedocles’ evolutionary scheme begins to unfold. In the testimonia we also find explicit corroboration of the thesis I have been promoting, that all mixtures in Empedocles, including ostensibly homo68

The rest of the second line is corrupt. Reference to another flavor (the salty?) is likely. See Bollack 1969, II, pp. 198-201; III, pp. 462-64; cf. Bollack 1965, pp. 246-49.

4. Quality, Structure, Emergence

115

geneous ones, are, fundamentally, mechanical aggregations. Aristotle says this in De gener. et corr. II.7, likening an Empedoclean aggregation to a wall made of bricks or stones (A43); and Galen makes a similar point in two different passages of In Hippocratis De natura hominis (A34). The simile Galen introduces is that of a composite powder, made from mixing the filings of four different metals. 69 Galen contrasts Empedocles with Hippocrates: both had a four-element theory; but whereas Hippocrates’ elements can fuse into genuine blends, Empedoclean elements produce compounds “by being juxtaposed in small particles and making contact” (țĮIJ੹ ıȝȚțȡ੹ ȝȩȡȚĮʌĮȡĮțİȚȝȑȞȦȞIJİțĮ੿ ȥĮȣȩȞIJȦȞ, A34; repeated in A43). X. A CHEMISTRY AND PHYSICS OF ILLUSION We now need to take in the full implications for Empedocles’ ontology of the line of interpretation I have pursued in the previous Section. If the qualitative mixing of stuffs results from an underlying meshing of discrete components, what are we to say of the appearances of smooth homogeneity in the case of bone, or blood, or wine, or the color brown? We have three options. We may postulate that Empedocles was prepared to take the bold approach of eliminative materialism that was later taken by Democritus. Or we may assign to Empedocles the position of an epistemological dualist; i.e., he may have been inclined to a reality-appearances doctrine, positing some metaphysical factor that somehow veils or compromises the real. Or we may assume that he had in mind a doctrine of illusions. That Empedocles should have espoused a position akin to that of eliminative materialism is not out of the question. In B8 we read that what men call “coming to be” and “perishing” are, in reality and respectively, a mixing together and a coming apart; and In B9 Empedocles states his willingness to continue employing the vocabulary sanctioned by “convention” (QRPǀL) in this respect. What he does not do, however, is to take the crucial next step – the one that was taken by Democritus – to declare that the compound objects of the manifest world that result from this mixing and separating are themselves, either in all or in a fair number of instances, entities of convention. 70 There is

69

The simile almost certainly does not go back to Empedocles, who would have preferred one that conveys the idea of reciprocation in the grain of the components. 70 One might conceivably make a case for such an eliminativist reading of Empedocles by arguing that physis at B8 should be given not the sense “coming-to-be” but the usual sense of “nature” or “essence.” If Empedocles had said that compounds do not have a “nature” of their own, he would have said, in effect, that they are products of nomos. But the rhetoric of the fragment makes this interpretation quite unlikely: physis is paired with mixis, a pairing that corresponds to that of WHOHXWƝ with diallaxis (B8.1-3); a contrast of “nature” vs. “perishing” would be weak and not easily intelligible. Cf. Wright 1981, p. 175. It is also worth noting that the pair of personifications 3K\VǀWH 3KWKLPHQƝWHat B123.1 probably refers to “Birth and Death” (so Wright 1981, p. 281 f.); but even

116

Alexander P. D. Mourelatos

better textual support for the other two options. The recurring formula of the exclusive reality of the four elements appears, with a distinct and notable “but”-clause appended as sequel, in B26.3-4: “These things alone exist; but running through one another they come to be [or ‘there come to be’] men and the tribes of other animals.” In the recurrence of the formula at B17.34-35, the sequel reads: “. . . they come to be at other times different, and yet continue to be (ƝQHNHV) at all times the same.” Then, in its third occurrence, at B21.13-14, the conventional language of “coming-to-be” that had been employed on the two other occasions is given a remarkably phenomenological gloss: gignetai DOORLǀSD, “they come to be of a different appearance.” 71 The phenomenological theme is heard also in B23.10, where the “marvel” of ostensible emergence is specifically referred to “those of mortal things that have come to be manifest, XQVSHDNDEO\ ZRQGURXV LQ WKHLU VRXUFH´ șȞȘIJ૵Ȟ ੖ıĮ Ȗ੻ įોȜĮ ȖİȖȐțĮıȚȞ ਙıʌİIJĮʌȘȖȒȞ B23.10). 72 Remarkably, there is a discrepancy between underlying reality and surface appearance even in the case of the Sphairos, within whom “not even the swiftly moving limbs of the sun are discernible (dieidetai)” (B27.1). The reference here is, of course, not to the sun as such but to elemental fire; and the implication is that the other elements cannot be made out either. 73 Empedocles goes on to speak of the Sphairos as “held fast in the state of thick concealment brought about by cohesive structure” (ਖȡȝȠȞȓȘȢ ʌȣțȚȞ૶ țȡȣij૶ ਥıIJȒȡȚțIJĮȚ %  DQ HORTXHQW H[SUHVVLRQ RI WKH LGHD WKDW Love’s work of combination suppresses the underlying diversity. And at least one more fUDJPHQW PD\ EHORQJ WR WKLV SKHQRPHQRORJLFDO WKHPH % ȝȓĮ ȖȓȖȞİIJĮȚਕȝijȠIJȑȡȦȞ੕ȥ³RIWKHWZRWKHUHFRPHVWREHEXWDVLQJOHDSSHDUDQFH [or ‘vision’].” This fragment has been preserved with no indication of its context, and is generally thought to refer to the phenomenon of stereoscopic vision. But it is just as likely that it refers to effects of the reduction of two distinct features into one through mixing – e.g., from red and black, brown. the translation “Growth and Decay” supports the dynamic sense of “growing into being” for physis in B8. 71 The rest of the line is, unfortunately corrupt. The two attested readings are IJંȖȠȞ įȚ੺țȡȚıȚȢ ਕȝİ઀ȕİȚ and IJંȖȠȞ įȚ੺țȡĮıȚȢ ਕȝİ઀ȕİȚ. See Wright 1981, p. 179. The most widely adopted emendation is by Diels: IJંıȠȞ įȚ੹ țȡોıȚȢ ਕȝİ઀ȕİȚ, “to that extent does mixing change (them).” But Diels’ emendation has three unattractive features: asyndeton; no corresponding relative for IJંıȠȞ; shaky syntax for įȚĮȝİ઀ȕİȚ. 72 For the construction, see above at note 14, and the corresponding passage in the main text. 73 B27 is a conflation of two fragments. Bollack and Wright have restored the distinction that is clearly supported in our sources. DK31B27.1 and 27.3-4 = Bollack fr. 92 = Wright fr. 21 comes from Simplicius, refers to the Sphairos, and is the fragment I am quoting here. Bollack fr. 171 = Wright fr. 19 consists of a variant of DK31B27.1, as its first line, and DK31B27.2, as its second. This fragment comes from Plutarch. It is assigned by him not to the Sphairos but to a condition of precosmic chaos brought about by Strife. The Plutarch fragment speaks not of the “swift limbs” of the sun but of its “splendid form,” the reference being quite obviously to the heavenly body. It is the Plutarch fragment that alludes to the non-evident status also of earth and sea, conceived here not as elements but as cosmic masses. See Bollack 1969, II, pp. 41 and 71, and III, pp. 134-38 and 185-86; cf. Wright 1981, pp. 103-04 and 185-88.

4. Quality, Structure, Emergence

117

The phenomenological theme supports either a dualist (appearance/reality) or an illusionist reading of Empedocles. It is surely the latter that makes better sense for Empedocles. Had he been a dualist, would he not have found occasion to sound the standard dualist complaint, that the senses often deceive us? Indeed, what a strange type of dualist he should have made – one who would have accepted the intrinsic sensible characteristics of the four elements as real, but branded those characteristics that emerge under conditions of mixture as appearances. Most seriously, a dualist position would bring back with a vengeance all those perplexities of quality emergence that the theory of poroi had sought to dispel. What would it serve us to be told that brown arises through a process whereby the poroi characteristic of red things mesh with those characteristic of black things? How would we explain the fact that this particular meshing constitutes brown, whereas the meshing involved in the mixing of red and yellow constitutes orange? This sort of emergence would indeed have to count as irreducible. For it would involve precisely that category jump, in this instance from structure to quality, that has been the hallmark of emergence in modern philosophical discussions. Illusionism, by contrast, provides just the right supplement to the theory of poroi. Earlier, in pondering what I called Empedocles’ “mechanical calculus” of mixing, I took note of the contrast between the confused appearance of the quasi-mixture of oil and water, in which the streaks of the two ingredients remain visible, and the homogeneous appearance of tempered wine. The mechanical difference between the two cases is real enough. But so far as the visual effect is concerned, the two cases are different only in the circumstances of viewing. That confused appearance of the quasi-mixture of oil and water is an effect that obtains only for a viewer at mid range. At close range the viewer sees clearly enough the distinct streaks of the two liquids. Seen from a distance, even that quasi-mixture presents a smooth and homogeneous appearance. The move Empedocles needs to make, if he is to complete his project of dispelling effects of quality-emergence, is obvious: All those intermediate colors that seem to emerge when elemental colors are combined are actually distant views of a microscopic patchwork of streaks of the elemental colors. At closer range, those intermediates, too, would assume the sort of blurred appearance characteristic of stirred oil-and-water. At microscopic range, to the eyes of the mythical Lynceus (one of the Argonauts who was endowed with superlatively sharp vision), the individual streaks and Love’s intricate latticework of the poroi would become manifest. There is obviously a host of everyday observations that may have inspired this illusionist solution. One of Empedocles’ Sicilian bays may have a shore that looks brown when seen from a promontory; another may have a shore that gives the distant appearance of grey. The person who descends the promontory to walk the two shores realizes, however, that the overall appearance is produced by a mixture of red and black pebbles, in the first case; by a mixture

118

Alexander P. D. Mourelatos

of black and white ones, in the second. Also, as weavers must have known from time immemorial, intermediate color shades – often especially vivid ones – can be obtained by weaving together threads of contrasting color in various patterns of close alternation. This analogue from weaving has an obvious affinity with Empedocles’ conception of poroi: the warp wends its way through the poroi of the woof, and vice versa. Effects of this genre, especially where the mixing of colors is involved, are known to modern theorists by the term “optical fusion.” The effect of optical fusion, which involves either visible texture (oil and water) or color (beach pebbles, colored threads), can be generalized, of course, to all our outer senses. It is indeed rather magical (cf. Empedocles’ thespesiethen) that brown comes to be from the close juxtaposition of red and black; but the magic is that of illusionism. There is also dazzling magic in the production from the four elements of the special tactile quality and grade of dryness characteristic of bone; but for one who understands the conjurer’s trick, the right object of marvel is no longer the effect itself but the skill shown in the trick’s execution. That Empedocles’ Love should also – in addition to its roles of carpenter and potter – take on the guise of a conjurer is mythologically appropriate. Empedocles himself was something of a magician – as is suggested by his public stance of “divine man” (theos, see B112.45) and miracle-worker (B111), and by his reputation in the biographical tradition for “playing the wizard” (JRƝWHXRQWL, Diogenes Laertius VIII.59; cf. DK31A1). The simile of painting – which I earlier discussed as evidence of the recognition among Neo-Ionian philosophers of the problem of quality emergence – now assumes a far more profound significance. Some of Love’s tricks will not be merely “analogous to,” they will in fact be identical with trompe l’oil devices used by painters. We saw earlier that B23 is best interpreted as envisaging a direct mixing of paints. Empedocles – who (as has been argued in the preceding pages) is not an emergentist, and who does not have either Aristotle’s theory of qualitative interaction or Anaxagoras’ ontologically lavish supply of latent ingredients – has only one way to explain the intermediate colors that result from the mixing of paints: optical fusion. Reasonably enough, he theorizes that the intermediate shade of brown arises because miniscule streaks of pure red and pure black pigment have been alternately juxtaposed. 74 It is relevant and fascinating to take note in this connection of a thesis advanced by Eva Keuls, that in the course of the fifth century Greek painters began to develop not only such techniques as linear perspective (in order to 74

It was only in the nineteenth century that it was realized that the laws of color mixing through optical fusion are not quite the same as those either of the direct mixing of paints or of mixing of translucent paints through superposition (as in glazing). Optical fusion conforms to the “additive” system, in which the primaries are green, red, and blue-violet; the other two methods, to the “subtractive” scheme, in which the primaries are red, blue, and yellow. See Keuls 1978, p. 67; cf. also Keuls 1975, pp. 3-4.

4. Quality, Structure, Emergence

119

create the illusion of distance) and color shading (so as to create the illusion of plasticity), but also a certain “divisionist technique exploiting optical color fusion; patches of color contrasted sharply to the nearby viewer, but seemed to blond when observed from the appropriate distance.” 75 Unfortunately, we possess almost nothing of direct archaeological evidence for Greek painting between 480 B.C.E (the date assigned to the Tomb of the Diver at Paestum) and the mid-fourth century. So, both Keuls and others who have attempted to reconstruct the history of Greek painting in its classical period – the time of such masters as Polygnotus, Apollodorus, and Zeuxis – have had to rely on the literary evidence. A terminus ante quem is furnished by certain passages in Plato that strongly suggest that Keuls’ “divisionist technique” was, by the early fourth century, both current among painters and familiar to the cultured public. An earlier terminus ante quem might be the start of Apollodorus’ artistic career, circa 430 B.C. If Keuls is right, the revolution in pictorial technique for which Apollodorus became famous included deployment of the divisionist technique. Admittedly, none of this gives us solid evidence for divisionism among Empedocles’ contemporary painters – almost a generation earlier than Apollodorus, by the usual chronology for Empedocles. But some obviously sage remarks by Vincent J. Bruno, who has studied the whole phenomenon of the development of color shading in fifth-century painting, give us a proper measure of the bearing of the indirect evidence cited above: [I]t seems highly probable that the first use of shading considerably predated the career of Apollodorus. Some method of showing three dimensions was very possibly first attempted in the period of Polygnotos, a period that seems to have witnessed many daring experiments in the technology of all the arts. . . . Shading, in its earliest beginnings, must have taken the form of tentative and only partially successful experiments by the artists of the early and mid-fifth century B.C. Apollodorus’s high success and fame as a painter probably rested on the fact that he restricted himself to the simplest shading methods, perfecting them, and imparting to his paintings a more convincingly three-dimensional appearance. 76

What Bruno says concerning the development of shading applies with equal force to Keuls’ thesis of the development of a divisionist technique. It would be completely in character for the intellectually versatile Empedocles to have apprised himself with engaged interest of experiments in illusionism in the visual arts. His contemporary Anaxagoras, it should be recalled, wrote a treatise on linear perspective. 77 In any event, my thesis concerning Empedo75

Keuls 1978, p. 79; cf. pp. 59-87, or Keuls 1975, pp. 5-16. Bruno 1977, p. 28 f. 77 Vitruvius 7. praef. 11 = DK59A39. 76

120

Alexander P. D. Mourelatos

cles’ awareness of effects of optical fusion does not depend on Keuls’ thesis of divisionism in fifth-century painting. The basic phenomenon of optical fusion must have been familiar from the sort of everyday situations – colored pebbles on the beach, weaving of textiles – I mentioned earlier. 78 I must now return to the two observations that initially gave credence to an emergentist reading of Empedocles: the contrast between Anaxagoras’ scrupulous avoidance of the language of genesis and Empedocles’ tolerant adoption of it; the ontological status of the Sphairos. We are now in a position to do justice to the first observation. The contrast with Anaxagoras is not at all lost. Anaxagoras explains away putative effects of emergence. Once one has accepted the message of Anaxagorean metaphysics, not only is emergence tout court abolished, even emergence qualified by the seemingly disarming adjectives “putative” or “ostensible” becomes an illegitimate concept. For Anaxagoras absolutely nothing emerges: no appearance, no epiphenomenon, no mirage, no illusion. Our senses grasp exactly what is real; they simply do not grasp the whole of what is present. Empedocles, by contrast, gives a reductive explanation, but an explanation, nonetheless. The existence of a certain illusion, of qualities that are other than the qualities of the four elements, is not denied. Both these illusory qualities and the ephemeral compounds that give rise to them manage to last long enough to make it possible for us to use – with no bad philosophical conscience, and with considerable plausibility – the language of birth, death, and change with respect to them. The problem raised by the Sphairos is more recalcitrant. Dare we say that the homogeneous appearance of the Sphairos is an illusion? I have noted earlier that the language used to describe that state of homogeneity has a strongly phenomenological flavor, which makes a prima fade case for the illusory character of the unity. And yet, other aspects of the language Empedocles deploys in speaking of the Sphairos – a tone of religious awe toward that “god” – make it seem almost blasphemous that any feature of the Sphairos should be called “illusory.” We have, I believe, another option: If the scenario of the origin of the poroi I sketched earlier is true to Empeclocles’ thought, then there are no poroi in the Sphairos. It is only under the conditions of adversity created by Strife that a structure of bolts and sockets is needed to bring things back together into harmonia. It is logical to assume that once the influence of Strife has been decisively neutralized, then the four elements unite readily and pervasively. This does not involve anything like Aristotelian interactive blending; the ele78

The metaphysical bearing of the pictorial technique of divisionism was noticed by Barnes 1979a, II, p. 24. But he did not go beyond drawing a simile: “Atomists are physical pointillistes: their world is made up of microscopical dots . . . Anaxagoras was a painter of the traditional type: his world is made of stuffs mixed through and through. . . .” Wright 1981, who denies (pp. 38-39 and 180) that Empedocles envisages direct mixing of paints in B23, does not consider the possibility of mixing through optical fusion.

4. Quality, Structure, Emergence

121

ments are not demoted to potential being. We have rather, for this one phase of the Empedoclean scheme, the same conception of through-and-through reciprocal interpermeation that we had in Anaxagoras. The model I gave for Anaxagoras was that of the compatible mixing of the hot, the white, and the liquid in a cup of warm milk. In the latter case, those three features can all manifest themselves together when present in equal portions, since no contrariety is involved. By contrast, the features that constitute the Empedoclean four elements do involve contrarieties. This commits us to the consequence that the Sphairos is both hot and cold, dry and wet, light and heavy, bright and dark – and so forth. Earlier, when we were considering what Empedoclean mixture might be like without the benefit of poroi, the possibility that ordinary compounds (plants, animals, etc.) should have contrary features struck us as absurd. But the Sphairos is a unique, extraordinary, and eschatological entity. For the theme of B27a, that “neither dissension nor conflict has its season within the limbs [of the Sphairos],” 79 what would be a more natural ontological basis than a conception of the Sphairos as the locus of a coincidentia oppositorum? This interpretation of Empedocles has the support of Plato in that famous passage of the Sophist (242D-E) in which Empedocles is dubbed “the more relaxed Muse of Sicily,” inasmuch as he posits only epochal advents of a unification of the many, in contrast to Heraclitus, “the stricter Muse of Ionia,” whose “one God” is permanently the locus of contrary characteristics (DK22B67). The “thick concealment” that suppresses the individuating features of the four elements in the Sphairos need not be any veil of appearances, or a blurring caused by distance; it is simply the neutral, nondescript state reached by a stuff that is exactly as hot as it is cold, exactly as dry as it is moist – and so forth for all relevant pairs of opposites. The featureless calm of the Sphairos is an ontological counterpart, so to speak, of the epistemological state of quietude the later Pyrrhonists achieve when all beliefs have been suspended, after repeatedly reaching judgments of the form “X is no more F than it is not-F.” Philoponus was right to gloss the Sphairos as apoios, “qualityless” (In De gener. et corr., 19 = A41).

79

Ƞ੝ ıIJ੺ıȚȢ Ƞ੝ įોȡȚȢ ਥȞĮ઀ıȚȝȠȢ ਥȞ ȝİȜ੼İııȚȞ. Bollack’s retention of the attested reading ਥȞĮ઀ıȚȝȠȢ, which he translates “de saison” (cf. ਥȞ Į੅ıૉ) makes perfect sense: see Bollack 1969, II, pp. 44-45; III, pp. 143 f. With the exception of Wright 1981 (pp. 255 f.), interpreters of Empedocles have referred this fragment to the Sphairos. She gives it a context of moral edification, which is the train of thought of Plutarch when he quotes this one-line fragment. But it is more likely that Plutarch is borrowing only the rhetoric of Empedocles’ saying. The statement is, of course, strictly false for any compound other than the Sphairos; and there is no reason to suppose that Empedocles would have used the metaphysically charged language of “the season of total concord” loosely or informally. Of course, even if the line Plutarch quotes did not refer to the Sphairos, the condition it describes fits the Sphairos better than it fits any other entity. The borrowing of the rhetoric would, in that event, be mine.

122

Alexander P. D. Mourelatos

XI. THE PREFIGURING OF STRUCTURE IN EMPEDOCLES There remains the question of the emergence of structures in Empedocles. There is surely no evidence either in the fragments or in the testimonia of a doctrine of preformation – in the sense in which this term applies to Anaxagoras’ “seeds.” Similarly, there is no evidence that Love makes use of exemplars – something like Plato’s Forms. Shall we then say that with respect to the emergence of new structures Empedocles has no recourse and must simply concede their quasi-miraculous supervenience? Before we answer Yes, we are challenged by the principle of interpreter’s charity to explore the relevant conceptual alternatives. Might not the theory of poroi make a significant contribution toward dispelling the mystery of effects of the growth of new structure, as well as of effects of qualitative mixture? In modern debates about organic evolution, biologists are often called upon to respond to a familiar argument advanced by creationists and skeptics: A chimpanzee pounding on a typewriter is not likely, not even if it had a life span of millions of years, to come up with a sequence of strikes that constitutes a complete Shakespeare play. The evolutionists’ rejoinder is that the colligations of atoms that ultimately lead to the evolution of organisms, random though they are in one sense, are not random in the sense in which the strikes by the chimpanzee of the thought experiment are so. The fact that the chimpanzee has once struck the sequence “HAMLET” does not at all enhance the likelihood that this non-human typist will do so again, let alone enhance the likelihood that it will soon strike the sequence “OPHELIA.” By contrast, every time matter achieves a new stable form of organization, the probabilities of certain newer and more complex forms of organization are enormously increased. We do not go straight from atoms to amoeba: we go to inorganic compounds, to organic compounds, to amino acids, to peptides, to proteins, and – on a separate but rotated line of development – to nucleotides, to nucleic acids, to viruses, to cells. Once the rudimentary chemistry of, say, nucleic acids has a toehold on planet earth, the double helix of DNA will not be long in coming. It is a far cry, to be sure, from the poroi of Empedocles to the electrochemical bonds of modern physical chemistry and molecular biology. And yet, the difference is one not of conceptual category but rather of type of material involved and of degree of structural complexity envisaged. According to our sources, the theory of poroi is appealed to also in explaining growth (Theophrastus De sensibus 12). 80 The poroi function to facilitate the aggregations that come to be manifested as growth; and, even more importantly, they promote certain types of growth and impede others. Indeed, at any stage in the Empedoclean evolutionary scheme, the existing texture of poroi places significant constraints on processes both of growth and of decay. Some of the sting of 80

Cf. Stratton 1917, p. 168 n. 42.

4. Quality, Structure, Emergence

123

the Melissan argument concerning PHWDNRVPƝVLV, “rearrangement,” is removed: not any old transformation is possible; just the ones that involve the proper alignment – either in coupling or in decoupling – of the poroi. An important additional constraint is provided by the continuous activity of Love. Empedocles enjoys, in this respect, a certain advantage over the modern evolutionist. It may, in fact, be easier for him to meet either the creationist or the skeptical challenge. His scheme is one of directed evolution, after all. Love may not have archetypes to work from; but she does have an unswerving determination to bring things back to the state of unity present in the Sphairos. At any given stage, she will coax from things the highest degree of unity permitted by the existing degree of fragmentation – or, to put it in other words, by the lingering influence of Strife. Shortly after Strife is at the height of its power, the closest to unity Love can bring things is to begin to assemble the smithereens into a fabric of strands and pathways, the poroi. In a later stage, even those non-viable monsters of B60 and B61 are probably bridge-structures, transitional forms of unity made necessary by the exigencies of a mechanics of poroi in a still hostile environment. Whether Empedocles’ scheme is as inexorably deterministic as that of Leucippus and Democritus is an issue that can be left open. The implication that some of the poroi are deformable (both flexible and distensible) would seem to make him – if I may indulge in this philosophical pun – a “soft determinist.” What matters for our present purposes is that the scheme is deterministic enough to provide for the sort of causal continuities that would make for intelligibility in the transitions from one structure to another. But now what about those smithereens that precede the laying out of the poroi? Broadly speaking, their chaotic scatter constitutes some sort of a network of relations – not a kosmos, to be sure, but not a nothing either. Indeed, is it not likely that before Love’s triumph there will be enough Strife in the world so that things will tend to fall apart not only as they uncouple along the smooth tracks provided by poroi but also through the sort of violence that either flattens the poroi by cutting along their length, or breaks their continuity by cutting across their width, or squeezes them tightly shut, making their walls collapse? How are we to account for the advent of these events of akosmia, disarray, and collapse? The question may seem jejune, since the answer is obvious: they are direct effects of Strife. The interest of the question lies in the special character of the causality at issue. Strife makes its mark on the world by breaking things up. Its nature is reflected on its products. Clearly, what is operative here is the socalled Transmission Model of Causality: the cause communicates its character to the effect. 81 And what is said of Strife in this connection applies equally to

81

See Mourelatos 1984a, pp. 1-3.

124

Alexander P. D. Mourelatos

Love. She too stamps the imprint of her nature on the world. Empedocles is explicit in his recognition of this metaphysical affinity of cause and effect: From earth there is a pouring forth (proreousi) of things that are [foundations? rooted?] and that are solid in appearance (VWHUHǀSD). But at the time of Hate all things turn into split kinds (diamorpha) and go separate ways (pelontai . . . andicha). Then, at the time of Love, they come together or are longingly sought (potheitai) by one another. (B21.6-8) 82

The image of “pouring forth” need not be taken metaphorically. Empedocles may be thinking of the ash and lava that pour out of the crater of a volcano (Aetna or Vesuvius), building up the volcano’s cone – and he may have theorized that all mountains are volcanic in origin. Almost certainly he is thinking of the extrusive process whereby earth forms the shell of mollusks and turtles (B76), arid of the percolating of earth into the trunks, limbs, and foliage of trees. In any event, this percolating of earth stuff, being a phenomenon of growth, takes place along the channels of poroi; so, the image of flowing is quite apt. 83 But the rhetoric of “pouring forth” spills inevitably and naturally on to the next two lines, making us think of the “outpouring” of Love and Hate. Significantly, the idea that F things are caused by an F agent is conveyed here by a verb that was destined to furnish one of the root metaphors of the Transmission Model of Causality. Here, then, is yet another way in which advents both of kosmos and of akosmia, structures and scatterings, have relevant antecedents in Ermpedocles’ scheme. Any advent of kosmos is pervasively imbued by the nature of Love; any advent of akosmia is imbued by the nature of Strife. There is, then, in Empedocles, too, something of the mechanisms of prefiguring that Philolaus and Democritus exploited. The philosopher who spoke with the greatest eloquence and with quasi-religious awe about the miracle of emergence, has shown himself wonderfully resourceful in naturalizing the miracle. XII. THE PRIMACY OF STRUCTURE IN PLURALIST SOLUTIONS TO THE PROBLEM OF EMERGENCE Our initial survey of the evidence concerning solutions to problems of emergence in Empedocles gave us the impression that this Protean philosopher might have been rather eclectic in his theoretical proposals. It should now be clear, I hope, that this is far from being the case. The conception of the poroi 82

The translations “[foundations]” and “[rooted]” are based, respectively, on the excellent emendations șİȜȣȝȞ੺ and șİȜİȝȞ੺ that have been proposed, in place of the obviously corrupt ș੼ȜȘȝȞĮ IJİ and șİȜ੾ȝĮIJĮ of the manuscripts. 83 Wright 1981, pp. 101 and 178, speculates that the ʌȡȠȡ੼ȠȣıȚ of line 6 is reason to suspect reference to water; perhaps as ș੺ȜĮııĮ, in the corrupt segment (see preceding note) that immediately follows in that line.

4. Quality, Structure, Emergence

125

and of Love’s artful exploitation of their mechanical logic is at the heart of the solution both in the case of compound structure and in the case of mixed quality. The ancillary thesis of illusionism makes it possible for Empedocles to argue that all combinations in the world we know, including those of qualitative mixture, are cases of structural assembly. That the Sphairos is an exception does not weaken the theory, since this creature is different enough from ordinary mortal compounds to command its own special principle of coherence. The crucial point is that Empedocles does not have to appeal, and does not seem to appeal, to any intuitive principle of purely qualitative convergence. The essentially structural character of the theory of the poroi was perceived by Aristotle. In a move that is as insightful philosophically as it is hermeneutically mischievous, he assimilates Empedocles’ theory to that of the atomists. 84 If the poroi are passages, Aristotle comments, they must have exactly the function of the atomists’ void: to permit motion, and thus coming-to-be, perishing, alteration, and growth. And if there are passages, there must be unperforated discrete bodies – call them “undivided” bodies, call them atoms. The difference between the systems of Empedocles and of the atomists, Aristotle seems to imply, is in the degree of congestion and suppleness of the bodies at issue – not a metaphysically significant difference. Though not unobjectionable as an interpretation of either of the two systems it compares, Aristotle’s comment is intriguing enough to invite adjustment and elaboration. What we have in Empedocles is indeed a transitional stage to atomism. Democritus’ mechanics is one of hard and rigid structures; that of Empedocles, one of structures that are relatively soft. In Democritus we have an absolute contrast between (i) sheer void and (ii) completely full bodies. In Empedocles we have correspondingly layered explanations that are appropriate for an emphatically plenum-scheme of four bodies that differ from one-another in their inherent grades of plasticity. Instead of explaining aggregations in terms of the entanglement of mostly convex structures (spherical, polyhedral, knobbed, barbed, or spiked atoms), Empedocles explains such aggregations in terms of reciprocities between deeply concave structures, the poroi. Instead of supposing one factor, the void, to provide room for the motions that lead to aggregation, and a separate factor, the geometry of atomic shapes, to provide the nexus involved in the emergence of compounds, Empedocles’ scheme is emphatically monist, and in that respect conceptually more coherent. For with the poroi of Empedocles we have the single factor that provides both grooves for the motion and devices of nexus for all parts of the cosmic whole. Interestingly, all but one of the four Pluralists that have been discussed, either briefly or at length, in the present Chapter or the previous one, appeal to 84

See Gener. et corr. I.8.325b6, ıȤİįઁȞ į੻ țĮ੿ ૃǼȝʌİįȠțȜİ૙ ਕȞĮȖțĮ૙ȠȞ Ȝ੼ȖİȚȞ ੮ıʌİȡ țĮ੿ ȁİ઄țȚʌʌંȢ ijȘıȚȞ. Cf. entire passage from 325b1 ff.

126

Alexander P. D. Mourelatos

structure for solving problems not only of the emergence of structure but also of the emergence of quality. The exception is, of course, Anaxagoras. His seeds may help explain why a certain reallocation of qualitative ingredients takes place at a given time; but the qualitative effect itself is accounted for purely by the three principles I discussed In Section V (UM, dominance/latency, sensory limitations). Even Philolaus’ doctrine of the tempered convergence of qualitative opposites in harmonia has recourse to structural principles – those of mathematics. As soon as the problem of the emergence of structure was focused on, the insight seems to have immediately taken hold that the transformational logic inherent to structures – arithmetical, or geometrical, or mechanical, or biological ones – offered the best promise for a solution to that one node of the two-fold problem of emergence. It was the logical next step for Empedocles, Philolaus, and Democritus to apply the solution to the case of quality-emergence. To bring this off some reductionist assumptions had to be intercalated. Philolaus could help himself to the various synaesthetic and quasi-synaesthetic associations between qualities and structures that had gained currency in the Pythagorean tradition (the Pythagorean Table of Opposites being one expression of this thought pattern). Democritus opted for an eliminativist version of materialism. There are serious problems, as we well know, with either of these approaches. The Empedoclean approach of illusionism – both too ingenious and too ingenuous, almost playful, and bizarre as it might appear to us – makes an impressive claim of intuitive plausibility. No category shifts are involved. The elements are endowed with more than structural-mechanical properties; they possess an appropriately thrifty set of basic sensible qualities. The immense variety of colors we see, flavors we taste, feelings of heat and cold we experience – and similarly with other examples – are effects of the perceptual fusion of those few basic sensible qualities of color, flavor, etc., that are respectively inherent in the four elements. Accordingly, Empedocles can claim for his system not only the virtue of theoretical parsimony, he can also claim a good share of that thorough-going realist and empiricist vision that is a captivating feature of the system of Anaxagoras. Trusting on the ministry of Lynceus and on his own knowledge of the magician’s tricks, he could even afford to flaunt a variant of the Anaxagorean challenge: How could color come from what is not color?

5.

The Concept of the Universal in Some Later pre-Platonic Cosmologists

The history proper of the concept of the universal has an obvious and familiar start: the philosophies of Plato and of Aristotle. Both of these classical philosophers recognized – indeed they made it the central part of their metaphysical teaching – that we grasp reality in two significantly different modes: in terms of concrete individuals or particulars (e.g., Socrates, the Parthenon, Miltiades’ sword, the jury’s verdict in Socrates’ trial); and in terms of types or kinds (e.g., man or animal, temple or building, sword or weapon, justice, wisdom). 1 The term used by both philosophers with reference to the second of the two modes is, of course, eidos or idea, “form,” “kind”; and Aristotle also employs the ancestor to the modern term “universal,” viz., to/ta katholou (in contrast to to/ta kath’ hekaston), as well as the term to koinon, “the common,” and other terms as well. 2 Within and beyond classical ancient philosophy, and down to our present day, the concept of the universal undergoes many transformations; and the so-called “problem of universals,” i.e., the long-standing debate between realists and nominalists, plays a major role in these transformations. Accordingly, modern scholars who write on Plato or Aristotle have often questioned whether either Plato’s or Aristotle’s “forms” may qualify as “universals” under one or another of variously more restrictive definitions of the concept of the universal. 3 In this Chapter, I seek to rely on a reasonably minimalist specificaPreviously published in Mary Louise Gill and Pierre Pellegrin, edd., A Companion to Ancient Philosophy, Blackwell Companions to Philosophy. Oxford: Blackwell Publishing, 2006, pp. 56-76. The study is reprinted here with some revisions. An earlier shorter version had appeared in Greek, under the title, “ਝʌĮȡȤ੻Ȣ IJોȢ ijȚȜȠıȠijȚțોȢ ਩ȞȞȠȚĮȢ IJȠ૨ «țĮșȩȜȠȣ»,” in ȆȡĮțIJȚț‫ ޟ‬IJ߱Ȣ ݃țĮįȘȝȓĮȢ ݃șȘȞࠛȞ [Proceedings of the Academy of Athens], 75 (2000), 509-26. – I thank Spiros Moschonas for transmitting to me copy of the full text of Rau (1923). 1

There is an excellent concise account of the development of the concept, and of related metaphysical debates, beginning with Plato and Aristotle, in the lexicon for ancient philosophy, Bächli–Graeser 2000, s.v. “Allgemeines,” pp. 18-24. 2 The locus classicus for the introduction of the distinction between universals and particulars in Aristotle is De interpr. 7.17a38-17b1. 3 For a comprehensive recent comparison of how the concept of the universal is introduced and how it functions in the two philosophies, respectively, of Aristotle and of Plato, see Segalerba 2003, esp. pp. 67-70, 82-83; cf. Segalerba 2001, pp. 95-106, 223-27.

128

Alexander P. D. Mourelatos

tion of the content of the claim that the two classical philosophers have grasped the concept. By the qualification “minimalist” I signal my intention to assume a low-level, intuitively accessible, conception of the universal; and by the qualification “reasonably,” my concern that the specification should not be so broad as to allow any act of attending to the distinction between referring expressions and predicates (common nouns, adjectives, verbs) to count as recognition of the universal. But before I attempt such a specification, let me confirm what may be perceived as implied in the title of this Chapter: my account is not one about the history proper of the concept but about what might be called its “prehistory.” D. M. Armstrong’s elegant modern philosophical introduction to our subject (1989) opens with this remark: “The topic of universals is a very old one. It goes back to Plato at least, perhaps to Socrates, perhaps to even earlier times (p. 1).” My interest is precisely in those “earlier times.” I seek to identify and to discuss contributions to the grasp of the universal that were made by the earliest philosophers in the Western tradition, the Presocratics. 4 Socrates himself, no doubt, made major contributions to the development of the concept; that is Aristotle’s explicit testimony (Metaph. I.6.987b1-4; XIII.4.1078b17-19 and 27-31; 1086b1-5), and that is what Plato, in his portrait of Socrates, would have us believe. But the Socratic problem is a scholarly issue I need to skirt. For our purposes here, Plato (including Plato’s Socrates) will serve as the terminus ante quem. Still, before proceeding, some thought is in order concerning the level of understanding of the distinction between particulars and universals we may assume for the Socratic-Sophistic circles of the fifth century that form the context and background of many of Plato’s early and middle dialogues. For it has been widely supposed by modern scholars that in this group of dialogues, fifthcentury characters other than Socrates (his interlocutors) are presented as inept in handling the concept of the universal. The alleged ineptness is that of confusing – at least initially in the conversation – universals and particulars. But Alexander Nehamas (1975 and 1999) has shown that this interpretative diagnosis has little support in the relevant texts (1999, pp. 159-76) Typically the initial error is rather one of offering too narrow a characterization of the universal at issue, not the error of citing merely an instance or example. More significant is the fact that Socrates often asks, “Is there such a thing as the F (e.g., “the pious/piety,” or “the beautiful/beauty”)?” or “Do we say that the F is something?” and his interlocutors unhesitatingly answer, “Yes, there is,” or “Yes it is.” Plato and Plato’s Socrates seem to regard it a prerequisite – indeed one they typically assume as having been fulfilled before the conversation starts – that the two sides in dialogue have an understanding of the difference 4

As it has often been pointed out, more than half of the “Presocratics” are contemporaries or nearcontemporaries of Socrates. The term is best understood as referring to fifth-century philosophers who were not influenced by Socrates.

5. The Concept of the Universal in Some Later pre-Platonic Cosmologists

129

between “Which thing is/things are F” and “What it is to be F.” It is not unreasonable to imagine, therefore, that issues concerning either the distinction between a universal and its instances or the distinction between universals of different scope (ones at different levels of generality, genus-species distinctions) would have been broached in the course of Sophistic instruction and Sophistic debates. Strongly suggestive evidence of this is in the reports concerning Prodicus’ practice of RQRPDWǀQ RUWKRWƝV, “correct use of words,” which entailed both the formulation of definitions and the canvassing of issues of sameness and difference between abstract concepts (typically expressed by infinitives, articulate adjectives, or nouns; see DK84A9, A11, A13-A19). 5 Also relevant and suggestive is the evidence implied in Plato’s testimony that Gorgias would admit only special kinds of DUHWƝ, “excellence,” but no single encompassing type (Meno 71B-73C). And it is worth noting that the theme of sameness/difference involving abstract concepts is conspicuous in the Sophistic treatise Dissoi logoi (DK90). The internal dating of this treatise, however, places its composition some time after the end of the Peloponnesian war (DK90.I.8); so, the Dissoi logoi is rather too late to qualify as pre-Platonic. But what about the main line of the Presocratics, the cosmologists or natural philosophers of the tradition (Aristotle’s physiologoi)? What thematically significant uses of the concept of the universal are found in their sayings and doctrines? What evidence of antecedents, anticipations, or pre-formations of the concept might be found in what we know of the philosophical writings and projects of these philosophical pioneers? In this Chapter I shall concentrate on contributions made by the most prominent of the fifth-century physiologoi – all of them, of course, contemporaries of Socrates. I. CRITERIA USED FOR THE CONCEPT OF THE UNIVERSAL Let me now offer what was promised at the start, a “reasonably minimalist” specification of the concept of the universal. Using the philosophies of Plato and of Aristotle as our benchmark, we can safely claim that the concept is recognizably present in the thinking of the two classical philosophers on the basis of three criteria. The criteria are not disjunctive; all three must be met if we are to claim that the concept has been grasped by a particular philosopher or in a certain period. I give the criteria first in a bald statement, and then I append needed explications and comments. In both the two classical philosophers we find: (1) the type-token distinction in thematically significant contexts; 5

As indicated above in the table of Abbreviations reeferences to fragments from the works of the Presocratics or to testimonia (reports) concerning their doctrines, is, as usual, by the chapter number and section (“A” for testimonia, “B” for putative excerpts from works) in the edition by Hermann Diels (revised by Walther Kranz), standardly referred to as “Diels-Kranz” or “DK”.

130

Alexander P. D. Mourelatos

(2) uses, in thematically significant contexts, of a distinction between two types, a type1 (subordinate) and a type2 (superordinate); (3) some awareness, explicit or implicit, of the distinction between (a) the type-token relation, or the type1-type2 relation, and (b) the mereological relation that connects the whole of any mass-entity (partitive entity) with the component parts or portions of that whole.

The first and most famous formulation of the first criterion is, of course, in Aristotle’s distinction between KHQDULWKPǀL and KHQHLGHLORJǀLJHQHL “one in number/numerically one” vs. “one in form/definition/genus/species.” In Plato this helpful terminology has not yet been established, though it is often implied in context. I shall need to dwell some on the first criterion, not only so as to justify my seemingly anachronistic preference for C. S. Peirce’s now familiar terminology of “type” vs. “token” (see Armstrong 1989, pp. 1-2; cf. Audi 1995, s.v.) but also because ultimately this criterion will emerge as especially apt for the purpose of assaying contributions to the concept of the universal by fifth-century cosmologists. Three texts merit our attention: first, the passage in Peirce which introduces the terminology and explains its rationale; and then – most interestingly – two ancient texts in which Peirce’s distinction is prefigured. Here is the key passage in the Peirce corpus: A common mode of estimating the amount of matter in a MS. or printed book is to count the number of words. There will ordinarily be about twenty “the”s on a page, and of course they count as twenty words. In another sense of the word “word,” however, there is but one word “the” in the English language; and it is impossible that this word should lie visibly on a page or be heard in any voice . . . . Such a definitely significant Form, I propose to term a Type. A Single event which happens once and whose identity is limited to that one happening, or a Single object or thing which is in some single place at any one instant of time , . . . I will venture to call a Token. 6

D. M. Armstrong notes that Peirce’s distinction is one “that practically all contemporary philosophers accept” (1989, p. 1). This is ostensibly because of its “minimalist” character, which permits capture of the essentials of the distinction between universals and particulars with the lightest burden of metaphysical interpretation or question-begging. Often the distinction is illustrated not with Peirce’s example of individual words but with that of individual letters. Amazingly, in this version the distinction had already been formulated in antiquity. For it occurs unmistakably in Galen:

6

Peirce 1931-1960, vol. 4, pp. 423-24 (= section 537). I thank Fred Kronz for pointing me to this passage. Later in his career, Peirce changed his terminology, calling type a “famisign” and token an “actisign.” But eventually (in 1908), he conceded that the “former names are better than the ones I now use”: Peirce 1998, vol. 2, p. 488.

5. The Concept of the Universal in Some Later pre-Platonic Cosmologists

131

When an utterance (SKǀQƝV) signifies a single thing (KHQ VƝPDLQRXVƝV), necessarily the type of the thing at issue is one. Nonetheless, a given letter can be many in number (DULWKPǀLSROOD), as in the case of the letter alpha. And for this reason when we say that there are seven vowels in our language, and twenty-four elements of the alphabet in all, clearly we are considering the common form underlying them all (WRNRLQRQKDSDQWǀQHLGRV), and not the particulars of them (ta kata meros) written on papyrus, wood, parchment, and stone. . . . [W]hen we say that they number twenty-four in all, we canvass (procheirizomenoi) not the particulars but rather just the types (Ƞ੝ IJ੹ țĮIJ੹ ȝȑȡȠȢ . . . IJ੹ įૃ İ੅įȘ ȝȩȞȠȞ), to which, in my view, we attach the names. (De methodo medendi II.7.8-9; Kühn 1825, pp. 131-132) 7

The full context in Galen’s treatise shows that use of the alphabet example is motivated by Galen’s adherence to an Aristotelian doctrine of universals as inherent in things, universalia in rebus. Nonetheless, as with Peirce’s use of the example of words, Galen’s alphabet example seeks to make the concept accessible and intuitive, metaphysical baggage being kept to a minimum. 8 Even more remarkable – and perhaps sufficient for abating any lingering concern that my use of the type-token terminology may be anachronistic – is the fact that Galen’s use of the alphabet example has its precedent in Aristotle. In Metaphysics XIII (Mu), ch. 10, Aristotle uses this very example in order to distance himself from the Platonic tenet that the distinction between universals and particulars requires a separately existing Form, a transcendent “the F itself.” The relevant text is intricately dialectical, but the rationale in the use of the alphabet example comes through clearly enough: [T]here is nothing to stop there being many alphas and betas, as with the elements (VWRLFKHLǀQ) of speech (WƝVSKǀQƝV), without there also being over and above the many (para ta polla) a certain “alpha itself” (auto alpha) and “beta itself” (kDLDXWREƝWD). (Metaph. XIII.10.1087a7-10) 9

It is telling that the Neoplatonist commentator Syrianus, in his discussion of this passage, resists Aristotle’s deflationary exploitation of the alphabet example. 10 And it is likewise telling that a modern commentator adopts Peirce’s terminology of types and tokens in paraphrasing the same text of Aristotle’s: 7

Cf. Hankinson 1991, pp. 66, 213-14. I have modified Hankinson’s translation. He very perceptively gives the translation “tokens” for ta kata meros; but this might well seem tendentious in the context of my own argument. 8 “And it is so evident, and it so naturally belongs to everything, be it man or beast, to recognize . . . the unitary form (KǀV HLGRV KHQ), that even donkeys, . . . the stupidest creatures of all, manage to distinguish between things which are one in form and those which are one numerically”: De meth. med. II.7, Kühn 1825 p. 133; transl. from Hankinson (1991), p. 67. 9 My translation, drawing on Annas 1976, p. 115. I am grateful to Dr. Segalerba for calling my attention to this important passage, 10 I translate from Syrianus, In Metaph. (Kroll 1902, pp. 163-64): “He [Aristotle] claims that just as the elements (stoicheia) of speech (SKRQƝV) make for infinitely many utterances, even though they are

132

Alexander P. D. Mourelatos

There can be many elements or letters of the same kind, e.g. many As, without this implying either that there is a Form of A or a mysterious universal A. Tokens can be tokens of the same type without this leading to the manufacture of an exalted status for the type over and above its tokens and separate and independent of them. 11

The fact that Platonists may find the type-token distinction unacceptably deflationary is not surprising. The important fact in the present context is that, no matter how Plato and his followers would have ultimately conceptualized the distinction between the F-itself and F-things, that distinction involves at a minimum the distinction between the F-type and F-tokens. Turning now to both the first and the second of the three criteria for the universal listed above, let me explain the qualification “thematically significant.” I call a distinction “thematically significant” not only if, in the context at issue, the distinction itself becomes a topic of discussion, but also if major philosophical claims are being put forward by virtue of deploying the distinction. Examples of thematically significant uses of either the type-token or the type1-type2 distinction in Plato and Aristotle are far too numerous to list. It may suffice just to allude here to two famous passage in Platonic dialogues. In Phaedo 74A-C Socrates prompts Simmias’ agreement that we ought to distinguish between “the equal” as it applies concretely to sticks or stones and auto to ison, “the equal itself,” or auto ho estin ison, “what it is to be equal,” or LVRWƝV, “equality.” In Meno 71D-E Socrates calls on his interlocutor to specify the encompassing type referred to as DUHWƝ, “virtue, human excellence” (type2); but Meno (adhering to the teachings of Gorgias) responds by citing plural subordinate or special types, such as “virtue for a man,” “virtue for a woman,” “virtue for a slave,” etc. (type1). Briefly after this exchange, Socrates distinguishes between VFKƝPD, “shape” (type2) and VFKƝPD WL, “a shape,” e. g., strongylon, “curvilinear,” or euthy, “rectilinear” (type1). The third criterion may seem at first blush too technical to qualify as part of a “minimalist” specification; and yet it is absolutely essential and elementary. For, without it, the type-token distinction (and thus the distinction between a universal and its particulars) might be confused with the distinction between a partitive whole and its parts. Though there is a certain one-many relation between, say, scoopfuls, or even stretches, of water in a lake and the lake water taken as a whole, that relation is “mereological,” it is not one between either a token and a type or between a subordinate type1 and a superordinate type2. A

11

not separate from utterances (WǀQ SKǀQǀQ), in like fashion the principles of things are inseparable from individuals . . . . [But] we would not be able to produce an infinite number of utterances out of the twenty-four specific elements of the alphabet if we did not have within ourselves both the form (eidos) of each and also the patterns for combining them all (ORJRXVV\QWKHVHǀV).” I thank Dr. Gerald Bechtle for calling my attention to this passage. Annas 1976, p. 190. And, without using Peirce’s terminology, but, in effect, applying the same distinction to the passage from Metaph. XIII.10: Segalerba 2001, pp. 210-12.

5. The Concept of the Universal in Some Later pre-Platonic Cosmologists

133

thinker who grasps the mereological relation may not necessarily be en route to grasping the universal. Plato took pains to distinguish the relation that holds between a Platonic Form and its instances from the one that holds between a mass-entity and its portions. In Parmenides 131B-C the istion, “Awning” argument (not the “Sail” argument, as istion is commonly mistranslated) is directly aimed at an assimilation of the universal to a mass-entity. 12 II. SOME CONCEPTUAL BARRIERS TO EARLY GRASP OF THE UNIVERSAL One of the reasons why the pre-history of the universal is an engaging topic is that there are certain thematic barriers that may have served to block or to delay recognition or significant deployment of the concept. In cosmologies that advocate some version of material monism, the concept of the universal inevitably lacks prominence, or indeed relevance. We today, as students of Anaximander, might distinguish between (a) his apeiron as a massive individual and (b) such properties as are attributed to it in our sources, viz., “eternal,” “unaging,” “deathless,” “indestructible,” “encompassing.” That is, we are free to envisage these properties as in principle or potentially detachable, applicable also to individuals other than the apeiron. But that is precisely what Anaximander’s monism precludes. These properties are possessed uniquely and exclusively by the apeiron. For Anaximander, the cosmic individual and its properties form an inseparable whole. The cosmology gives us no encouragement or stimulus to detach the properties, to begin to view them as universals. What is said here about Anaximander’s apeiron applies correspondingly to the cosmic air of Anaximenes, which is likewise the subject of such attributes as “infinite,” and “divine” (DK13A5, A6, A9, A10). But it also applies with reference to the various states of rarefaction-condensation of cosmic air. For, however natural it might be for us to view like states of rarefaction (or of condensation) in different regions or at different times as distinct tokens of the same type, for Anaximenes these manifestations in different regions or different times are integral parts of a global individual. Pursuing this line of thought, one might suppose that the one God of Xenophanes, the unitary what-is of Parmenides, and the One of Melissus should perhaps also be judged barren as conceptual soil for the universal. And yet, most interestingly, pluralism as well may be inhospitable to the concept of the universal. As W. A. Heidel pointed out more than a century ago (1906), a thought-structure that holds sway in philosophical and scientific thought before 12

The text, in translation as in Allen 1983, p. 8, slightly revised: [Parmenides to Socrates] “You make one and the same thing be in many different places at once, as if you’d spread an awning (istion) over a number of men and then claimed that one thing as a whole was over many. . . . But would the whole awning be over each man, or part of it over one and part over another? – [Socr.] Part. – Therefore, Socrates, the Forms (HLGƝ) are divisible, and things that have a share of them have a share of parts of them; whole would no longer be in each, but part of each in each.”

134

Alexander P. D. Mourelatos

Plato is that of composition/re-composition of partitive entities, or massentities, of stuffs. The “opposites” in Anaximander and in Alcmaeon, the intermingling earth and water of Xenophanes, the opposite morphai in Parmenides’ “Doxa,” the four elements of Empedocles, the infinite FKUƝPDWD of Anaxagoras’ cosmology, and (in parallel to these more properly philosophical contexts) the dynameis, “powers,” of the medical treatises – all these entities represent a hybrid category of thing-stuff-power-quality (see Mourelatos 1973, pp. 17-30). Even the four that became the canonical “contraries” in Aristotle’s scheme – the hot, the cold, the dry, the moist – are conceptualized by many of the Presocratics not as qualities or properties but as “the-hot-burning-stuff” or “the-drydesiccating-stuff,” fully on a par ontologically with those stuffs that bear familiar mass-noun designations, such as air, soil, bone, or sap. Accordingly, the hot in the oven and the hot in the kiln are not two tokens of the type hot but rather two portions of a single mass-entity that have been parceled out to different regions. There is good reason to believe that this dominantly mereological thought structure is a hold-over from a characteristically pre-philosophical conceptual scheme (Mourelatos 1973, pp. 21-22). For the reasons that have already been suggested above (when I specified criteria with respect to Plato and Aristotle’s grasp of the universal), the hybrid category that is at issue here may have served more as a block or deviation and less as a proper stage in the development of the universal. Interference from this hybrid category may also render moot the contribution to our account that could have been made by two noteworthy conceptual devices that are well-attested in Presocratic thought: the principle Like to Like; and the so-called Synonymy Principle of Causation, which stipulates that the presence of a certain property F in an effect must be traced to the presence of F in the cause (Barnes 1979a, I, pp. 88, 119). Clearly, the Like to Like principle can be applied to entities of the hybrid category – and this, almost certainly, is the original version of the principle. So applied, if the-hot-burning-stuff consorts with its like, the assimilation may be understood in concrete, physical, and mereological terms as attraction or agglutination between different portions of the same stuff. As for the Synonymy Principle, there are famously special versions of it in Plato and in Aristotle that point straight to form as an abstract entity and a universal: the Platonic formula, F-things are F because of the form F-ness; the Aristotelian metaphysical slogan, DQWKUǀSRV DQWKUǀSRQ gennƗi, “a human being begets a human being.” Applied, however, to entities of the hybrid category, the Synonymy Principle envisages a mere physical transfer (imparting, invading, encroachment, spillage, contamination) from cause to effect: when the hot brick is put in the water, the-hot spills out of the brick and heats (invades) the water. The community of the F-character is not one involving different tokens of the same type but rather one of redistribution of portions of the same individual.

5. The Concept of the Universal in Some Later pre-Platonic Cosmologists

135

Another factor that may blunt awareness of the type-token relation is attention to cyclically recurring phenomena. A modern student of philosophy would not require much prompting to realize that such recurring events as nightfall, dusk, daybreak, dawn, new moon, or – for that matter and more broadly – each of the four seasons, are all properly types, and that under each type there are distinct tokens distributed over more or less regular intervals in time. But early students of the cosmos, who were indeed pre-occupied with such recurrences, are likely to have perceived in them successive reappearances or returns of the same token individuals. Winter and summer would have been conceived as individuals of that hybrid category: they periodically make their advent, and then recede – even today, we speak of the coming of dawn, or the coming of winter. Other cyclically recurring events would have been perceived as temporal parts of such conspicuous individuals as the sun or the moon, no more detachable from either luminary than the traits and behavior of Achilles are detachable from the person of that unique individual in Homer. Those modern scholars who have looked for anticipations of the Platonic concept of Form independently of Aristotle’s suggestions on the topic have focused their investigations on that hybrid category of thing-stuff-powerquality, 13 the category I have specifically set outside the scope of the present account. This focus in scholarship is understandable, in view of the extensive overlap, both lexical and syntactic, between the early mereological scheme and that of Plato’s theory. In both schemes we have substantives formed from the articulate neuter (to thermon, “the hot” / to hosion, “the pious”); in both we have the language of “sharing,” “partaking,” “participating,” “communion” (metechein, NRLQǀQHLQ), as well as language for the converse relation of “inherence” or “presence in” (pareinai or eneinai with dative); and in both schemes the “sharing,” etc., admit of degrees (“more” and “less”). But it has generally been recognized that the mereological model can be applied, at best, only to a limited range of Platonic Forms. 14 It obviously fails in the case of arithmetical and geometric Forms – indeed in the case of structure-Forms, generally – also in the case of biological Forms, and in other cases. Given that the focus of the present account is on the universal as an abstract type (the paradigm being the type-token and the type1-type2 distinctions) “anticipations” that draw on the mereological scheme are simply off our topic.

13

14

See, e. g., Furley 1976, pp. 81-82; Furley 1987, pp. 69-70, 171-173; Moline 1981, pp. 84-97; Meinwald 1992, p. 375. See, e. g., Brentlinger 1972. I am not convinced by the argument in Moline 1981, pp. 143-55, that the mereological model persists into Plato’s middle and late periods.

136

Alexander P. D. Mourelatos

III. EMPEDOCLES: FORMULAE FOR COMPOUNDS; BIOLOGICAL FORMS; TYPEIDENTITIES ACROSS CYCLES

Aristotle recognizes three Presocratic antecedents for his concept of to katholou, “the universal”: numbers and proportions in Pythagoreanism; 15 the formulae of composition for compounds in Empedocles; 16 and Democritus’ attention to issues of “form” and “essence.”17 Aristotle is characteristically chary in all three cases, which may to some extent be justified by his narrow focus on “defining essence.” We, however, can be more generous. I shall take up Empedocles first. We find in the Empedocles fragments arithmetical formulae not only for bone (B96), which is the only case cited by Aristotle, but also for “the forms (eidea) of other tissue (DOOƝVVDUNRV)” (B98). Blood requires exactly equal parts of the four elements (LVƝ PDOLVWD, B98.1), whereas other tissues and body fluids involve specific variations from the norm of equality: slightly more of earth, or slightly more of the other three (B98.4). The numerical formula for bone envisages transparently a universal, as Aristotle noted. Moreover, the principle of variation of proportion allows Empedocles to construct an entire domain of types of biological tissue. The term eidos and its cognate idea are prominent in yet another context in Empedocles. The plurals (HLGƝ, eidea, ideai) are employed along with the plural of ethnos, “tribe, kind” (ethnea) in reference to kinds of living beings, including gods (who for Empedocles are GROLFKDLǀQHV, “long-lived,” but not immortal). 18 Emphasizing the immense variety of these forms and kinds, Empedocles describes their evolutionary advent in these words: . . . countless tribes (ethnea myria) of mortal beings poured forth, structured in all sorts of forms (SDQWRLDLVLGHHVLQDUƝURWD), a marvel to behold (thauma idesthai). (B35.16-17; cf. B35.7)

15

Metaph. XIII.4.1078b17-23: Socrates was first to seek katholou horizesthai, “universal definitions”; Democritus “touched on it” (see below); the Pythagoreans, asked “what is opportunity (kairos) or the just or marriage” and reduced these to numbers (HLV WRXV DULWKPRXV DQƝSWRQ). Cf. Metaph. I.5.985b23-31, 987a13-22. For an explanation of such reductions of abstract entities to numbers, see McKirahan 1994, pp. 91-111. 16 Phys. II.2.194a20-21: “Empedocles and Democritus touched in small part on form (eidous) and essence (WRXWL ƝQHLQDL).” Cf. PA I.1.642a17-24: Empedocles “driven by truth itself is compelled to say that constitutive nature (physis) is the ratio, as when he explains what bone is”; Metaph. I.10 “even Empedocles says that bone is what it is by virtue of a ratio, which is the essence (WR WL ƝQ einai) and reality (NDLKƝRXVLD).” 17 See Phys. text in preceding note. Cf. Metaph. XIII.4.1078b17-21: “Democritus just touched on it slightly (the katholou), and roughly (SǀV) defined the hot and the cold.” Also PA I.1.642a17-24: Democritus was the first to touch on WR WL ƝQ HLQDL, “essence,” and on WR KRULVDVWKDL WƝQ RXVLDQ, “working out definitions of the reality at issue.” – See also the corresponding Sections on Empedocles, Philolaus, and on Democritus in Chapters 3 and 4, above. 18 In addition to the citations from DK31B98 (above in the text), plural forms HLGƝHLGHD also appear in B22.7, B23.5, B71.3, B73.2, B115.7, B125.1.

5. The Concept of the Universal in Some Later pre-Platonic Cosmologists

137

Not limiting himself to this global characterization, he frequently sorts out the “mortal beings” into sub-genera: “human beings” (further subdivided into “men” and “women”), “beasts,” “birds,” “fish,” “trees,” “shrubs.” 19 Of course, the fact that we have in Empedocles classificatory uses of eidos and related terms does not by itself satisfy the demand for thematic significance I have stipulated. If, however, we follow the majority of interpreters in assuming that in Empedocles’ cosmology we have a cosmic cycle, the distinct phases of which are indefinitely or eternally repeated, then the issue of eidea and ethnea becomes philosophically momentous. We saw earlier how cyclical phenomena may serve to block awareness of types by allowing that different tokens of the same type should be conceptualized as recurring advents of the same token (e.g., last year’s and this year’s summer). In Empedocles’ scheme, most significantly, there is no possibility of such blockage. For, with the possible exception of transmigrating daimones, the vast majority of individuals (the tokens corresponding to the types) do not survive through the phase of maximum Love (the sphairos), and probably also not through that of maximum Strife. What recurs are unmistakably types. Theoretically scrutinizing any two complete cycles M and N, out of the entirety of elapsed cycles, any hearer or reader of Empedocles’ verses would have been prompted – as today any reader would likewise be prompted – to wonder, “Are the eidea of cycle M the same as those in the corresponding phase of cycle N?” Indeed, even within a single cycle – assuming that cycles are counted from total Love (the sphairos) to next total Love – the question quite naturally arises, “Are the eidea that are temporally separated by the intervening period of maximum Strife the same or different?” IV. PHILOLAUS: GENUS, SPECIES, AND THE RELATION TO PARTICULARS With respect to my second of the three antecedents acknowledged by Aristotle, it is important to stress that Aristotle speaks of “Pythagoreans,” or even of “ones who call themselves Pythagoreans,” not of Pythagoras. It is certainly not clear whether the reference is to Pythagoreans of the early fifth or the late fifth century, or, for that matter, of the fourth century. Indeed, the most penetrating and critically balanced account of the subject of Pythagoras and Pythagoreanism in modern scholarship, Walter Burkert’s magisterial book of 1962 (original German edition), would lead one to surmise that Pythagoras’ contribution to the development of the concept of the universal would have been, at most, a proclamation of mystical faith in universal affinity. Even the much-discussed discovery of the arithmetical ratios for the musical concords was probably not, 19

See these fragments (all DK31B): 9, 20, 21, 23, 26, 76, 77-80, 117, 127, 130.

138

Alexander P. D. Mourelatos

according to Burkert (1972, esp. chs. 5 and 6), one made exclusively by Pythagoras’ successors – let alone by the founder of the school. If we limit our purview to this later period, we may appreciate with some confidence the contributions made by Philolaus of Croton, certain important fragments of whose writings are now securely acknowledged as authentic. It is especially significant that Philolaus names his cosmological principles not apeiron and peras but (using the plural) apeira and perainonta, “things unlimited” and “limiting things”/“limiters” (DK44B1, B2, B6), thus openly proclaiming his purpose to classify ta eonta panta, “all existing things,” under these two broad types (see Huffman 1993, pp. 37-53, 93, 101). A conceptually more articulate typology, together with some notably sophisticated distinctions, is implied in a context of narrower scope, the domain of numbers. The viewpoint, nonetheless, is equally abstract as in Philolaus’ statement of cosmological principles, and the implied metaphysical bearing is no less fundamental: Number, indeed, has two proper kinds (LGLDHLGƝ), odd and even, and a third from both mixed together, the even-odd. Of each of the two kinds there are many forms (morphai), of which [scil., the forms] each thing, in and of itself, provides a sign (KHNDVWRQDXWRVƝPDLQHL). (DK44B5) 20

We need not be side-tracked by the vexed question of the nature and scope of the derivative “third” type, the “even-odd”; 21 for it is clear from the second sentence that the focus is on the two kinds that were just denominated “proper” or “intrinsic” (idia). So read, the fragment offers a statement of major importance, as has rightly been emphasized by Carl Huffman: Philolaus here speaks not only of a relation between numbers as individual species and their two genera, even and odd, but also of “the relation between numbers and things” (1993, p. 178). In the terminology used in the present study, Philolaus takes note both of a type1-type2 relation and also of the relation between individual tokens and their appropriate type. He is contrasting the mediated relation that obtains between, say, four pebbles and the type picked out by the expression “even” with the more immediate relation of signification that holds between the tetrad of pebbles and the type picked out by “four.” In the latter case, the physical tetrad “in and of itself” (hekaston auto) is a token of the type – indeed, “betokens” would be an excellent alternative translation for VƝPDLQHL.

20

21

Except for slight variation in the translation of the final three words, text and translation is as in Huffman 1993, p. 178; see also pp. 184-93. I agree with Huffman that the reading Į੝IJĮȣIJં (Doric reflexive pronoun) of the MSS needs to be emended to simply Į੝IJં. Huffman has pointed out to me, however (in private correspondence), that Į੝IJĮȣIJં could be retained, provided we insert įȚ(੺), with elision, to read įȚ’ Į੝IJĮȣIJં, and this might help explain the origin of the initial delta in the nonsensical įȘȝĮ઀ȞİȚ (emended by Huffman and other editors into ıȘȝĮ઀ȞİȚ). The solution in Huffman 1993, p. 190, is plausible enough: “[T]he even-odd is a derived class of numbers whose first member is, as the ancient tradition indicates, the one, but which also includes . . . even and odd numbers combined in ratios (e.g. 2:1, 4:3, and 3:2).”

5. The Concept of the Universal in Some Later pre-Platonic Cosmologists

139

By contrast, in the former case, the relation of the particular to the eidos referred to as “even” is mediated by the morpKƝ referred to as “four.” So explicated, Philolaus B5 gains even more in significance. Its phrasing shows Philolaus groping for a metaphor that may convey the relation between a particular and a universal. Unmistakably it is in this text, not in Plato, that we have the first recorded attempt in the history of metaphysics at coining a term for this relation. Philolaus opts not for “participates in” or “shares in” or “imitates” or “exemplifies” but rather for VƝPDLQHL, “provides a sign of,” “betokens” – remarkably, an expression that is a close congener of the modern term “instantiates.” One is naturally led to wonder how differently the history proper of the universal may have unfolded had it started with philosophical reflection on the relation conveyed by Philolaus’ VƝPDLQHLQ rather than with the other metaphors cited above – the ones introduced and favored by Plato. V. DEMOCRITUS: AN INFINITY OF ATOMIC TYPES, ATOMIC TOKENS I turn now to the third case of anticipation Aristotle recognizes. The philosophy of Democritus 22 was the subject of a lost treatise by Aristotle; 23 and it is clear from the preserved treatises of the Aristotelian corpus that Aristotle viewed Democritus as a major philosophical rival. Aristotle’s own testimony belies his grudging comments on the topic of the universal (“Democritus just touched on it slightly . . . roughly defined the hot and the cold”: text cited above at note 17). For it is Aristotle himself who tells us that Democritean atoms are “infinite in multitude . . . and have all sorts of forms (pantoias morphas) and all sorts of shapes (VFKƝPDWD SDQWRLD) and differences in size” (Aristotle fr. 208, cf. DK68A37). And it is Aristotle who, in other passages, leaves no doubt that “all sorts” must be given the strongest possible reading, in other words, that Democritus envisaged infinity in both the relevant respects: The indivisible bodies are infinite both in multitude and in the shapes. (GC I.1.314a22, cf. DK67A9) Since [ordinary] bodies differ in shape, and the shapes are infinite, then the simple bodies as well are infinite. (De caelo III.4.303a11, cf. DK67A15)

Especially striking is the inference from the infinity of types of shape to the infinity of atoms: the range of shapes is intrinsically infinite; there must be at least one specimen for each type (Democritus was no believer in purely ideal 22

It is, by and large, not possible to distinguish between the contributions to the original philosophy of Atomism made by the founder of the theory, Leucippus, and those made by his successor Democritus: see McKirahan 1994, p. 304; Taylor 1999, pp. 157-58. I adhere to the widely adopted practice of using the name “Democritus” in synecdoche for the sayings and doctrines of both Leucippus and Democritus. 23 Perhaps two treatises, “On Democritus” and “Problems from Democritus.” See DK68A34.

140

Alexander P. D. Mourelatos

possibilities); therefore, the number of atoms bearing shape must be infinite. The inference here could not have been drawn without a firm grasp of the typetoken distinction. The degree of insight into that distinction these reports assign to Democritus compares favorably with what may be assigned not only to corresponding passages in Epicurus and Lucretius 24 but also to the many similar contexts in Plato and in Aristotle. Moreover, it is likely that Democritus went further: he posited an infinity of token atoms for each of the infinite types of shape. This is implied by testimony that presents the early Atomists as having recourse to a version of the Principle of Sufficient Reason, the so-called argument from ou mallon, “no more this than that”: They [Leucippus and Democritus] say that the multitude of shapes in the atoms is infinite since there is no more reason (GLDWRPƝGHQPDOORQ) that [a given atom] should have this shape rather than another (WRLRXWRQƝWRLRXWRQ einai). (Simplicius In Phys. 28.15 ff. = DK68A38)

As in the De caelo text cited earlier, the inference does not go beyond establishing the infinity of types of shape. But if the ou mallon argument works to produce this inference, it should also work to establish that the number of token atoms for any one of the infinity of types of atomic shape is infinite. 25 Furthermore, relying on the same argument, it may well seem reasonable to strengthen Aristotle’s “all sorts of sizes” into “an infinity of atomic sizes,” and then proceed to the corollary, that with respect also to any one size-type there is an infinite supply of token atoms. 26 In any event, just as there is an infinity of types with respect to at least one of the two properties possessed by the atoms intrinsically (shape, size), there is a corresponding immense multitude 24

25

26

Cf. Epicurus Letter to Herodotus 42-43; Lucretius II.478-531. The Epicureans’ doctrine of minimal parts dictated adjusting Democritean doctrine: the token atoms are infinite; but the types of atomic shape are finite, though vastly numerous, DSHULOƝSWD, “unencompassable.” The inference is indeed drawn by the Epicureans for each of the types of shape – in this respect probably in adherence to Democritean doctrine – even though the number of atomic shapes is finite: see preceding note. The evidence is, admittedly, ambiguous. See Taylor 1999, pp. 173-75. But the emphasis placed by the Epicureans on limiting the variety of atomic sizes does perhaps indicate that Democritus, by contrast, had envisaged such an infinity. It is sometimes assumed – and I so once assumed myself – that any infinite variation of size would have to obtain between a lower and an upper bound (smallest and largest atomic size). The positing of any such bounds would be unfortunate; for it would significantly qualify Democritus’ adherence to the ou mallon principle. But (as Victor Caston and others have pointed out to me in discussion), the unqualified infinity of the Democritean universe could well allow for finite atomic sizes that recede or proceed ad infinitum to smaller-and-smaller and to larger-and-larger, respectively. The Like to Like principle would guarantee that kosmoi, “worlds” (which in all ancient atomist cosmologies are large-scale regional clusterings of atomic compounds) would involve only compatible atomic sizes – ones that fall within a range appropriate for the dimensions of the kosmos at issue. To be sure, if atoms are to be thought of as indivisible magnitudes in a strong sense (metaphysically or mathematically, not just physically), then there would have to be a lower bound atomic size. But I strongly doubt that Democritus would have entertained the self-contradictory concept of ideai, “shapes,” that are indivisible.

5. The Concept of the Universal in Some Later pre-Platonic Cosmologists

141

(perhaps an infinity) of types for what might be called extrinsic or relational properties, those acquired by atoms after collision and rebound, or in the course of aggregation into compounds: motion (including the differentiae of speed and direction); taxis, “array, order, position” (the difference between AN and NA); and thesis, “posture, tilt, orientation” (the difference between N and Z). 27 Finally, at the widest compass, in the infinite expanse of the void there is an infinity of kosmoi, “world structures” (DK67A1, A21, 68A1[44]). Democritus may well appear intoxicated by the theme of infinity, obsessed with it. In the overall ramified scheme, we have five superordinate types (the five distinct intrinsic or extrinsic properties of atoms cited above: shape, size, motion, array, posture); for at least the first of these there is an infinity of subordinate types of that property; and for each of the subordinate types (be they infinite or finite) there is an infinity of token atoms. One of our ancient sources comments, with evident irritation, that “it is a consequence of their [Leucippus’ and Democritus’] theories that there is something more infinite than the infinite.” 28 What drives this exuberant proliferation of entities – here is the crucial point for the present account – are the two distinctions: type-token and type1- type2. Given the important work these distinctions accomplish within Democritus’ system, it would not have required too difficult an act of abstraction for him to focus on the infinite domain of any one of the five super-ordinate types cited here. Certainly in the case of atomic shape, we have evidence that he did approach the subject abstractly and typologically. For he spoke not only globally of an infinity of shapes; he took note also of such sub-genera as “convex,” “concave,” “lop-sided,” “curviform,” “spherical,” “angular,” “polygonal,” “hook-like”; 29 and he wrote treatises titled 3HUL GLDSKHURQWǀQ UK\VPǀQ, “On Different Shapes,” and 3HUL DPHLSVLUK\VPLǀQ, “On Changes of Shape” (DK68A33), as well as several treatises on topics of geometry – five or six out of the twelve listed under PDWKƝPDWLND at DK68A33. VI. COMMENTS BY DEMOCRITUS ON THE UNIVERSAL My argument so far has been that the uses of the type-token and type1- type2 distinctions by Democritus have enough thematic significance to justify the claim that this philosopher has a well-articulated grasp of the concept of the universal. But I believe an even stronger case can be made: there is good evidence that Democritus not only made use of the concept of the universal but 27

Aristotle Metaph. I.4.985b4-21; DK67A6. I deliberately select the Aristotelian terms over the Democritean GLDWKLJƝ and WURSƝ which refer, I believe, to the dispositions that correspond, respectively, to the two occurrent and acquired properties of taxis and thesis. 28 John Philoponus commenting on Aristotle GC I.1.314a15. Cf. Taylor (1999), p. 73. 29 See DK68A37, A101-102, A132, A135. Cf. Stückelberger (1979), pp. 319-20.

142

Alexander P. D. Mourelatos

also that he “thematized” it, that he made the concept itself a topic for explanation and comment. The evidence is in two seldom discussed fragments. The next Chapter in the present collection offers a philological analysis of these two fragments, as well as of the contexts in the ancient sources in which these two quite striking quotations from Democritus have been preserved. Here I limit myself to reporting the conclusions of that analysis. The first of these fragments is B124. The part in which Democritus is quoted reads: ਙȞșȡȦʌȠȚİੇȢ਩ıIJĮȚ țĮ੿ ਙȞșȡȦʌȠȢʌȐȞIJİȢ Human beings will be one [scil. “human being”] and (they) all (= everyone) will be (undrstood esontai or eisi) a human being.

The source is pseudo-Galen Medical Definitions (439, Kühn 1830), and the theme in the paragraph in which the quotation occurs is history of the genetic or embryological doctrine of pangenesis, the doctrine that seed is drawn from the whole of the body of the father or (whole of the body of each) of the parents. But remarks attributed to Presocratics in later sources are often placed in contexts quite different from those in which the remark at issue may have originally been made. My proposal is that the original context was metaphysical. The future form estai does not envisage embryological development; it is a futurum consequentiae (cf. “Given what was said, this will be the case . . . ”). Moreover, the grammatical subject is not heis, “one,” but rather DQWKUǀSRL, “human beings” – and this construe is the more natural one, given the word order. 30 My metaphysical reading of B124 can, accordingly, be conveyed in this paraphrase: If one adopts the synoptic view, the many human beings will be (may be viewed as) a single human being – the type Human Being. Nonetheless, the many human beings, taken severally, as separate individuals (tokens), will each count as (or simply “is”) a human being.

If this interpretation is correct, Democritus did not only conform to the third of the criteria listed above in Section I of the present Chapter; he actually stated that criterion. In effect, he intuitively anticipates the solution to Plato’s istion, “awning,” argument in the Parmenides: the “one” that is collected in a universal is not a “one over many” in the way an awning extends over many spectators. Discernible in the Democritean saying are the following two points. (a) As we look for the universal, the plural individual human beings will come to be viewed as one, the type Human Being. (b) This, however, does not mean that Human Being is a partitive entity, one that can be distributed mereologi30

Cf. Kahn 1973, p. 427: “There seems to be no doubt that the statistically favored order, both in Homer and in classic prose, is . . . N [subject] F [predicate] is.”

5. The Concept of the Universal in Some Later pre-Platonic Cosmologists

143

cally; rather, each and every human being is in himself or herself fully and exhaustively “a human being.” The other Democritean saying that shows, I believe, our philosopher attending to the concept of the universal is in B165, which in DK is printed as a fragment in two parts, joined by a dash. In the source text, in Sextus Empiricus, the two parts are separated by two intervening sentences of Sextus’ own comments. My interest, like that of other scholars who have discussed this fragment, is only in the second part: ਙȞșȡȦʌȩȢ ਥıIJȚ ੔ ʌȐȞIJİȢ ੅įȝİȞ. Human Being is what we all know. (B165) My proposal is that the saying combines epistemological and metaphysical import. It introduces and explains the universal implied in DQWKUǀSRV, “Human Being,” by pointing out a disparity between our knowledge of universals and our knowledge of particulars. There is no way that all human beings, certainly not across all the ages, but not even (assuming ancient modes of transport) over some limited stretch of time, should know one single individual – say, Democritus, or Leucippus, or Protagoras, or the present king of Persia. And yet we all know what it is to be human, we know the universal DQWKUǀSRV. It is not at all difficult to imagine what circumstances might have prompted Democritus to make this comment. Other fragments show him contrasting the clarity and security with which we know general and fundamental truths about his two fundamental entities, atoms and the void, vis-à-vis the obscurity that is inherent in our knowledge of hekasta, “things in particular”: Even so (kaitoi), it will become evident that it is not possible to know what each particular thing is (hoion hekaston) in reality (HWHƝL). (B8) Surely, on the one hand (men), it has been shown in many ways that, in terms of reality (HWHƝL), we do not grasp what each particular thing (hoion hekaston) is or is not . . . (B10)

But the example of DQWKUǀSRV in B165 may also have served the same purpose the example of the alphabet had served in Galen (above, Section I). If a critic had objected that concrete particulars are more easily knowable than types, B165 would have provided an effective rejoinder. VII. DEMOCRITUS AND ARISTOTLE: ORIGINS OF THE TYPE-TOKEN DISTINCTION Emerging from the preceding two Sections is a historical diagnosis that is absent in standard accounts of the Presocratics. It would appear that in Democritus we have an anticipation of the Aristotelian solution to the metaphysi-

144

Alexander P. D. Mourelatos

cal problem of the status of universals: universalia in rebus. 31 But if this is right, then Aristotle’s use of the alphabet example in Metaph. XIII.10 (above, Section I) gains enormously in significance. For it is also in another passage in Aristotle, at Metaph. I.4.985b4-21, that we find the suggestion that Democritus had used the alphabet example in order to illustrate the basic differences in the fundamental properties of the atoms – VFKƝPD, “shape,” taxis, “array, order,” and thesis, “posture, tilt.” Since the Greek word for letter, stoicheion, is the one that is ultimately adopted (clearly by Plato’s time) as the term for metaphysical “elements,” and Democritus pre-eminently and famously has a metaphysics of elements, there is no reason to doubt that the example Aristotle used came from Democritus himself. And once this connection is made, it is difficult to resist the further inference that Democritus’ choice of the alphabet example made it possible for him also to illustrate the distinction which, as we saw in Section V, is absolutely essential for formulating the doctrine of dual pluralities and infinities: the distinction between types and tokens. Just as the letter alpha can refer either to a token or to a type, so too “cube” can refer either to a particular atom or to a type of atomic shape. And, more broadly, ideai may be, as we saw in Section V, either types of atomic shape or atoms viewed individually, token atoms. One may wonder whether there is a historical chain that binds Galen’s use of the type-token distinction with its counterpart in Peirce. In the case of the ancients, at any rate, there is good reason to suppose that we do have such a chain. The chain’s later links show up in Galen and in Syrianus; the chain certainly runs through Aristotle; and the chain’s start and mooring is in Democritus. VIII. DEMOCRITUS AND PLATO In Section V we saw evidence that Democritus posited a domain of infinitely various types of shape. This could be envisaged as an enormous upside-down typological tree, with prodigious branchings and sub-branchings – in effect, a hierarchical inventory of all possible shapes, the entirety of the conceptual matrix of shape. Naturally, one is drawn to a comparison with the Platonic universe of ideai, “forms.” That universe is vastly greater in compass than Democritus’ inventory of the totality of ideai, “shapes.” Nonetheless, the special region of Plato’s ideal universe which is allocated to shape has exactly the same content and the same logical structure as Democritus’ hierarchical inventory.

31

The alternative, that he might have intuitively assumed some version of nominalism, is quite unlikely. For there would have been little point in Democritus’ insistence on the actual infinity of shape-types if all he posited as present in reality were gross resemblances between atom-tokens.

5. The Concept of the Universal in Some Later pre-Platonic Cosmologists

145

Plato’s grand metaphor in the Sophist of a battle between “the Gods and the Giants” (the idealist friends of the Forms and the materialists, 246A-C), has given rise to simplistic, almost hackneyed, contrasting of Democritus and Plato in histories of philosophy. And yet, as is suggested by the remarkable fact that the term idea plays a key role in the system of both philosophers, there are also striking affinities, both on the theme of the importance of form as structure or shape and – as the present Chapter has sought to demonstrate – in their respective approaches to the concept of the universal. 32 To be sure, there are also radical and stark differences. What for Plato is a universe of ideal possibilities (e.g., all the possible shapes) is for Democritus an actual universe of realized atomic shapes. For it would be remiss of us not notice that the exuberant application of the ou mallon argument has the effect of making Democritus an adherent not only of the Principle of Sufficient Reason but also of Lovejoy’s aptly named Principle of Plenitude (1936, p. 52, and generally his ch. 2). Oher adherents to this second principle envisage that all possibilities are realized eventually, in infinite time. For Democritus, all possibilities are realized at every moment – if the whole of the spatially infinite universe is taken in consideration. Any of the types of atomic shapes which at some particular time are represented by no tokens at all in some particular kosmos, “world structure” (e.g., our world, here and now) are so represented at that same time in other kosmoi – elsewhere in the infinite expanse of the void. 33 Since Democritus not only outlived Socrates but may have still been active philosophically into the first ten or twenty years of Plato’s philosophical career, the trenchant saying in B165 (“Man is what we all know”) may have also served as a vehicle for conveying the difference Democritus himself perceived between his concept of type – and, more pointedly, of idea, “shape,” as a type – and the more exalted concept of the Platonic idea as a sublime exemplar. The crucial point in the present account is that, so far as the logic of the concept of the universal is concerned, Plato would not have had to teach Democritus much – certainly not what he presents Socrates as explaining to Meno about the logic of shape. 34 32

For affinities on topics other than the ones taken up in the present study, see ȆĮȞȑȡȘȢ,1984, esp. pp. 80-81; also Nikolaou 1998, esp. pp. 140-52. 33 See Taylor 1999, pp. 94, 96, 197. For the distinction between kosmos and universe (to pan), see Furley 1987, p. 136. 34 Earlier versions of this Chapter were presented as follows. In Greek: Academy of Athens; University of Crete; University of Patras. English versions were presented at: University of Aarhus; University of Edinburgh; Florida State University; Texas A & M University; Central European University, Hungary; University of Utrecht; Center for Hellenic Studies, Washington, D.C.; Marquette University. – I am grateful to Johanna Seibt of Aarhus University for her elegant German translation, which made it possible for me also to present the work at the University of Bern, and at Freie Universität Berlin. – I am indebted to the respective audiences for their input, and in particular to Gianluigi Segalerba and Gerald Bechtle.

6.

Democritus on the Distinction Between Universals and Particulars: the Relevance of Fragments B124 and B165

Andreas Graeser in memoriam

The article “Allgemeines” in Bächli and Graeser’s very useful and admirably concise dictionary of Grundbegriffe der antiken Philosophie (2000) starts with Aristotle’s distinction between universals and particulars (ta katholou, “things in general”, and ta kath’ hekaston, “things in particular”). The authors immediately focus on three issues: (1) What are universals? (2) Do universals exist in reality? (3) What is the relation between universals and particulars? Since it is for Plato and for Aristotle that these issues are paramount, Bächli and Graeser do not discuss antecedents to the concept of the universal in Presocratic philosophy. I believe the preceding Chapter has shown that the distinction is certainly grasped and thematically utilized by several thinkers in the second half of the fifth century, in the later stages of early Greek philosophy. In the present Chapter I seek to amplify and reinforce the evidence that Democritus, quite apart from deploying the type/token distinction at the most fundamental level of his metaphysics, may also have shown some reflective or meta-theoretical awareness of the distinction. Specifically, I believe he does this in two intriguing fragments, B124 and B165, which are generally bypassed by philosophically-minded readers inasmuch as the first is taken to bear on genetics or embryology and the second is generally given a minimalist or reductivist interpretation. In the pages that follow there will be much discussion of the term antKUǀSRV, -oi, “human being(s).” I shall be leaving occurrences of the latter mostly without translation. I do so because of the now well-established avoidance in English of “man/men” as the “unmarked” or “default” use. Greek does, of course, treat the grammatically masculine form generally as the unmarked or default. But with DQWKUǀSRV, -oi we have the advantage of a common noun which in its morphological root and root meaning is gender-neutral. In English,

First published under the title, “Democritus on the Distinction Between Universals and Particulars,” in Andreas Bächli and Klaus Petrus, edd., Monism, Philosophische Analyse 9, Frankfurt and London: Ontos Verlag, 2003, pp. 43-56. It is reprinted here with the kind permission of Walter de Gruyter.

6. Democritus on the Distinction Between Universals and Particulars

147

by contrast, uses of the corresponding term “human” not as adjective but as common noun (as in “one human”/ “many humans”) still sound harsh and are avoided in favor of “one person”/ “many persons” or “people.” In any event, with the currency in English of derivatives such as “anthropology/ anthropic/ philanthropic/ anthropo-morphic,” the transliterated term should pose not problem to Greekless readers. I In pseudo-Galen Medical Definitions 439 (Kühn 1830) we read: ਫțțȡȓȞİIJĮȚ IJઁ ıʌȑȡȝĮ, ੮ıʌİȡ ȆȜȐIJȦȞ ijȘı੿ țĮ੿ ǻȚȠțȜોȢ, ਕʌઁ ਥȖțİijȐȜȠȣ țĮ੿ ȞȦIJȚĮȓȠȣ. ȆȡȦIJĮȖȩȡĮȢ į੻ țĮ੿ ǻȘȝȩțȡȚIJȠȢ ਩IJȚ IJİ ੊ʌʌȠțȡȐIJȘȢ ਥȟ ੖ȜȠȣ IJȠ૨ ıȫȝĮIJȠȢ੘ ȝ੻ȞǻȘȝȩțȡȚIJȠȢȜȑȖȦȞਙȞșȡȦʌȠȚİੈȢ਩ıIJĮȚțĮ੿ ਙȞșȡȦʌȠȢ ʌȐȞIJİȢ. ੘ į੻ ੊ʌʌȠțȡȐIJȘȢijȘıȓ . . . Seed is secreted, as Plato and Diocles say, from the brain and the spinal cord. But Praxagoras and Democritus, as well as Hippocrates, [hold that it is secreted] from the whole body – In the case of Democritus, inasmuch as he says, “one will be DQWKUǀSRL [plural] and [they] all will be DQWKUǀSRV [singular].” Hippocrates, however, says . . . . 1

In the above extract, the underlined text is generally listed as a Democritus fragment: B124 in Diels-Kranz (DK henceforth). There should be no doubt concerning the validity of the framing pseudo-Galen testimony, scil., that Democritus held that “seed is secreted from the whole body” – what in the history of genetics is now generally referred to as the “pangenesis” theory. The attribution of pangenesis to Democritus is independently attested in the Aëtius compendium (DK68A141); and the doctrine does comport well enough with other themes in Democritus’ natural philosophy. 2 Indeed, and quite interestingly, it has been argued that Aristotle’s attack on pangenesis in book I of Generation of Animals is in the main directed against Democritus, in spite of the fact that this dialectical opponent is not explicitly mentioned in this connection till book IV (3.769a17-19). 3 There are, however, strong doubts about the soundness of the text after 'ƝPRNULWRV OHJǀQ, “Democritus . . . says” (see the underlined text above). For it is not at all clear how what has been preserved in the MSS can be at all plausibly or naturally read as proclaiming the pangenesis doctrine.

1

Note that the translation given above is the one commonly offered: it assumes that heis is the subject and DQWKUǀSRL is syntactically the predicate complement. Shortly below I argue for the reverse construe. The citation of B124 in the preceding Chapter has given already the translation I prefer. 2 Taylor 1999, p. 198. 3 See De Ley 1980, pp. 130 ff. Cf. Morel 1996, pp. 157-59; Salem 1996, p. 235.

148

Alexander P. D. Mourelatos

The notes ad loc. in DK record the following proposal by Heinrich Gomperz: the entire body of the father is constituted of a huge number of homunculi; a small subset of these are carried off as sperm; they subsequently (that’s presumably how Gomperz reads the future estai) coalesce in the womb to form a new colony of homunculi, a new human being. This rather too imaginative proposal has found no takers; 4 and since the fragment is often bypassed in discussions of Democritus, it would appear that most scholars have acquiesced in Diels’ suspicion that B124 is corrupt and should be emended. Diels’ own conjecture is ਙȞșȡȦʌȠȢ ਥȟȑııȣIJĮȚ ਥȟ ਕȞșȡȫʌȠȣ ʌĮȞIJȩȢ. The restored text should, accordingly, be taken either as a variation on, or as corruption of, part of B32: ȟȣȞȠȣıȓȘਕʌȠʌȜȘȟȓȘȝȚțȡȒāਥȟȑııȣIJĮȚȖ੹ȡਙȞșȡȦʌȠȢਥȟਕȞșȡȫʌȠȣ [Orgasm in] sexual intercourse is a minor apoplexy [stroke, seizure]: an DQWKUǀSRVhas sprung forth out of an DQWKUǀSRV.

Early support for the remedy of emendation was provided by Rau (1923), who, however, opts for a somewhat different reconstruction: ਙȞșȡȦʌȠȢਥȟਕȞșȡȫʌȠȣਥțıİȓİIJĮȚțĮ੿ ਕʌȠıʌ઼IJĮȚ. An DQWKUǀSRV is shaken out of, and is torn off, another DQWKUǀSRV.

Rau makes the quite germane observation that DQWKUǀSRV SDQWHV is palaeographically rather close to apospatai. For, given standard late-antiquity abbreviations, and in uncials, the former would have been written: ANOCȆANTEC (compare AȆOCȆATAI).

Both of these proposals adhere to tenets of philological practice that lost favor in the middle and late decades of the twentieth century. In the philology of today and of at least the last two generations, proposals for emendation must pass a stringent test of necessity. The proposal should be considered viable only if all plausible alternatives have been eliminated. There has been no such elimination of alternatives in the case of Democritus B124. To begin with, the literary flavor of the received text is strikingly that of Democritean apothegm. Overlaid on the main doubling (anadiplosis) of the noun-pronoun pair (antKUǀSRL –“one”; DQWKUǀSRV –“all”) are four other rhetorical effects: (1) Chiasmus-like (ABBA) alternation in the grammatical number: plural (DQWKUǀSRL)/ singular (“one”); singular (DQWKUǀSRV)/ plural (“they all”). (2) Poignantly dissonant associations: the many is singular; the one is plural.

4

Cf. Lur’e 1970, pp. 130, 180.

6. Democritus on the Distinction Between Universals and Particulars

149

(3) Paired contrasts: DQWKUǀSRL vs. DQWKUǀSRV ; one vs. all. (4) The consonance of the trio of syllables, -poi, -pos, pan- (DQWKUǀpoi, DQWKUǀpos, pan-tes), interwoven with the consonance of the other trio (monosyllabic heis, plus two es- syllables), (heis, estai, pantes).

And as for palaeographic considerations, the striking fact is that it must have taken enormous discipline on the part of many generations of copyists to preserve the syntactically incongruous alternation of singular and plural forms ਙȞșȡȦʌȠȚ İੈȢ ਩ıIJĮȚ țĮ੿ ਙȞșȡȦʌȠȢ ʌȐȞIJİȢ. The temptation to regularize this conundrum into ਙȞșȡȦʌȠȢ (singular) İੈȢ ਩ıIJĮȚ țĮ੿ ਙȞșȡȦʌȠȚ (plural) ʌȐȞIJİȢ – sliding into a lectio facilior – must have been enormous. This suggests that the Democritean apothegm had been widely memorialized. Its fame or notoriety thus had worked as an effective preservative for its ostensibly paradoxical syntax. What has made Gomperz’s proposal appear gratuitous is that both he and his critics assumed that the meaning of the fragment in the pseudo-Galenic context must have been the same as the meaning of the fragment in its original Democritean setting. No doubt, it is methodologically sound to scrutinize the context in which a Presocratic saying has been embedded in later sources so as to find clues concerning the likely context of the original statement. But there are innumerable known cases in which the context in the later source is wildly different from that in the original; and apothegmatic utterances lend themselves rather easily to thematic transformation. Specifically in the case of Defin. med. it needs to be observed that allusions to the doctrines of the philosophers are often rather tenuously connected to the medical topics that are the subject of “definitions”. Testimonia and quotations are quite brief, and they seem to serve more as displays of erudition. For example, the section on bodily fluids (61-74) concludes with the physiologically irrelevant generalization that the “motions within us are the six distinguished in Aristotle’s Categories”. A hypothesis that shows some charity toward the author of Defin. med. is that the original Democritean context was broadly embryological, though it was not specifically a statement of Democritus’ version of the pangenesis theory. Evidently Democritus wrote a treatise with the title Peri physios DQWKUǀSRX, “On the nature of DQWKUǀSRV” (see below, Section II), and perhaps B124 appeared somewhere in that work. In fact, this is precisely how B124 is interpreted in the article by B. ten Brink (1853), which is the only study specifically devoted to reconstructing the contents of Democritus’ anthropological treatise. We are offered this paraphrase of B124: omnes homines unius hominis referunt naturam, unus autem omnium, “all human beings reflect the nature of one human being, whereas one reflects that of all” (p. 415). The paraphrase takes rather too many liberties with the text. Charitable though it may be to the author of Defin. med., this exegetic hypothesis is hardly charitable to Democritus. For it seems to be a

150

Alexander P. D. Mourelatos

trivial truth of (pre-Darwinian) genealogy that if many human beings are descended from a single human being, the whole of human progeny will likewise be human beings. At best, Democritus would be presented as offering a needlessly wordy version of Aristotle’s principle, DQWKUǀSRVDQWKUǀSRQJHQQƗL The more likely and attractive hypothesis is that the author of Defin. med. appropriated for his purposes a saying of Democritus that was not specifically embryological. The future tense estai need not have been developmental; it could have been a futurum consequentiae, a use that has parallels in other Democritean fragments, e.g., B8 “Nonetheless [given certain assumptions, or on the basis of an argument given or projected] it will become clear (GƝORQHVWDL) that there is no way to know what each thing is.” 5 I propose that we explore the possibility that the original context was metaphysical. Let us examine the syntax of each of the two parts of the saying. In the first part, we have estai in the singular, which strongly suggests that we are to take heis, “one”, as subject, whereas DQWKUǀSRL is to serve as predicate complement. In the second half, however, it is surely the pronoun pantes, “all”, that serves as subject; and DQWKUǀSRV must be the predicate complement. Indeed, if in that second half of B124 we were to insist on having the singular DQWKUǀSRV as subject, so as to retain an understood estai, “will be (singular)”, as the copula, the second part becomes indistinguishable in sense from the first: “(one) DQWKUǀSRV (will be) all (pantes)” – a sheer redundancy. The inescapable inference is that we have a slight anacoluthon in the transition from the first to the second part: the understood copula in the second part is no longer the singular estai of the first part but either esontai, “will be (plural)”, or simply eisi, “are”. Of the two parts, the second is presumably the less paradoxical one. For it could be understood as: “all will (count as) human”; or “any will (count as) a human being”; or “all are human”. It is the first part that is the stumper: “One DQWKUǀSRV will be (many) DQWKUǀSRL”. We think immediately of the Heraclitean parallel: “one (heis, i.e., KHLV DQWKUǀSRV) is for me as good as ten thousand (myrioi), if he is the best (aristos)” (DK22B49). 6 The first part of the Democritean saying, by contrast, is far more enigmatic and provocative – more so than either the Heraclitean fragment or the second part of Democritus B124. The key to solving the enigma, I believe, is in the switch from estai, “will be (singular)”, to either esontai, “will be (plural)”, or eisi, “are”, in the second part. Given the anacoluthon, we need not necessarily assume that the sense of the verb “to be” in the two parts is identical. The pronominal subject “all” in the second part makes a predicative construe of that part inevitable: “all will 5

At B155, the perplexing logical consequences (for Democritean mathematics) of making a cut on a cone immediately above the cone’s basis are likewise expressed with futures (parexousi, estai, phaneitai). 6 Among the fragments of Democritus listed as “spurious” in DK, there appears to be a loose version of B124 that may have arisen through conflation with Heraclitus B49: Democritus ait: unus mihi pro populo est et populus pro uno (Seneca, Epist. 7.10; DK68B302a). Cf. Taylor 1999, p. 7.

6. Democritus on the Distinction Between Universals and Particulars

151

be/are [of the kind] human being.” But in the first part the logical structure is rather more suggestively that of identity – and the Heraclitean parallel certainly supports this construe – “One (DQWKUǀSRV) = (many) DQWKUǀSRL.” It is worth noting immediately that if the copula here has the sense of identity, the rule that verb-number agrees with subject-number need not apply: “DQWKUǀSRL is (identical to) one DQWKUǀSRV” is perfectly possible if a suitable interpretation of this speculative identity statement is forthcoming – and this I am about to show. Without the sequel provided by the second part of B124 one might indeed be tempted to conjecture that the original Democritean saying was political or ethical or epistemological – perhaps another expression of philosophical disdain for “the many”. But taking the plural DQWKUǀSRL as grammatical subject in the first part and putting the two parts together, we have the possibility of a powerful metaphysical reading of B124. Let me convey this in an interpretative paraphrase: The many DQWKUǀSRL, viewed collectively as a class, will be (may be viewed as) a single DQWKUǀSRV – the universal Human Being – just as each of the totality of (pantes) human beings, taken severally, as separate individuals, will by herself or himself count as (or simply “is”) an DQWKUǀSRV

One can see immediately the felicitousness in the use of the singular “is” with plural subject DQWKUǀSRLin the first part of the statement. A heuristically crucial first step toward grasping the universal is to realize, for example, that “DQWKUǀSRL is $17+5ƿ326”. 7 Had Democritus used the plural “are” in the first part, we would not only have missed the striking message, we would also have tautology in the two parts: “All DQWKUǀSRL are human”; and “Each and everyone of the DQWKUǀSRL is an DQWKUǀSRV.” Let me incur the risk of anachronism in order to explicate in terms that may be helpful to a modern reader. My proposal is that Democritus is taking note of the two complementary aspects in grasping the relation between particulars and universals: we realize that there is the common nature, $17+5ƿ326, which is present in all the individual DQWKUǀSRL; but we also realize that the universal is not parceled out mereologically; for it is not parts of $17+5ƿ326 that are apportioned to the individual DQWKUǀSRL; rather, each individual is fully and self-sufficiently an instance of the entirety of the nature $17+5ƿ326. Democritus’ point happens to have been made with use specifically of the example DQWKUǀSRLDQWKUǀSRV. But the metaphysical distinction the example conveys reaches into the deepest levels of Democritean cosmology and meta7

This “first step” does incur the risk (as the late Scott Austin had reminded me) of confusing the type with a mere collection of tokens. But this risk is an inevitable part of the discovery process. – My use of capitalization is intended merely to mark the type. I do not, of course, wish to suggest that Democritus posited transcendent Forms or essences, in the manner in which this is done by Plato or by like-minded metaphysical “realists”.

152

Alexander P. D. Mourelatos

physics. For we might say that the infinity of, e.g., spherical atoms viewed as a class, prompts us to grasp the idea (Democritus’ own technical term for “shape”) of sphere; and yet, each member of that infinite class of spherical atoms, viewed as a token, with no account taken of any others, is all by itself simply a sphere. II Democritus is also reported to have said, ਙȞșȡȦʌȩȢ ਥıIJȚ ੔ ʌȐȞIJİȢ ੅įȝİȞ, “DQWKUǀSRVis what we all know” (second half of B165 in DK). The saying is preserved in two separate contexts in Sextus Empiricus, but it is also implied in remarks concerning Democritus in Aristotle’s Parts of Animals. The tenor of the skeptical argument in both the Sextus passages is that there is no science of DQWKUǀSRV, inasmuch as any attempt to define human nature relies on terms that are either too broad, or question-begging, or otherwise off the mark. I quote here the fuller of the two Sextus passages, as well the relevant passage in Aristotle: Of those who sought the concept [at issue] (ʌİȡ੿ IJોȢ ਥʌȚȞȠȓĮȢ ȗȘIJȘıȐȞIJȦȞ), Socrates ended up in skeptical perplexity. He said that he was ignorant both as to what he himself is and how he relates to things generally (Į੝IJઁȞ ਕȖȞȠİ૙Ȟ IJȓ IJૃ ਩ıIJȚ țĮ੿ ʌ૵Ȣ ਩ȤİȚ ʌȡઁȢ IJઁ ıȪȝʌĮȞ). “For I don’t know,” he says, “whether I am a human being or also some other beast more multiform than Typhon.” But Democritus, whose sayings have been likened to [or who models his sayings on] the pronouncements of Zeus, and who [presents himself as] “saying the following about things in general” (ȜȑȖȦȞ IJȐįİ ʌİȡ੿ IJ૵Ȟ ȟȣȝʌȐȞIJȦȞ), did make an attempt to expound the concept (epinoian ekthesthai). He did not, however, succeed beyond a trivial subjective avowal (LGLǀWLNƝV DSRSKDVHǀV), stating that “DQWKUǀSRV is what we all know” (ਙȞșȡȦʌȩȢ ਥıIJȚ ੔ ʌȐȞIJİȢ ੅įȝİȞ). For, in the first place, dog too is what we all know, but DQWKUǀSRV is not dog; and all of us know horse also, and plant, but DQWKUǀSRV is none of these. Moreover, he has begged the question. For no one will allow that what DQWKUǀSRV is can be known from what is readily available (ek procheirou) – didn’t the Pythian Apollo posit “Know thyself” as a greatest challenge? But even if this knowledge were allowed to be obtainable, it would be so not to all but to the most perspicacious (tois akribestatois) of philosophers. (Sextus Empiricus, Adv. math. VII.264-67; cf. Pyrrh. hyp. II.22-23) [In answer to the “what is it?” question] it is not enough to state the “out of which it is” (HNWLQǀQHVWLQ), say, out of fire or earth. For likewise if we were speaking even of a bed, or of some other artifact, we would be seeking to specify the form of it (WR HLGRV DXWƝV) rather than the matter (Ɲ WƝQ K\OƝQ). . . . Formal nature (țĮIJ੹ IJ੽Ȟ ȝȠȡij੽Ȟ ijȪıȚȢ), surely, is of greater

6. Democritus on the Distinction Between Universals and Particulars

153

import (N\ULǀWHUD) than material nature (K\OLNƝVSK\VHǀV). Indeed, if each of the animals and their parts is what it is by virtue of their shape or color (IJ૶ ıȤȒȝĮIJȚ țĮ੿ IJ૶ ȤȡȫȝĮIJȚ), Democritus would have spoken correctly. For that is how he appears to have understood things: he says that it is evident to everyone what sort of thing DQWKUǀSRV is in terms of form (WƝQ PRUSKƝQ), inasmuch as DQWKUǀSRV is familiar to us in terms of shape and of color But a dead DQWKUǀSRV too has the same form of shape (VFKƝPDWRVPRUSKƝQ), and yet he is not an DQWKUǀSRV. (Aristotle, Parts of Animals, I.1.640b22-35)

The “inasmuch as” clause gives itself away as Aristotle’s own explication; the reference to “color” is obviously intended to emphasize the supposed superficiality of Democritus’ approach; and the terms PRUSKƝ and VFKƝPD are standard in Aristotle for merely prima facie accounts of the formal cause. In both Aristotle and in Sextus the issue Democritus addressed comes across as that of the essence or nature of DQWKUǀSRV at a high level of generality. Sextus takes the Democritean remark to be an attempt at giving a strictly logical definition, “DQWKUǀSRV = df. the X such that we all know X”. This is what permits Sextus’ (transparently captious) critique by means of counterexamples. Aristotle, on the other hand, has a more forgiving formulation of the Democritean doctrine at issue: the form or nature of DQWKUǀSRV is grasped by Democritus in terms of such manifest general characteristics as shape and color, which Aristotle judges insufficient. We have no indication at all of the context in which this anthropological remark of Democritus’ originally occurred. One might initially suppose that it appeared somewhere (possibly at the start) of the treatise “On the nature of DQWKUǀSRV” (mentioned above), which is one of the alternate titles for a work on this subject preserved in the ancient tabulation of Democritus’ works (DK68A33, B5d). Certainly the testimonia give ample evidence that Democritus wrote in detail about human embryology, anatomy, and physiology. 8 In the reconstruction by ten Brink (1853) of the contents of Democritus’ anthropological treatise, we find material that bears on more than twenty-five themes, ranging from embryology through the anatomy of the viscera to the physiology of death. What might have been the point of B165 in the context of such concerns and investigations? The easiest, most straightforward interpretation of the fragment would have Democritus voicing a reductivist, almost Philistine, attitude: there’s nothing mysterious about DQWKUǀSRV; human nature is something manifest; something evident to all. 9 But this is hardly consonant with the range and specialized tenor of the topics covered in the testimonia. And, most pertinently, it would be 8 9

See Lur’e 1970, fragments 522-44, pp. 130-34. Lur’e 1970, p. 430, assumes that DQWKUǀSRV in B165 refers primarily to human sensations and beliefs. Accordingly, he judges that the fragment is very close in its content to Protagoras’ homo-mensura doctrine. I find this unconvincing.

154

Alexander P. D. Mourelatos

hard to reconcile such an attitude with Democritus’ celebrated profession of zeal for finding explanations: “better to find one single explanation rather than attain to the kingship of the Persians” (B118). An intriguing non-reductivist gloss on B165 is offered by Michael of Ephesus in his commentary on the relevant passage from Aristotle’s Parts of Animals: Democritus approached the whole investigation as though it were about the material cause (K\OLNƝVDLWLDV), and he neglected final cause (tinos heneka) or formal cause (HLGLNƝQ), saying: “It is evident to everyone (KDSDQWLGƝORQ) what sort of thing DQWKUǀSRV and each of the animals is in terms of shape and color; it is what they are in terms of matter that is non-evident (țĮIJ੹ į੻ IJ੽Ȟ ੢ȜȘȞ ਙįȘȜȠȞ). But if this is so, then our inquiry ought to be concerned with the non-evident, not with what is most evident (SKDQHUǀWDWRX). (In De part. anim. 640b29)

The passage presented in direct discourse is certainly not a quotation from Democritus; it does little more than rearrange the words found in the corresponding Aristotelian passage. Moreover, it is immediately followed by words that announce another quasi-quotation (“Aristotle objects to this, saying . . .”), and what is then presented, again in direct discourse, is quite transparently a (rather wordy) paraphrase of Parts of Animals 640b29-641a21. In the quasi-quotation from Democritus, 10 after the semicolon, from “it is what they are in terms of matter,” Michael does appear to offer something new; yet even here the implicit Democritean thought is given Peripatetic garb. Michael’s focus is on the Aristotelian contrast between form and matter. His point is that Democritus, 11 having had a simplistic conception of form as the “evident” visual shape, wrongly opted for the “non-evident” matter. But a form/matter contrast is precisely what is not found in Democritus, for whom atoms are inseparably both units of body and units of structure, ideai. 12 In all likelihood, Michael knows nothing more about Democritus’ anthropology than what he gleans from the Aristotle passage he is paraphrasing. Still, could Michael’s gloss perhaps serve to inspire a viable reading of B165? The message of “DQWKUǀSRV is what we all know” might have been this: ignore or set aside what is manifest; go beyond it; search rather for the underlying realities (the atoms, presumably), which are hidden. If B165 is read in this way, the fragment would have affinities with such Democritean fragments as B117:

10

It is relevant to note that in the corresponding passage in manuscript F (Parisinus gr. 1859, a MS that represents a significantly different tradition of Michael’s commentary), Democritus’ view is not presented as a quotation. See Hayduck 1904, p. viii. 11 That the attribution is to Democritus is clear. The remark at issue immediately precedes “Aristotle objects to this, saying . . .”. Besides, the thought is conspicuously un-Aristotelian. 12 See preceding Chapter and, in Greek, Mourelatos 1984b.

6. Democritus on the Distinction Between Universals and Particulars

155

“We know nothing in truth; for reality lies in the depths.” 13 And yet there is something unsettling even in this more generous (but also loose) reading of B165. For it does not at all seem in the spirit of Democritean inquiry to ignore or to set aside a body of data from the realm of the “evident.” We have more than a superficial glimpse of how Democritus proceeds from the evident to the non-evident in one major part of his anthropology, his theory of perception. Amply expounded in Theophrastus’ De sensibus, that theory, far from ignoring the manifest data of perception, seeks to explain these data by connecting them intelligibly and in detail with the corresponding, underlying atomic structures. Both the reductivist and the Michael-inspired interpretations of B165 give the fragment the import of promoting a certain philosophical or scientific methodology. Neither makes any significant use of the specific and overarching theme that was first noticed above, after the two source-passages from Sextus and from Aristotle were quoted: the theme of high-level generality. It is significant that Sextus starts by recalling Socrates’ “what is it” question, as raised in Phaedr. 230A. In both Sextus and in Aristotle, the issue that prompts the allusion to Democritus is one of grasping the essence of such types as DQWKUǀSRV, dog, horse, plant, animal. With this clue, an attractive alternative interpretation of B165 may be forthcoming – if we attach to it a certain logical sequel or qualifier: DQWKUǀSRV is what we all know. Individual DQWKUǀSRL, however, are known at most to only some of us.

Let me elaborate on the suggested interpretation. We all know the type DQWKUǀSRV. There is no way, however, that all of us – all of humanity across the ages and across the infinity of kosmoi in Democritus’ universe – could possibly know Leucippus, or Democritus, or Protagoras, or the king of Persia. We could not all know the same token DQWKUǀSRV. 14 Even superlatively prominent and famous individuals are known, at most, to many others, not to all. Indeed, there is no guarantee that a particular individual should be known by anyone else; yet each of us, to the extent one has any self-awareness, is aware of oneself as an DQWKUǀSRV Finally, it is a fortiori true that neither any one of us nor all of us can know DOO LQGLYLGXDO DQWKUǀSoi. Knowledge of DQWKUǀSos (type) is complete, unrestricted, and public; knowledge of DQWKUǀSoi (tokens) is inevitably incomplete, restricted, and parochial.

13 14

For all the relevant epistemological fragments, see Taylor 1999, pp. 8-12. Interestingly, this implication of Democritus’ B165 is recognized in the other of the two Sextus passages that are a source for the fragment: ਥțİ૙ȞȠȢ ȝȑȞ ijȘıȚ įİ૙Ȟ ਫ਼ʌઁ ʌȐȞIJȦȞ ȖȚȖȞȫıțİıșĮȚ IJઁȞ ਙȞșȡȦʌȠȞ, Ƞ੝įȑȞĮ į੻ ਙȞșȡȦʌȠȞ ੅ıĮıȚ ʌȐȞIJİȢ ਙȞșȡȦʌȠȚ (Pyrrh. hyp. II.23-24).

156

Alexander P. D. Mourelatos

III In the light of what was presented in Sections I and II, it would appear that B165 is closely related to B124. The two fragments, taken together, serve to explicate the distinction between types and tokens, universals and particulars, the one and the many. B124 has purely metaphysical tenor; it focuses directly on the one-many relation that binds entities across the two metaphysical levels. B165 is no less metaphysical; but it aims to capture that same relation by differentiating between the two levels in epistemological terms. Still, is it reasonable to suppose that Democritus will have concerned himself with the type-token distinction? I am convinced that the answer must be Yes. For, as I said at the start of this Chapter, and as has also been argued in the preceding one, it is impossible for Democritus’ doctrine of many-tiered infinities to be formulated without recourse to the distinction. Besides, we have the evidence from Plato’s Socratic dialogues that tends to suggest that the distinction must have frequently come up in philosophical debates in the later decades of the fifth century, as it did throughout the fourth century. (And it is well to remind ourselves, from time to time, that Democritus was younger than Socrates, and that he was still alive in Plato’s later years.) So, then, inevitably either some tough-minded interlocutor or critic, or Democritus’ own dialectical conscience, would have raised challenging questions such as these: What are these types of atomic ideai? What knowledge could we have of them? Is the knowledge of types more secure or is it less secure than the knowledge of tokens? In dealing with such difficulties the analogy with familiar and ordinary objects would have been in order: there are individual DQWKUǀSRL; and there is the type $17+5ƿ326. And in response to any skepticism concerning our knowledge of types, nothing would have been so effectively disarming as B165: we all know what DQWKUǀSRV is. On this reading of B124 and B165, we have significant resonance with several other Democritus fragments. B164 (which is only accidentally a neighbor to B165 in DK) speaks first of the general tendency of animals to herd with their own kind (KRPRJHQHVL ]ǀRLV V\QDJHOD]HWDL), e.g., doves with doves, cranes with cranes. Then it notes that there is also a tendency of lifeless objects to aggregate like-to-like after sustained agitation: on the beach, pebbles of like shape; in the farmer’s sieve, lentils with lentils, barley with barley, wheat with wheat. The fragment concludes with this observation: “as though the likeness present in these things has something that acts to bring them together” (੪Ȣ ਗȞ ıȣĮȖȦȖȩȞ IJȚ ਥȤȠȪıȘȢ IJ૵Ȟ ʌȡĮȖȝȐIJȦȞ IJોȢ ਥȞ IJȠȪIJȠȚȢ ੒ȝȠȚȩIJȘIJȠȢ). The “as though” construction strongly suggests that this final observation is Sextus’ own comment; but the phrase “likeness present in . . . things” very probably reflects a Democritean conception. In any event, the theme of types is conspicuous in the fragment. With or without the final observation, B164 can serve to buttress B165, when the latter is deployed to

6. Democritus on the Distinction Between Universals and Particulars

157

answer challenges to the assumption that we can grasp types. The challenge could come either from one who was more broadly a skeptic or from the sort of empiricist that Democritus was; or it could be inspired by the idealist alternative, the one represented by Plato’s metaphysics. There is nothing unlikely or mysterious, Democritus would be saying through B164, in our ability to recognize types; 15 we certainly recognize our own human type; and this is grounded in a like-to-like tendency that is widely in evidence, throughout both realms, animate and inanimate. There is also resonance with the fragments in which Democritus himself appends certain skeptical qualifications to his epistemology. The same philosopher “whose sayings have been likened to the pronouncements of Zeus”, when the scope of such sayings is that of “things in general” (SHUL WǀQ [\PSDQWǀQ, Sextus passage, above) 16, nonetheless proclaims agnosticism about “things in particular”: Even so (kaitoi), it will become evident that it is not possible to know what each particular thing is (hoion hekaston) in reality (eteƝL). (B8) Surely, on the one hand (men), it has been shown in many ways that, in terms of reality (HWHƝL), we do not grasp what each particular thing (hoion hekaston) is or is not . . . . (B10)

There is simply no way to know individual atoms; but this is not due wholly to the extreme minuteness of their dimensions; even at the level of ordinarysize objects, our knowledge of individuals is limited in scope. Nonetheless, there is no bar to knowing atoms as ideai, as types of shape, and to know all the geometric and kinematic properties (also types) that are entailed by these ideai.

15

The passage in which Galen explains the type/token distinction takes a similarly demystifying approach: “And it is so evident, and it so naturally belongs to everything, be it man or beast, to recognize . . . the unitary form (KǀVHLGRVKHQ), that even donkeys, . . . the stupidest creatures of all, manage to distinguish between things which are one in form and those which are one numerically” (De meth. med. II.7, Kühn 1825, p. 133; transl. from Hankinson 1991, p. 67). 16 Might xympanta have been Democritus’ term not for “the universe” but for “universals”? The sense fits quite well the Sextus passage in which the term occurs: Adv. math. VII.264-67, translated above. It even fits nicely inside the phrase pros to sympan that is used by Sextus to recall Socrates’ “what am I” question in Phaedrus 230A, where there is no reference to “the universe”. Tempting as it is to consider this possibility, other evidence does not support it. We read in Cicero’s Academica, II.23.73: “What shall I say about Democritus? . . . the one who dared to begin (qui ita sit ausus ordiri) with these words: ‘These are my utterances about the totality of things (haec loquor de universis)’? He excludes nothing from the scope of his announcement – for what can exist outside the totality of things (quid enim esse potest extra universa)?” Given his testimony, we must assume that SHUL WǀQ [\PSDQWǀQ did appear at the start or near the start of a major Democritean treatise, where it could not have borne the rather technical sense “about universals”. But we need not adhere specifically to Cicero’s construe of universa as “the totality of things, the universe”. At the start of Democritus’ treatise, the phrase SHUL WǀQ [\PSDQWǀQ may have had the force simply of “things in general, things broadly considered”.

158

Alexander P. D. Mourelatos

The new reading of B124 and B165 proposed here is, of course, speculative. It is offered as an exploration of a philosophically attractive alternative that has been overlooked. 17 And it is in this spirit that I conclude with an even more venturesome bit of speculation. In Diogenes Laertius and in Cicero (in Latin translation) we find an autobiographical remark by Democritus: “I came to Athens and not a single person knew me” (B116). This is generally taken to express either Democritus’ disenchantment with the smugness of Athenian intellectuals or his reveling in self-effacement and modesty. But perhaps B116 too had metaphysical tenor. Could it not have been Democritus’ poignant example of limitations in our knowledge of individuals? Might it not be the case that B116 and B124, separated only by an adversative particle, originally formed a single statement? . . . ਷ȜșȠȞ Ȗ੹ȡ İੁȢ ਝșȒȞĮȢ țĮ੿ Ƞ੡ IJȚȢ ȝ੻ ਩ȖȞȦțİȞ. ਙȞșȡȦʌȩȢ ਥıIJȚ ੔ ʌȐȞIJİȢ ੅įȝİȞ. . . . For I came to Athens and not a single person knew me. DQWKUǀSRV, , is what we all know.

17

The ideas developed here were first presented informally to an ad hoc discussion group at The University of Texas at Austin in spring 2002. I am grateful to Richard Sorabji for convening the group, and to all participants in the discussion for their comments and for their encouragement. I wish to acknowledge in particular the very helpful suggestions offered by the late Herb Granger and by Ravi Sharma. I also thank Dina Garmong for translations from the Russian of crucial passages of S. J. Lur’e’s Demokrit. For comments on the penultimate version, I thank Richard D. McKirahan, Jr., Lesley Dean-Jones, and I also acknowledge here my debt to the late Scott Austin, who was the respondent when this paper was read (under the title “Democritus on the Distinction Between Types and Tokens”) at the 26th Annual Workshop in Ancient Philosophy at Texas A&M University in March 2003.

7.

Intrinsic and Relational Properties of Atoms in the Democritean Ontology

Aristotle’s survey, in book I of the Metaphysics, of the contributions of the earlier natural philosophers provides us with a major source-text for the fundamentals of fifth-century B.C.E. atomic theory: 1 Leucippus and his associate Democritus declare the full and the empty [the void] to be the elements, calling the former “what-is” (to on) and the latter “what-is-not” (WR PƝ RQ) . . . They declare that the differentiations (also “differences,” diaphorai) [of “what-is,” or of “the full”] are the causes of all else. Now they say that the differences at issue are three: shape (VFKƝPD), and array (taxis), and posture (thesis). 2 For they say that what-is differs only in “rhythm,” in “junction,” and in “bearing” (૧ȣıȝ૶ țĮ੿ įȚĮșȚȖૌ țĮ੿ IJȡȠʌૌ ȝંȞȠȞ  2I WKHVH ³UK\WKP´ (rhysmos) is shape, “junction” (GLDWKLJƝ) is array, and “bearing” (WURSƝ) is posture. For A differs from N in shape, AN from NA in array, and Z from N in posture. 3 Concerning motion – the origin of it, and how it is present in the things-that-are (੖șİȞਲ਼ ʌ૵Ȣਫ਼ʌ੺ȡȤİȚIJȠ૙Ȣ Ƞ੣ıȚ – they, more or less like the others, did not bother to give any account ʌĮȡĮʌȜȘı઀ȦȢIJȠ૙Ȣ ਙȜȜȠȚȢ૧Įș઄ȝȦȢਕijİ૙ıĮȞ Metaph. I.4.985b4-21) 4

Originally published in Metaphysics, Soul and Ethics: Themes from the Work of Richard Sorabji, Ricardo Salles, ed., Oxford, Clarendon Press, 2005, pp. 39-63. It is reprinted here with the kind permission of Oxford University Press. 1

2

3

4

In the discussions of atomic theory in his two magisterial surveys of ancient philosophy and ancient science – Time, Creation, and the Continuum; Matter, Space, and Motion – Richard Sorabji has offered us analyses, comments, and insights that are as permanently of value as those found in the best of monographs on the subject of early Greek atomic theory. It was with immense admiration, in awed humility, and with deep gratitude for his personal and academic friendship that this essay was composed as a contribution to a volume in Richard Sorabji’s honor. The usual translations for taxis and for thesis are “arrangement” and “position,” respectively. But “arrangement” has misleading connotation of purposeful ordering by an arranger, and “position” fails to capture the semantic component of orientation or (geometric) attitude. For the last of the three contrasts, Aristotle’s examples are capital eta and capital iota (the latter of which, in his script, is H turned 90 degrees). My translation, with borrowings from: Taylor 1906, pp. 90-91; McKirahan 1994, pp. 304-05; and Taylor 1999, pp. 72-73. This famous passage has many parallels in the ancient commentators on Aristotle and in other sources: see Lur’e 1970, 69-71.

160

Alexander P. D. Mourelatos

The differentiated bits of “the full” are, of course, the atoms. The remark about motion is tendentious. Our evidence is that Democritus did say a lot about the motion of the atoms; 5 he just did not say what Aristotle would have wanted him to say, viz., what the “natural” motion of atoms is, in that technical sense of “nature” and “natural” Aristotle introduces in Phys. II. Nonetheless, through this studied separate mention of the motion of the atoms, Aristotle conveys, in effect, an important theme of Democritus’ theory: the actual motion of atoms is not intrinsic to them; it is a property atoms acquire as a result of collision; 6 and thus it cannot be included in a statement of the causally primary differentiations. On a first reading, there are two obvious infelicities in this statement of “the three diaphorai.” To begin with, there is the curious omission of the property of size. That is surely as much of an intrinsic property of atoms as shape, and the causal role of atomic size is widely attested in the testimonia. 7 Indeed, there is a pattern in our sources for the omission of separate mention of size when shape is cited as a fundamental property of atoms. So, it is not unreasonable to assume that “shape” was used by Democritus in synecdoche for both shape and size; and that presumably is the case also in Metaph. I.4. The second infelicity is that only one of the “the three diaphorai” explicitly mentioned by Aristotle, shape (or shape-size), can specify properties that belong to atoms individually, in themselves, intrinsically. “Array” could only be a property of pairs, trios, or n-tuples of atoms. Likewise, “posture” necessarily implies a background or matrix, against which, or within which, an individual atom is placed. That background or matrix could not be merely the void. The latter, conceived as a “what-is-not,” lacks even the minimal degree of definiteness that would allow for some points of purchase or reference. The matrix or background for the single atom would have to be provided either by one other single atom (compare our own use of underlining to help distinguish the numeral 6 from the numeral 9), or in a configuration with two or more other atoms. 8 Aristotle himself gives an aptly more cautious statement in On Coming to Be and Passing Away: Democritus and Leucippus say that there are indivisible bodies out of which everything else is composed (ਥț ıȦȝ੺IJȦȞ ਕįȚĮȚȡ੼IJȦȞ IJਛȜȜĮ ıȣȖțİ૙ıșĮȚ); that these (indivisible bodies) are infinite both in number and in variety of shape; and that compounds differ from each other (Į੝IJ੹ ʌȡઁȢ Į੝IJ੺) in respect of the components [i.e., the number and the shapes of components] 5

See Lur’e 1970, 79-83; McKirahan 1994, pp. 317-20; Taylor 1999, pp. 88-89. See esp. fr. DK67A6. Cf. Taylor 1999, pp. 88 (texts 60a-e) and 179-84. 7 Cf. Taylor 1999, p. 172. 8 The infelicity has often been noted in modern accounts of this Democritus testimonium. See: Ross 1953, p. 140 ad loc.; Salem 1996, p. 33; Hankinson 1998, p. 206 n. 4; Taylor 1999, pp. 171-72. 6

7. Intrinsic and Relational Properties of Atoms in the Democritean Ontology

161

and also in respect of the posture and array of these components (įȚĮij੼ȡİȚȞ IJȠ઄IJȠȚȢ ਥȟ ੰȞ İੁıȚ țĮ੿ ș੼ıİȚ țĮ੿ IJ੺ȟİȚ IJȠ઄IJȦȞ). (De gen. et corr. I.314a2124) 9

Array and posture are here explicitly mentioned as properties possessed by atoms within compounds. The same cautious formulation is found in several other Leucippus/Democritus testimonia. 10 Let us then assume, for a start, that the infelicity in the phrasing of the passage from Metaph. I.4 is to be corrected by our understanding all three of the “differences” (including the property alternately referred to as “shape” and “rhythm”) as applying to atoms in relational complexes – either within stable compounds or as atoms are juxtaposed one to another in various configurations. 11 There is, however, yet another and more intriguing inconcinnity in Metaph. I.4. Aristotle offers “shape,” “array,” and “posture” as appropriate synonyms of the Democritean rhysmos, GLDWKLJƝ, and WURSƝ. But, as was noticed by Kurt von Fritz more than half a century ago, 12 there is a striking and systematic discrepancy between the Democritean terms and the synonyms Aristotle offers. The discrepancy is not to be explained away as one between two dialects, Aristotle’s Attic and Democritus’ Thracian Ionic. As the examples of letter shapes and syllables suggest, Aristotle understood the properties at issue as occurrent and static. This is clearest in the case of his translation VFKƝPD, “shape.” Moreover, from the examples Aristotle uses, it is also clear that in the case of the terms taxis, “array,” and thesis, “posture,” Aristotle envisages occurrent or actual states of atomic configuration. By contrast, the Democritean terms – all three of them – have a strongly dynamic and dispositional flavor. It is worth reviewing, and in some respects amplifying, the observations made by von Fritz. The noun WURSƝ derives as transparently and as directly from the action-sense of the verb trepein as the English gerund “turning” does from its cognate verb. 13 The term GLDWKLJƝ, “junction,” combines the preposition dia-, “trans-,” or “cross-,” with a noun derived from the verb WKLQJDQǀ, “to touch,” “to take hold of,” or – what is more likely – directly from the middle form, diathinganomai, “to come in mutual contact, touch and be touched.”14 In 9

Translation as in Williams 1982, p. 1, with minor adaptations. For the text and construction, see Joachim 1926, pp. 65-66. 10 For example, in the Aëtius compendium: IJ੹ įૃ ਥț IJȠ઄IJȦȞ ș੼ıİȚ țĮ੿ IJ੺ȟİȚ țĮ੿ ıȤ੾ȝĮIJȚ įȚĮij੼ȡȠȞIJĮ ıȣȝȕİȕȘțંIJĮ, “the compounds [literally “the things out of them”] differing, as they happen to do so, in posture, array, and shape” (DK67A32). For other texts, see Lur’e 1970, nos. 233, 433, 434. 11 In due course this provisional assumption will, in the present study, be qualified. It will turn out that whereas VFKƝPD can be understood as a property of atoms as distinct individuals (per se), rhysmos is shape that involves relations with other atoms and patterns of atomic movement. 12 See Fritz 1966, pp. 25-29. 13 Cf. Fritz 1966, pp. 26-28. 14 See below n. 43. In Fritz 1966 GLDWKLJƝ is explicated as follows: “das ‘sich hindurch bewegend in Berührung, in Kontakt, in Zusammenhang mit etwas kommen’ das ‘sich fügen’” (p. 28).

162

Alexander P. D. Mourelatos

either case, the derivation of the noun from an action verb by addition of the accented suffix -੾ or -੽ (eta with either acute or grave accent) is likely to convey not a state but, as in the derivation of WURSƝ from trepein, a disposition or activity. 15 Finally, and most interestingly, rhysmos is not to be given the facile translation “shape.” When referring directly and simply to the intrinsic shape of atoms, Democritus uses the term idea. As von Fritz explains, rhysmos, which is etymologically related to UKHǀ, “to flow” (a semantic nuance that survives in the modern term “rhythm”), conveys “the objective law of a pattern [that is characteristic] of an object in itself . . . form as an outflow of motion from the object itself.” 16 Democritus’ eschewing of the term idea in the present context, his choosing instead terms with pronounced dynamic connotations, must have a special rationale that may have been rubbed out in Aristotle’s flat translations. In this essay I propose to use the passage from Metaph. I.4 as the starting point for exploring the logical structure of Democritean 17 atomic properties – from the level of the utterly simple, independent, and fundamental, on to higher levels that are cumulatively more complex, dependent, and derivative. Some properties belong to an atom intrinsically, without any implied reference to anything outside it. Other properties can be made sense of only insofar as atoms are in some relation to other atoms. The outcome in this exercise will, I hope, provide something more than a tidied-up statement of the fundamentals of Democritean atomic theory. It ought to help exhibit Democritus’ overall project of physicalist metaphysics as logically systematic, gradualist, synthetic, and generative. A historically germane by-product of the present account will be an explication of the pointedly dynamic vocabulary Democritus had used in the original context that lay behind the passage from Metaph. I. It will emerge that the Aristotelian trio of shape/array/posture need not be taken merely as a slack, prosaic version of the Democritean trio of rhythm/junction/bearing. Both sets of terms have their function; but they come into play, at different levels in the scheme of properties. I shall myself initially use the semantically less charged 15

Consider the noun derivatives from verbs that have some semantic affinity to șȚȖȖ੺ȞȞȦ/įȚĮșȚȖȖ੺ȞȠȝĮȚ (all of them implying some form of contact): ਖȡʌĮȖ੾, “snatching away, seizure” (from ਖȡʌ੺ȗȦ); ਖij੾, “touching, kindling, faculty of touch” (from ਚʌIJȦ); ȕĮij੾, “dipping in dye” (from ȕ੺ʌIJȦ); ȖȡĮij੾, “drawing, painting, writing, art of writing” (from Ȗȡ੺ijȦ). The tendency of the suffix -੽ (-੾) to retain the dynamic connotation of the base verb is especially striking when the derivation involves, as in the case of IJȡȠʌ੾ from IJȡ੼ʌȦ, in ablaut vowel-shift, for example: ੒Ȝț੾, “drawing, tugging” (from ਪȜțȦ); țȜȠʌ੾, “stealing” (from țȜ੼ʌIJȦ ȝȠȝij੾, “blame, censure” (from ȝ੼ȝijȠȝĮȚ), etc. 16 Fritz 1966, p. 26: “das objektive Gestaltgesetz des Gegenstandes in sich; . . . die Gestalt . . . als Ergebnis einer Bewegung aus sich selbst.” 17 As is now standard in accounts of fifth-century (Abderite) atomism, I make no attempt to distinguish between the contributions of Leucippus and Democritus. I shall be using the proper name “Democritus” and the adjective “Democritean” to refer to Leucippus and/or Democritus and to aspects of the philosophy of either or both.

7. Intrinsic and Relational Properties of Atoms in the Democritean Ontology

163

Aristotelian terms. At a later stage, a richer context will demand and will help explicate the semantic rationale of the trio in Democritus’ own terminology. *** In constructing and exploring the scheme of properties, I need to deploy three conceptual devices. First, I shall be making reference to the “Parmenidean requirements,” by which I mean the four attributes of “what-is” famously deduced by Parmenides in fragment B8 of his “Truth”: (a) “unborn and imperishable;” (b) “homogeneous, indivisible, one;” (c) “immobile;” (d) tetelesmenon, i.e., “fully actualized, free of latent dispositional properties,” and thus totally unchangeable. 18 Even though there is no need to assume that Leucippus and Democritus saw their project as one of “replying to Parmenides,” or more generally “replying to the Eleatics,” 19 it is transparent that the fundamental realities of atomism are modeled on Parmenides’ “what-is.” Adherence to, or deviation from, the four requirements can thus serve as the historically most germane measure of the conceptual distance between stages in the systematic transition, referred to above, from static monism to dynamic pluralism. The second device is one I have already used, the familiar distinction between “dispositional” and “occurrent” properties. The distinction can be reviewed briefly with reference to an example: a ball has the dispositional property to roll; but when the ball is actually rolling, the property is occurrent. Related to this distinction, but not quite equivalent to it, is the third device I shall need: the distinction between determinable and determinate properties. All pots in a potter’s studio have shape in the determinable sense of “shape”: they have some-shape-or-other. But the particular pot on the showcase has shape in the determinate sense of shape: it is, say, an amphora of X, Y, and Z specifications. Technological limitations notwithstanding, it is possible for two or more pots to have exactly the same determinate shape. 18

The four attributes are announced as follows in the programmatic statement in the central fragment from Parmenides’ “Truth,” DK28B8: (a) ਕȖ੼ȞȘIJȠȞ ਕȞઆȜİșȡȠȞ OLQH   E  Ƞ੣ȜȠȞ ȝȠȣȞȠȖİȞ੼Ȣ IJİ (line 4), ੒ȝȠ૨ ʌ઼Ȟ  ਪȞ ıȣȞİȤ੼Ȣ OLQHV -6); (c) ਕIJȡİȝ੼Ȣ OLQH   G  IJ੼ȜİȚȠȞ RU IJİȜોİȞ OLQH  adopting either of the proposed emendations for the puzzling ਕIJ੼ȜİıIJȠȞ RI WKe MSS), or actually preserving the ਕIJ੼ȜİıIJȠȞRIWKH066E\JLYLQJLWWKHWUDQVODWLRQ³QRWto be accomplished” = “not (still) incomplete.” The announced attributes are then supported with arguments, in some cases with additional synonyms of the attributes appended: (a) argument for ਕȖ੼ȞȘIJȠȞLQOLQHV-E Ƞ੝į੻ įȚĮȚȡİIJંȞ੒ȝȠ૙ȠȞ-25); (c) ਕț઀ȞȘIJȠȞ- G Ƞ੝țਕIJİȜİ઄IJȘIJȠȞIJİIJİȜİıȝ੼ȞȠȞ-33, 42-49). Some readings of Parmenides treat “eternal” or “existing now” Ƞ੝į੼ ʌȠIJૃ ਷ȞȠ੝įૃ ਩ıIJĮȚਥʌİ੿ Ȟ૨Ȟ ਥıIJȚȞOLQH DVDGLVWLQFWDGGLWLRQDODWWULEXWHDQGDFFRUGLQJO\LQFUHDVHWKHWRWDOOLVWRIDWWULEXWHVWR five. But if this (in the order of the text) second-out-of-five attribute has the sense of sempiternity (always existing), it is not sufficiently distinct (for purposes of a Democritus-Parmenides comparison) from “unborn and imperishable.” If, however, it bears the sense either of “atemporal” or of “punctual” existence, then it has no counterpart in the Democritean system and would likewise not be relevant to the present investigation. 19 See above, Chapter 1 in Part I.

164

Alexander P. D. Mourelatos

Let me first present the overall scheme in outline form. I shall then expand on each of the lemmata of the outline: PROPERTIES OF DEMOCRITEAN ATOMS: INTRINSIC VS. RELATIONAL I. INTRINSIC A. Dispositional? 1. [Mobility is not an intrinsic property.] B. Occurrent (fixed) 1. Uniform a. [The atom has no special material-constitutive qualities.] b. Fullness, rigidity, solidity (properties opposite to those of the void) 2. Multiform a. Shape b. Size II. RELATIONAL (OF ATOMS IN RELATION TO ONE ANOTHER) A. Dispositional 1. Independent of shape-size a. Univalent i. Impassivity (except for rebound) b. Multivalent i. Bearing ii. Mobility iii. Capacity to rebound 2. Dependent on size a. Weight 3. Dependent on shape and size a. Conjunctivity b. Motility i. The tendency of atoms of like shape/size to aggregate ii. The kinematic disposition of different atomic shapes/sizes Į. Characteristic speed ȕ. Characteristic geometry of motion B. Occurrent (variable) 1. Determinate posture or tilt (cf. Aristotle’s thesis) 2. Determinate motion 3. Determinate array or join (cf. Aristotle’s taxis)

***

7. Intrinsic and Relational Properties of Atoms in the Democritean Ontology

165

I. In trin sic P ropertie s It is a fundamental principle of the Democritean theory that there is a plurality, indeed an infinity, of atoms. The “intrinsic” properties are the ones an atom could be meaningfully said to possess independently of any reference to other atoms, or even if – contrary to another fundamental principle of the Democritean theory – there was just a single atom in the infinite expanse of the void. IA. In trinsic a nd Disp o sitio nal? Democritean atoms – and, for that matter, atoms in all versions of ancient atomism – are famously unchanging and unchangeable in all intrinsic respects (see below IB1); they do, however, move through the void. Should we then not say that there is just this one intrinsic dispositional property even a single isolated atom should possess, that of mobility? Prima facie, it ought to make sense to say even of a single atom that at any time it has the determinable property of motion, in other words, that it is either at rest somewhere-or-other in the void, or moving in some-way-or-other through the void. And since atoms generally do exhibit motion as an extrinsic occurrent property, it should appear strange to hold that a single atom, per se, does not possess the ability to move. And yet that seems to be precisely what is required by the logic of Democritus’ conception. The same considerations that rule out “array” and “posture” as intrinsic properties of atoms dictate that motion even in the dispositional sense can only be conceptualized as a relational property of atoms. 20 It is relevant here to remind ourselves of the major contrast, alluded to at the start of this essay, with Aristotelian “elements”: Democritean atoms do not possess an “intrinsic principle of motion and rest;” 21 nor, unlike the atoms of Epicurus, may they be said to be constantly “falling.” Any motion exhibited by a Democritean atom results from collision with other atoms. 22 If occurrent and determinate motion implies reference to a prior collision (for that matter, a history of such collisions), mobility and determinable motion as properties of a given atom correspondingly and inevitably posit the existence of atoms other than the one at issue. The corollary at this point is that the only intrinsic properties for atoms are strictly occurrent. 20

21

22

This implication of the analysis pursued here was most helpfully pointed out to me by Richard McKirahan. In comments he offered on a draft of this paper (see below n. 47) he wrote: “If there were only one atom in an infinite expanse of void, it would be impossible to tell whether [that single atom] were in motion or at rest; motion and rest can only be relative to other atoms.” The “vibratory motion” (palmos) attributed to Democritus’ atoms in the Aëtius compendium (DK68A47) is almost certainly a pattern of occurrent motions that results from frequent collisions, not an intrinsic tendency for random motion. See below at IIB. And, as it has often been pointed out, no vicious regress is involved in allowing acquisitions of motion to regress to infinity. The regress is merely temporal, and Democritus expressly holds that time is infinite. See Taylor 1999, p. 181. Cf. McKirahan 1994, p. 318.

166

Alexander P. D. Mourelatos

IB. Occurrent and fix ed Accounts of Democritus do not always emphasize this point, but it is of some importance to note that in the domain of intrinsic properties that are occurrent and fixed (or unchanging) we need to distinguish between uniform and multiform properties. IB1. Fix ed and uniform One is sometimes tempted to ask: What is the stuff Democritean atoms are made of? What is the material inside the borders set by the shape of the atom? 23 It would make especially good sense for Aristotle to frame the question in these terms, since Aristotle is committed to a hylomorphic paradigm, one that sharply distinguishes between the figurational/formal/structural aspect of things and their material-constitutive aspect. Indeed, the form-matter distinction does come up in one of his references to the Atomists: [Leucippus and Democritus] say that the nature [of “the shapes,” i.e., the atoms] is one, as though each were a separate piece of gold. (Aristotle, De caelo I.7.275b-276a1 = DK67A19)

It is well to note that Aristotle hedges his “gold” example with an “as though.” The question I called “tempting” is in fact misguided; the analogy is, in the end, more misleading than helpful. The form/matter distinction is conspicuously absent at the fundamental level of Democritean metaphysics. 24 Atoms are certainly VǀPDWD, “bodies, corpuscles;” but they have no specifiable constitutive matter. Materials, such as gold, or the Empedoclean “elements” (earth, water, air, and fire), and indeed stuff and matter generally – all these are physically and metaphysically derived entities in all versions of ancient atomism; they are complex aggregations of atoms. Interestingly, Aristotle himself comes close to making this point in another context, in a comparison between Leucippus’ conception and Plato’s theory of elementary triangles in the Timaeus: This account is to this extent different from that given by Leucippus, that in the latter solids, while in the former planes, are regarded as the indivisibles. (Aristotle, De gen. et corr. I.8.325b25-27) 25

We might add that the early Atomists, true to their commitment to the indivisibility of the atom, showed as little interest in the interior of their atoms 23

Cf., for example, Hasper 1999, p. 4: “given the fact that all atoms are made of the same material.” I have argued in the immediately preceding Chapter that another distinction, that between universals and particulars, or types and tokens, is indeed paramount in Democritus. My claim that Democritus has thematic awareness of the type/token distinction does not compromise the point made above concerning the absence of a form/matter distinction in Democritus. For it would be a mistake to assume that the form/matter distinction is presupposed by the type/token distinction. 25 Translation from Williams 1982, p. 28. 24

7. Intrinsic and Relational Properties of Atoms in the Democritean Ontology

167

as Plato showed in what lay inside the triangles that form the faces of his elementary molecules. Indeed, some of the ancient sources speak of the atoms as apoia, “qualitiless” (DK68A125), which is an apt negative description of the interior of Democritean atoms. Nonetheless, if we are to characterize properly the atom in contrast both to the void and to non-atomic objects, there are some intrinsic properties that are independent of properties conferred by the shape of the atoms and which, in that sense, may be thought of as applying to the atom through and through. Like the void, the atom is apathes, “impassive.” 26 This attribute can obviously be taken in two senses: (i) internally unchanging, always in the same state; (ii) immune to any change that may be brought about through interactions with other atoms. In the second of these senses, impassivity expressly envisages the existence of other atoms. Sense (ii) of impassivity is therefore best discussed under IIA1 below. In sense (i), impassivity need not even be understood as a disposition (i.e., as immunity to change); it should rather be viewed as a state, as changelessness, an occurrent property that is uniformly possessed by each atom. 27 Moreover, unlike the void, indeed in direct contrast to it, the atom is an absolute plenum, free of gaps or interstices, absolutely solid, rigid. These multiple characterizations of the second property do not represent different properties but different entailments of the single property of to naston, “the solid.”28 It is now relevant to observe that with respect to these two uniform and unchanging properties (changelessness and solidity), all four of the Parmenidean requirements may be claimed as fulfilled – with no qualifying rider. Since what is presently at issue are properties that are occurrent and fixed, there is no need to dwell on the requirement of “unborn and imperishable.” The requirement “homogeneous, indivisible, one” is ostensibly fulfilled by virtue of the property of solidity. Indeed our sources offer “solidity” as the reason for the atom’s indivisibility. 29 The property of mobility, as we saw, may not be regarded as intrinsic; the atom per se may properly be regarded as “immobile,” and that is precisely how it is spoken of in one of our sources. 30 With even the dispositional property of mobility ruled out at this level of the analysis, the atom’s changelessness guarantees that the fourth Parmenidean requirement, “fully actualized,” is also met. 26

DK68A1, A49, and additional texts in Lur’e 1970, 65-67. For the denial of metaballein, “changing,” see, e.g., DK68A57. There are also several references to denials both of metaballein and of alloiousthai, “alteration,” in the texts cited at Lur’e 1970, 66. 28 DK67A7, 68A38, and additional texts in Lur’e 1977, 65-67. 29 DK68A43, A56 (Cicero: individua propter soliditatem), and additional texts in Lur’e 1970, 65-67. 30 6LPSOLFLXVǻȘȝંțȡȚIJȠȢij઄ıİȚਕț઀ȞȘIJĮȜ੼ȖȦȞIJ੹ ਙIJȠȝĮʌȜȘȖૌ țȚȞİ૙ıșĮ઀ ijȘıȚȞ³'HPRFULWXVDUJXLQJ that the atoms are by nature immobile, states that they are moved by ‘blow’” (DK68A47). The “nature” vs. “blow” contrast shows unmistakably that the distinction here is not between changelessness-in-general and specifically locomotion but rather between denial of intrinsic locomotive ability and extrinsically acquired motion. 27

168

Alexander P. D. Mourelatos

IB2. Fix ed and mu ltifo rm It is at this level that we have the first deviations from the Parmenidean requirements. Though we are not yet taking into account relations among atoms, we do recognize not only a plurality, indeed an infinity, of atoms but also a plurality, indeed likewise an infinity, of shapes and sizes that characterize them. It is worth noticing that it is not the first of these pluralities, that of token atoms, which makes for incompatibility with Parmenides’ second requirement of “homogeneous, indivisible, one.” For there have been models of Parmenides’ argument that allow the requirements to apply to a universe of multiple Parmenidean “ones” or monads. 31 The crucial incompatibility comes with the stipulation of shape as an intrinsic property of atoms. Even if all atoms had the same shape, and even if there were just a single spherical or cubical or however-shaped atom, we would immediately need to recognize that we are envisaging entities that have parts. Atoms, accordingly, may not be considered indivisible in an unqualified sense. 32 The violation of the second requirement is only compounded with the stipulation of diversity in shape and size. What is inevitably entailed in this stipulation are contrarieties of all sorts, when even one such contrariety would have violated Parmenides’ requirement of homogeneity. There are, for instance, atoms that have curved surfaces, and others that have surfaces bounded by straight lines and angles; there are atoms that have regular shapes and others that have irregular shapes; and there are larger atoms and smaller atoms. The other three Parmenidean requirements are in no way adapted or modified at this level; they apply at full strength. An occurrent, determinate shape and an occurrent determinate size are an atom’s unborn, imperishable, and unchanging possession throughout infinite time. The total rigidity-solidity of the atom is the secure warrant of this. Moreover, there is nothing either about shape as such or about size as such, nothing about the actual infinity of shapes and sizes, and nothing about the infinity of token atoms that should prompt us to reconsider the character of individual atoms as intrinsically “immobile.” With respect to the fourth requirement, that of “fully actualized,” it suffices to note that fixed shape and fixed size are occurrent properties. With 31

See, e.g., Mourelatos 1970/2008a, pp. 111-14, 130-35; Barnes 1979a, II, pp. 204-07; Barnes 1979b; Curd 2004, 65-75. 32 Some of the testimonia support the view that Democritean atoms were only physically indivisible; others, that atoms were also conceptually indivisible. Sorabji 1983, pp. 354-57, offers a comprehensive and judicious account of the controversy and proposes mediation: Democritus did not distinguish between physical and conceptual indivisibility. This I find unlikely. In the context of specific passages from authors who give testimony about Democritus there may well be a failure to distinguish between the two alternatives. But I have difficulty accepting that Democritus would have missed so glaring a contradiction at the heart of his theory – shapes that have no parts. For an account that lends support to the alternative that atoms are not indivisible magnitudes, see Hasper 1999, p. 4, esp. n. 5, and p. 12.

7. Intrinsic and Relational Properties of Atoms in the Democritean Ontology

169

no regard taken of relations among atoms, there is no hint of interactive powers, abilities, or dispositions that might compromise the fourth requirement. There is, however, an important respect in which the fourth requirement is enhanced by the stipulation of an infinite variety in atomic shapes. This actual infinity of types would certainly imply that all possible shapes are realized in the elements of the Democritean universe – the term “universe” understood, of course, as referring not to a particular kosmos, “world,” but to the totality of things, the totality of worlds. At least with respect to atomic shapes, there is no distinction in the universe of Democritus – as there is also no distinction within the “what-is” of Parmenides – between possibilities and actualities. This global or collective completeness of “the all” may or may not correspond to the completeness of Parmenidean “what-is” (the uncertainty here reflects problems in the interpretation not of Democritus but of Parmenides); it is nevertheless significant that in the atomic theory the requirement of completeness is fulfilled both distributively and holistically. The same double fulfillment may not, however, be claimed with respect to atomic sizes. These too are presumably infinite, notwithstanding the fact that “no source explicitly states that they [Leucippus and Democritus] argued from the Principle of Sufficient Reason to an infinity of gradations of size.” 33 For, as we saw earlier, there is a pattern in our sources of mentioning “shape” in synecdoche for “shape and/or size.” This unlimited diversity of atomic sizes, however, must obviously have both a lower and an upper bound. There certainly could not be an atom of infinite size; that would be tantamount to denying the existence of infinite void. Moreover, the whole thrust of the atomic theory is that there are certain bodies that are smallest. So, the infinite variety of shapes ranges between the largest and the smallest type of atomic size – these upper and lower bounds being set arbitrarily, at the cost of Democritus’ consistency in adhering to the Principle of Sufficient Reason. II. Re la tio nal prope rtie s At this level, properties involve reference to at least one other atom, beyond the atom that bears the property at issue. No relation between two, let alone more than two, atoms is permanent. Having concluded our exploration of the first level of the Democritean scheme (IA and IB), we have left behind us those properties of the atom that are both fixed and occurrent. Fixity at the second level will apply to dispositions as dispositions and to determinables as determinables.

33

Taylor 1999, p. 173. Cf. McKirahan 1994, p. 308.

170

Alexander P. D. Mourelatos

IIA 1. Dispositiona l a nd shape -size indepe nden t The distinction drawn at level IB between uniform and multiform properties has two counterparts at level IIA. First, we need to distinguish between dispositions that are “univalent,” in the sense that they express themselves (are actualized) in only one way, and dispositions that are “multivalent,” inasmuch as they express themselves (are actualized) in an unlimited variety of ways. Moreover, the diverse shapes and sizes established at level IB2 give rise to certain corresponding dynamic properties, as atoms of specific shapes and specific sizes exhibit patterns of array with one another. I shall speak of the latter group of dynamic properties as ones that are “dependent on shape and size.” But there are also dispositions that are dependent on size alone; and, finally, there are some that are independent both of shape and of size. Understandably, both those dispositions that are dependent on shape and/or size, as well as those that are dependent on size alone are multivalent. But within the domain of dispositions that are shape-size-independent it is of notable interest to apply the univalent/multivalent distinction. When we do so, it emerges that there is just one such disposition, viz., the property of impassivity (apatheia), in the second of the two senses distinguished above (IB1), reactive impassivity. No atom can cause another atom to undergo internal change; every atom is immune to internal change regardless of the behavior of other atoms with respect to it. Specifically, an atom’s shape cannot be deformed by the impact of another atom upon it; nor is it possible for one atom to penetrate another atom even slightly. The only way this disposition is expressed is with the atom’s steady retention of its shape and maintenance of its absolute distinctness from other atoms. The remaining shape-size-independent properties are all multivalent. Let me first take up the property of WURSƝ. When Aristotle refers to this property as thesis, “posture,” he is obviously thinking of the actual, occurrent tilt or slant that differentiates, as in his example, an eta from a iota (N from a Z in the translation). But the pointedly more dynamic nominalization WURSƝ, “turning” or (as I have rendered it earlier) “bearing,” is best taken as envisaging the full capacity of an atom to assume an unlimited variety of different tilts vis-à-vis another atom or reference-providing complex. The atom’s dispositional bearing is a permanent and essential characteristic of it: at any time an atom has sometilt-or-other in relation to its surroundings. Firmly grounded as a disposition and as a determinable, the property of “bearing” will allow the full gamut of its determinates (as “posture,” “tilt,” or “slant”) to play a decisive role at level IIB. The second shape-size-independent property is the one we had some conceptual difficulty placing at level IA: dispositional motion; the sheer mobility of atoms. In the first instance, it is the ability of a given atom to acquire motion (some-motion-or-other) upon collision with another atom that is already moving (in some-way-or-other). But it seems logical to extend this property to encompass also the generic capacity to rest, i.e., the ability of a moving atom

7. Intrinsic and Relational Properties of Atoms in the Democritean Ontology

171

(regardless of the specifications of that motion) to come to rest upon contact with another atom or group of atoms. On that basis, we can say that motion is an essential and determinable atomic property: at any time an atom is either at rest or in some-motion-or-other vis-à-vis another atom or reference-providing complex. Let me emphasize the shape-size-independent character of this property, as we shall soon have occasion to introduce the shape-size-dependent dispositional property of motility, which is the ability of atoms of a specific shape and size to move in a special pattern. Quite apart from that further distinction between mobility and motility, the distinction between dispositional and occurrent motion allows us to capture with greater precision the peculiar status of motion in the Democritean ontology. It is the occurrent motion of atoms that is wholly extrinsic. Mobility, however, implies a reference not to atoms per se (to any single atom as an individual) but to the relational complex of some-atoms-or-other that are in someway-or-other dispersed in the void. This distinction may help explain why some of the testimonia appear to suggest that the void is not only a necessary but even a sufficient condition for the motion of atoms 34. Speaking generally – that is, outside the context of Democritean atomism – the circumstances that may be regarded as jointly sufficient for the presence of an ability are far too numerous and complex to permit a priori determination and closure. 35 With atoms and the void, the situation is so starkly simple that there is no such indeterminacy. The void is not only the one and only necessary condition (as Democritus sees it) for the actual, occurrent motion of atoms; it is also the sufficient condition for the atoms’ mobility. Put in another way, the void is only a necessary, not a sufficient, condition for any determinate motion of a given atom; it is, however, a sufficient condition for the determinable motion of any one atom, i.e., for the fact or circumstance that at any given time a particular atom is either at rest somewhere-or-other in the void, or moving in some-way-or-other through the void. The point can be made yet more elegantly by exploiting the contrast between perfective (aoristic) and imperfective verb aspect in Greek: the void is 34

Aristotle, Phys. VIII.9.265b24 (= DK68A58): įȚ੹ į੻ IJઁ țİȞઁȞ țȚȞİ૙ıșĮ઀ ijĮıȚȞ, “they [Leucippus and Democritus] say that there is motion because of the void.” Aristotle has just referred to the Love and Strife of Empedocles and to the Mind of Anaxagoras as efficient causes. So, it is clear that, at least in this passage, a similar role is being ascribed to the Atomists’ void. Cf. Simplicius In Phys. IV.1.209a18 (= Lur’e 1970, 251, quoted in the note 36 here). For discussion, see: Guthrie 19621982, II (1965), pp. 398-99; Löbl 1989, pp. 72-73; Taylor 1999, pp. 194-95; Berryman 2003, pp. 183-91; cf. Hasper 1999, p. 7 n. 12. 35 Consider: What is the set of conditions that is sufficient for Jones’ ability to play the piano? or for sugar’s being soluble in water? or for the flammability of paper? The safe recourse in such cases is to the epistemic test, the one that demands the relevant performance: “Proof is in the pudding;” ab esse ad posse valet consequentia. Ontologically speaking, each necessary condition bears an “at least this” qualifier, and there is generally no a priori criterion for closing the set of necessary conditions just short of the relevant performance, at the immediately antecedent terminus that would determine purely the configuring of the ability.

172

Alexander P. D. Mourelatos

certainly never WR NLQƝVDQ, “the thing that did/does the moving,” but it may quite legitimately be viewed as to kinoun, “the thing that makes for moving.” 36 The third shape-size-independent property at level IIA1 is antitypia, the capacity of atoms to rebound after collision. Its multivalence resides in the unlimited ways in which the property is expressed in the circumstances of collision. We shall shortly attempt to encompass that variety under ten very broad but nonetheless distinct types. But it is relevant to consider, first, whether the reactive impassivity of atoms is compromised by antitypia. Certainly, if we should think of the collision between atoms A and B in the familiar terms of Newtonian physics, a full panoply of dynamic concepts come into play: action, reaction, elasticity, force, counter-force, transfer of momentum, and more. In these modern terms, it is obvious that bodies A and B have both undergone some internal change as a result of the collision. But it is not right to think of the collision of Democritean atoms in these terms. Let me first set aside some potentially confusing uncertainties. With one single exception, all testimonia that speak of contact and subsequent rebound appear to assume full contact and, if there is no rebound, interlocking. The one exception is John Philoponus, who reports that the closest possible “contact” between two atoms involves some measure of separation between them – else the two atoms would be one. 37 This uncertainty need not be resolved here. We can simply think of antitypia as the property that makes for rebound either upon contact or upon very close approach – I shall henceforth refer to either of these as “contact.” What is crucial is that rebound changes none of the intrinsic properties – whether uniform of multiform – of the atoms. All that is changed is occurrent motion; but that is an extrinsic property. Indeed, antitypia has scope much wider than that of rebound. Suppose Democritean atoms A and B are moving toward one another and then come in contact – what might be the possible types of outcome? We need to distinguish 36

37

In a discussion, at Simplicius In Phys. IV.1.209a18 (= Lur’e 1970, no. 251), of the putative causal efficacy of topos, “place,” Eudemus is quoted putting an imaginary question to Democritus (on the assumption that “place” is equivalent to Democritus’ “void”). Note the use of the aorist participle (țȚȞોıĮȞ) to convey perfective (inceptive) aspect: ਕȜȜૃ ਛȡ੺ Ȗİ, ijȘı઀Ȟ, IJઁ țȚȞોıĮȞ [scil. IJઁȞ IJંʌȠȞ șİIJ੼ȠȞ]; ਲ਼ Ƞ੝į੻ Ƞ੢IJȦȢ ਥȞį੼ȤİIJĮȚ, ੯ ǻȘȝંțȡȚIJİ; įİ૙ Ȗ੹ȡ țȚȞȘIJȚțઁȞ İੇȞĮȚ țĮ੿ ਩ȤİȚȞ IJȚȞ੹ į઄ȞĮȝȚȞ. “But are we to posit place, he [Eudemus] says, as the thing that brings on the moving? or is that too likewise impossible, Dear Democritus? For it must be generative of motion and must have some such power.” (My translation. Urmson’s otherwise fine translation in the “Ancient Commentators on Aristotle” series fails to capture the force of the aorist, and consequently the reference in this passage to occurrent, determinate motion: Urmson 1992, p. 26.) The sense in which the void is a cause of motion, without being IJઁ țȚȞોıĮȞ, “what brings on motion,” is accurately explicated by Aristotle himself at Phys. IV.7.214a24-25: Į੅IJȚȠȞ į੻ țȚȞ੾ıİȦȢ Ƞ੅ȠȞIJĮȚ İੇȞĮȚ IJઁ țİȞઁȞ Ƞ੢IJȦȢ, ੪Ȣ ਥȞ મ țȚȞİ૙IJĮȚ . . . “They [the Atomists] consider the void a cause of motion in this way, as that in which [something] moves.” For a different solution to the problem of Aristotle’s ascribing to the Atomists the doctrine that void has causal efficacy, see Berryman 2003, pp. 187-90. DK67A7. See Taylor 1999, pp. 84 and 186-88. “[T]here is a wealth of references in Aristotle and others to the possibility of touching and colliding atoms,” observes Hasper 1999, p. 7; and cf. his n. 13.

7. Intrinsic and Relational Properties of Atoms in the Democritean Ontology

173

between cases of clean rebound, of mere juxtaposition, and of entanglement. This yields, surprisingly, no fewer than ten possible types of outcome – only some involving rebound but all being the causal offspring and direct expression of antitypia. • Assuming that there is rebound: (1) Atom A carries atom B forward; (2) Atom B carries atom A forward; (3) The two atoms rebound and thus reverse their respective pre-collision courses; (4) Each of the two atoms rebounds at a certain angle with respect to its pre-collision course. • Assuming that there is no rebound, the two atoms simply making contact without forming a stable compound: (5) Both atoms stop moving; (6) The two atoms continue moving in a tandem motion. 38 • Assuming that the two atoms join to form a stable compound: (7) The compound itself does not move; (8) The compound continues to move in the pre-collision course of atom A; (9) The compound continues to move in the pre-collision course of atom B; (10) The compound moves in a pattern and course different than that of the pre-collision course either of A or of B.

The specific outcome in any case of contact between moving atoms is determined, of course, by the shape, size, tilt, and the direction and speed of the atoms involved. In closing this section, let me revert to the point made in the first of the sections concerning the relational properties of atoms. Dispositional and determinable properties certainly violate the fourth of the Parmenidean requirements, “fully actualized.” It must nonetheless have struck Democritus as significant that variability and multivalence are grounded in dispositions which, in themselves, are as permanent and as fixed in their nature as any of the intrinsic properties of the atoms. IIA 2. Dispositiona l a nd siz e -dep enden t It is at this level that the property of atomic weight comes into play. For this is the only property of Democritean atoms that is dependent purely on size. There has indeed been much debate and some confusion in modern scholarship, because of discrepancy in our sources on the topic of atomic weight. The 38

My thanks to Richard McKirahan for pointing out this possibility, one I had initially overlooked.

174

Alexander P. D. Mourelatos

emerging consensus appears to be that Democritean atoms have weight in a sense that corresponds – roughly – to the modern concepts of mass, momentum, and kinetic energy: upon collision, larger atoms have a stronger impact on smaller atoms than is true vice versa. In regions of atomic congestion, or in compounds, the larger atoms will tend to impart the direction and speed of such motion as they possess (as a result of whatever history of prior collisions) to the smaller ones. What Democritean atoms almost certainly do not have, as was pointed out earlier, 39 is weight in the Aristotelian-Epicurean sense of an intrinsic precosmic tendency “to fall downwards.” 40 Since atomic size is multiform, ranging in variety between the lower and upper bounds cited earlier, atomic weight exists in a correspondingly unlimited variety of types between those same two bounds. But unlike atomic size, which is an intrinsic property, atomic weight is pre-eminently a relational concept. Accordingly, the effects of some particular atomic weight will vary with the circumstances of the collision or contact and with the previously acquired motion of the atom whose weight is at issue. Specific atomic weights share in this respect the multivalence of properties at level IIA1b. If a large “heavy” atom is moving at very low speed just behind a marginally slower small atom, the two of them on the same line of motion, the smaller one will increase its speed only slightly after the larger one catches up with it and the two make contact, and there will be no change of direction for the smaller atom. Under the same conditions of shared line of motion, the same large atom moving at high speed will drastically change the speed of the same smaller atom (forcibly carrying it forward); and there will also be change of direction for the smaller atom if the collision takes place at an angle. The effect of the small atom on the large atom, in either case under these same conditions, will be negligible. IIA 3. Dispositiona l a nd dependen t on shape an d/or size The first of two properties at this level I shall call conjunctivity – postponing briefly taking up the question whether Democritus himself may have had a corresponding term. Every combination of different atomic shapes and different atomic sizes logically generates a repertoire of possible arrays with other atomic shapes and sizes. Modern LEGO sets provide an obvious model for this, albeit one that requires immediate adjustment in its application to the Democritean context. It is clear that individual LEGO pieces have a distinct repertoire in the exercise of their ability to form “joins;” and that repertoire can be either quite small or quite large, yet still finite. In the case of Democritean atoms, however, given infinite diversity in shapes and sizes, the repertoire is infinite both for those looser cases of array that involve no more than contact 39 40

See above at IA. See Morel 1996, pp. 75-83; Salem 1996, pp. 89-95; Taylor 1999, pp. 179-84.

7. Intrinsic and Relational Properties of Atoms in the Democritean Ontology

175

(in the sense of the term defined above) and in the case of joins. Nonetheless, as with the manageably finite repertoire of LEGO pieces, any one of the possible arrays of Democritean atoms, like any of the possible atomic shapes, sizes, and combinations thereof, can be distinctly specified and envisaged. An atom that has the stellated geometry of a sea urchin can be entangled in an immense variety of ways either with a like-shaped and like-sized atom or with one that has, say, the shape and size of a cone that can fit between the stellated atom’s spines. On the other hand, a stellated atom affords almost no possibility of a join with a like-sized spherical atom. Small spherical atoms, however, could lodge themselves between the spines of a large stellated atom. I have introduced conjunctivity as the “repertoire” a given atomic shape and size offers “of possible arrays with other atomic shapes and sizes.” But we can also think of it as a property that binds similar atomic shapes and sizes. For this sub-type of conjunctivity we have fairly detailed testimony: atoms of like shape and like size tend to aggregate. This appropriation by Democritus of the familiar “like to like” principle plays a major role in his cosmogony (DK67A10, A24). 41 The principle was evidently supported by analogy with patterns of aggregation observed in ordinary objects, both animate and inanimate: doves with doves, cranes with cranes; in a sieve, lentils with lentils, barley with barley, wheat with wheat; on a beach, oblong pebbles together, round pebbles together (DK68B164, A38, A128). 42 The second of the properties at level IIA3 I shall call motility. It is evident that the specific pattern of motion characteristically exhibited by an atom depends in some crucial respects on its shape: a spherical atoms rolls smoothly; an atom that has the shape of a cube or prism tumbles or slides; an atom of irregular contour is likely to exhibit an irregular path of motion as it is twirled and twisted after collisions. All of these effects are compounded and enhanced by atomic size. The two factors, shape and size of a given atom, taken together, make both for the tendency of that atom to display a certain geometrical pattern in its motion and for a certain tendency of speed. On the whole, small spherical atoms will tend to move at high speed. The reason for this tendency is twofold: the smallness and rolling surface of such atoms allow them to escape unhampered through areas of atomic congestion, and even through the interstices of compounds; and the cumulative effect of those collisions which the small spherical atoms do not escape can only be one of successive increases to their speed. Large stellated polyhedrons, by contrast, will tend to be sluggish, as their many protuberances offer frequent occasions of entanglement, thus constantly detracting from whatever speed they may have acquired. We saw earlier that Democritus’ WURSƝ corresponds to Aristotle’s thesis as disposition does to occurrence – “bearing” and “tilt,” respectively. The two 41 42

Cf. Taylor 1999, pp. 94-95. Cf. Taylor 1999, pp. 4-5, 121.

176

Alexander P. D. Mourelatos

concepts of motility and conjunctivity now make it possible for us to make sense of the other two pointedly dynamic Democritean terms which, in Aristotle’s account, are summarily equated to “shape” and “array.” The concept of motility provide us with an insight into Democritus’ intriguing choice of the term rhysmos for atomic shape. When the level of consideration is that of individual atoms considered in themselves, the proper term is, of course, idea (or, in Aristotle’s terminology, VFKƝPD). But when the level is that of atoms in relation to other atoms (which is, as we saw, the context in the Aristotle passage in which the term is quoted), rhysmos is strikingly the better term. My suggestion is that rhysmos was Democritus’ term precisely for “motility,” the pattern of motion determined by shape-size. The noun GLDWKLJƝ, best identified as derived from the verb diathinganomai, “to touch and be touched, to be in mutual contact,” 43 should logically have both a dispositional and an occurrent use. In the latter sense, as “junction,” it does indeed serve as a synonym of Aristotle’s taxis, “array.” But in the dispositional sense, as tendency or capacity for frequent and reciprocating contact, the term is aptly the Greek equivalent for the concept of conjunctivity introduced earlier in the present Section.44 The full logical space of types, sub-types, and species of motility and conjunctivity is, of course, infinitely vast. And yet, this nearly dizzying proliferation of logical possibilities should not obscure the fact that each token atom, given its particular combination of atomic shape and atomic size, has a distinctive combination of rhysmos, “motility,” and GLDWKLJƝ, “conjunctivity,” as its fixed and permanent dynamic endowment. That particular combination of dispositions is as unborn, imperishable, and unchanging – as “Parmenidean” with respect to these attributes – as is the shape and size of each token atom. But there is also a noteworthy reversal in the Parmenidean affinities of properties as we pass from level I to level II. At level I it was the properties of shape and size that introduced deviation from the Parmenidean requirement of indivisibility-homogeneity. At level II, while mobility and the other two multivalent properties hugely expand ramifications into the realm of plurality 43

The verb occurs in a medical context at 634a9 of the History of Animals in the Aristotelian corpus – in book X, which is regarded as spurious. The passage at issue draws on the widely held (false) doctrine of the wandering womb: “if the womb remains close [to the vagina] and is not capable of withdrawing, it will be less responsive because of recurrent touching [by the penis] (țȦijંIJİȡĮȚ ਩ıȠȞIJĮȚ, įȚĮșȚȖȖ੺ȞİıșĮȚ ਕİ઀), and will not open promptly [to receive the sperm]).” This occurrence of diathinganomai is missed in many Greek dictionaries; but it is picked up in Bonitz 1870, s.v.; also in Dimitrakos 1950, s.v. A search of the Thesaurus Linguae Graecae database confirms that this is the only attested occurrence of this passive/middle form. 44 Much of the argument in Fritz 1966 bears out strongly the dispositional sense of the noun GLDWKLJƝ. It might then, at first blush, appear surprising that von Fritz concludes his discussion by shifting emphasis to the stative aspect: “In GLDWKLJƝ . . . ist . . . in ganz eigentümlicher Weise der Zustand [von Fritz’s emphasis] des Eingefügtseins, d. h. des “Sich in einer Anordnung Befindens,” also etwas, das wir eigentlich als etwas Statisches empfinden, als Moment in einer Bewegung gefasst” (p. 28). I count this explication as enhancing the evidence for the dual character of the term.

7. Intrinsic and Relational Properties of Atoms in the Democritean Ontology

177

and change, it is those initial Parmenidean “apostates”, shape and size, that reintroduce principles of constancy. In this we have an aspect of what was referred to, at the start of this paper, as the “gradualist” character of the Democritean project. IIB. Occu rren t and variable It is only at this level that we build distance from the Parmenidean requirements. Atoms frequently change their position and the direction and other parameters of motion through the void. Indeed, there is some evidence that they do so constantly: one of our sources (Aëtius, in DK68A47) speaks of atoms having palmos, “vibratory motion,” which is best understood as the “bouncing back and forth of the atoms between collisions.” 45 These are all cases of occurrent, determinate motion. Moreover, as atoms make or unmake contact, the occurrent relational properties of array and posture change. I revert here to the Aristotelian terms of Metaph. I.4 advisedly, since at this stage the purview matches Aristotle’s assumption that the “differentiations” are to be taken as occurrents: and these are precisely the determinates that fall under the Democritean dispositions and determinables of conjunctivity and bearing (diaWKLJƝ and WURSƝ). The last of these determinates, WURSƝ, understood as “tilt,” now plays a decisive role. For whether two atoms may ultimately either join or collide depends totally on their tilt. However much their respective conjunctivities may favor their junction, if their mutual approach is not at the appropriate tilt, no join will result, perhaps not even an array in the weaker sense of mere juxtaposition. Likewise, however much their respective motilities may favor collision, if the tilt does not put them on a collision course, they will just bypass one another. Moreover, if tilt brings them in contact, it will play an important (though not exclusive) role in determining the outcome of contact, viz., the direction and speed of the rebound. *** We are now in a position to survey, in an overall outline, the entire logical scheme of atomic properties in Democritus. The earlier outline had the rationale of highlighting the distinction between intrinsic and relational properties. If we should now take as our major divide the distinction between fixed and variable properties, the outline should be drawn as follows:

45

McKirahan 1994, p. 319.

178

Alexander P. D. Mourelatos

THE LOGICAL TIERING OF PROPERTIES OF ATOMS IN DEMOCRITUS: FROM FIXED TO VARIABLE I. FIXED A. Occurrent 1. Uniform a. Fullness, rigidity, solidity 2. Multiform a. Shape b. Size B. Dispositional 1. Univalent a. Impassivity (except for rebound) 2. Multivalent a. Dependent on shape and size i. Conjunctivity ii. Motility Į. The tendency of atoms of like shape/size to aggregate ȕ. The kinematic disposition of different atomic shapes/sizes b. Dependent on size alone i. Weight c. Independent of shape-size i. Bearing ii. Mobility iii. Capacity to rebound II. VARIABLE A. Occurrent 1. Determinate posture or tilt (cf. Aristotle’s thesis) 2. Determinate motion 3. Determinate array or join (cf. Aristotle’s taxis) The above outline brings out strikingly the gradualism and progressivity in the transition from fixed to variable, from simple to complex, from fulfillment to non-fulfillment of Parmenidean criteria – in sum, from Being to Becoming. Note how all the dispositional properties retain in their own way the fixity of the occurrent properties; how the first of the dispositions recalls the character of the first of the occurrent properties; how the shape-size dependent dispositions both look back to the multiform occurrent properties but also, in their multivalent character, mediate the much greater variability represented by the

7. Intrinsic and Relational Properties of Atoms in the Democritean Ontology

179

lemmata that follow; and, finally, how the scope of variability is at its freest in the case of dispositions that are shape-size independent. And yet, even after we pass to consideration of variable and occurrent properties (IIA above, stage IIB in the earlier outline), we are only at the threshold of Becoming. For we are still, after all, speaking strictly of properties realized in atoms. To be sure, all the atomic properties at issue are utterly ephemeral; at any time they are subject to change. What gives us license, nonetheless, to speak of a “logically progressive transition” is that all these changes in determinates are traceable to collisions between or among atoms, and that these events follow a deterministic logic in which geometry and kinematics are the only factors at play. For one who takes the longest possible perspective – and for Democritus that would be nothing short of the temporal infinity of the universe – the two properties in the middle of the outline above, motility and conjunctivity, serve as crucially invariant factors. An atom in its career through infinite time will have assumed every spatially possible tilt, and it will have undergone every spatially possible collision with other atoms. Apart from its intrinsic properties throughout this career in eternity, there are only two other constants: motility and conjunctivity. One might adapt Heraclitus’ saying, “character (ƝWKRV) is a man’s fate-dispensing deity (GDLPǀQ),” to coin a metaphysical slogan for Democritus: “shape and size are an atom’s destiny.” *** At the start of this essay, intrinsic properties were defined as those which even a single solitary atom would possess, specifically in the absence of other atoms. The relational properties may correspondingly be defined as those which atoms possess even in the absence of compounds. The next stage in a logical reconstruction of the Democritean ontology would, accordingly, take up what might be called “colligative” powers of atoms, viz., those properties that can be conceptualized insofar as atoms exist in certain arrays within compounds. It is at that level that the Democritean project provides explanatory constructions (still in essentially geometric terms) of effects and phenomena at the macro-physical level: e.g., the combination of heaviness with softness or brittleness that is characteristic of lead; the lesser heaviness but greater hardness and rigidity of iron; the friability of certain substances; the glutinous texture, or the acidic properties, of others – to mention just topics that come up in Theophrastus’ account of Democritus in De sensibus. 46 And it is likewise by recourse to such colligative properties that Democritus can claim, as Theo-

46

Reproduced in DK68A135; cf. A129-133. See McKirahan 1994, pp. 330-33; cf. Cartledge 1997, pp. 27-30.

180

Alexander P. D. Mourelatos

phrastus’ account would suggest, some success in “reducing” certain of the effects of taste and color to atomic structures. Investigation into that third level of the Democritean scheme would require a much longer study, in which analysis of the Democritus section in Theophrastus’ De sensibus would be central. But even at the two initial stages explored here, one may discern the potential of logically progressive advance from the domain of atoms to the complex and variegated “Manies” of the world of ordinary-size objects. For, as careful readers of the De sensibus would appreciate, much can be metaphysically constructed (or, proceeding top-tobottom, much can be metaphysically reduced) before Democritus should find himself compelled to have recourse to the move characteristic of “eliminativist” materialism or physicalism – scil., of declaring the “raw feels” of sensation as existing only “by convention (QRPǀL, DK68B9 and cf. 68A49).” 47

47

My sincere thanks to Ricardo Salles, editor of the collection in which the present study was originally published, for helpful comments and suggestions on an early draft of it. – As indicated passim in the notes above, I have also sought to make further improvements on that early draft by taking into account detailed critical comments by Richard McKirahan; and I am happy to record again here my warm thanks to him.

Part III: The Sophist’s Demurrer

8. Gorgias on the Function of Language

Gorgias made his fame as an orator, teacher of rhetoric, and diplomat. He was certainly fully conscious of the special genius of his art. This is shown by the panegyric on the powers of logos he inserted in one of his model speeches, the Encomium of Helen. There the power of logos is compared to that of divine intervention, physical compulsion, overwhelming passion, magic, or the influence of drugs. And yet, the same Gorgias, who also authored On Non-Being or On Nature, argued in the third part of that treatise that logos can never convey to a hearer the information a speaker may possess – in effect, that intelligent verbal communication is impossible. The most satisfying reconciliation of these seemingly antithetical conceptions and assays of logos was offered in one of the twentieth-century classics of Presocratic scholarship, Guido Calogero’s Studi sull’Eleatismo (1932). Calogero argued that Gorgias’ epistemology is indistinguishable in its thrust from that of Protagoras: Gorgias too is a relativist. And so, Calogero wrote, for Gorgias “logos [does not have] the theoretical function of mirroring reality, or mirroring our knowledge of reality; it has merely a certain practical and rhetorical function, viz., to put the mind in a [certain] affective state through persuasion.” 1 The panegyric of logos in Helen is, accordingly, the positive and declaratory side, which is complemented by a negative and elenctic side presented in the third part of Non-Being. Though I did not set out to vindicate the Calogero solution, 2 the actual results I have reached in this study can be suggestively viewed as though In its earliest version, this study was presented in Greek at the First International Symposium on the Sophistic Movement, organized by the Greek Philosophical Society and held in Athens 27-29 September 1982, and later published in the proceedings of the Athens conference: “੘ īȠȡȖȓĮȢ ȖȚ੹ IJ੽ ȜİȚIJȠȣȡȖȓĮ IJોȢ ȖȜȫııĮȢ” in ‫ܻ ݠ‬ȡȤĮȓĮ ȈȠijȚıIJȚțȒ: The Sophistic Movement (Athens, 1984), pp. 22331. An intermediate version was presented at the International Congress on Gorgias, held in December 1983 at Lentini and Catania, Sicily, and then published in Gorgia e la sofistica, eds. Luciano Montoneri and Francesco Romano, in the series Siculorum Gymnasium N. S. a.XXXVIII, nos. 1-2 (1985). In expanded and revised version it was subsequently published in Philosophical Topics (Mourelatos 1987). It is the latter that is reprinted here, with the kind permission of the editors of both the Siculorum Gymnasium series and of Philosophical Topics. 1 2

Calogero 1977, p. 262. Indeed, though Calogero is surely right in seeing Gorgias as a relativist, in a suitably broad sense of the term, the Protagorean homo mensura does not, and properly cannot, play any role in Gorgias’

184

Alexander P. D. Mourelatos

emerging from an attempt to develop the insight contained in the remark of Calogero’s just quoted. My overall argument is that in the third part of NonBeing Gorgias attacks two captivating conceptions of the nature of linguistic meaning, viz., that meaning is reference, and that meaning is mental image or idea. The attack is in the form of a series of puzzles. These are by no means “sophistic,” in the pejorative sense. Indeed, the puzzles have recurred in the history of philosophy and – in their modern formulations, as I argue – quite similar puzzles have played a role in the development of twentieth-century philosophical analysis. The panegyric of logos in Helen presents, in effect, a third alternative, a behavioral conception of meaning. The three alternatives – referential, ideational, and behavioral – constitute a systematically coherent triad, as is suggested by the fact that the three, in respectively various guises or variants, have classically been the major contenders in attempts by linguists and philosophers to offer an account of the nature of meaning. Parts I-III of this paper are devoted to an analysis of the argument in the third part of Non-Being. The better source, unquestionably, for this part of the treatise is the paraphrase by the anonymous author of the Peripatetic work, De Melisso, Xenophane, Gorgia (MXG hereafter, 980a19-b21). 3 The corresponding paraphrase in Sextus (Adversus mathematicos VII.83-87) omits entirely the second prong of Gorgias’ elenchus, the puzzles of perceptual sameness. I am not, however, persuaded by Newiger’s argument4 that Sextus is completely dependent on MXG. Accordingly, I shall not desist from also making use of the evidence in Sextus – both parallel passages and material for which there is no counterpart in MXG. 5 To anticipate a result that seems especially intriguing: the logos passage in Helen may not be our only evidence of Gorgias’ endorseargument concerning incommunicability: see below, note 31 and corresponding remarks in the main text. 3 For the text, which is not included in Diels-Kranz, see Diels 1900 or Untersteiner 1961. The latter prints essentially the text of Diels 1900. For a newer critical edition, with French translation and extensive commentary, see Cassin 1980. Unfortunately, Cassin, who analyzes the text with notable philosophical acuity and literary sensitivity, adheres to ultra-conservative principles of editorship, often preserving readings that are manifestly corrupt. The text of MXG is so poorly preserved that extensive philological restoration is unavoidable. It is well to remember that one of the scribes felt desperate enough to write on the margin at the end: “The original contains many errors; no one should blame me; I just copy what I see (țĮșઅȢ Ȗ੹ȡ ੒ȡ૵ Ƞ੢IJȦ ȖȡȐijȦ).” – For English translations of MXG, see Loveday-Forster 1913, in the Oxford translation of Aristotle; also Hett 1936 in the Loeb Aristotle. – It has come to be apparent that the author of MXG has Pyrrhonist leanings, and that parts of the treatise (not, ostensibly, the parts on Gorgias) betray the influence of a Middle Platonist or Pythagoreanizing source. Nonetheless Mansfeld 1985, who has pointed up these later influences (pp. 243-44), has also stressed that there are enough parallels between MXG and either Gorgias’ epideictic speeches (Helen, Palamedes) or other fifth- and fourth-century sources to give us a good measure of confidence that MXG has indeed captured authentically Gorgianic themes and arguments. 4 Newiger 1973, passim, but esp. pp. 161-70. 5 The text of the Sextus paraphrase is most easily available in Diels-Kranz (82B3) and in Bury 1935. – For criticism of Newiger’s thesis of dependence, see Guthrie 1975, pp. 706-07; Graeser 1978, pp. 64-69; Kerferd 1981b, p. 321; Mansfeld 1985, pp. 244, 260, 262.

8. Gorgias on the Function of Language

185

ment of a behavioral conception of meaning; there is also evidence of that endorsement in a passage in Sextus. I. THE ARGUMENT FROM CATEGORIES: MXG AND SEXTUS The first argument of the third part of Non-Being starts inauspiciously in both versions. In MXG 980a19-21 we are asked to envisage a situation in which a speaker has seen and knows certain things which the listener has not seen (DNRXVDQWL    PƝ LGRQWL). 6 How can the speaker under these circumstances, asks Gorgias, make the object of his visually acquired knowledge evident (GƝORQ) to the listener? The corresponding passage in Sextus elaborates: external realities are visibles, audibles, and generally perceptibles; but how can visibles be conveyed through audibles, or vice versa (Adv. math. VII.83)? In either version, the argument, so far, is pathetically weak. For it appears to concede that logos, being audible, may well serve to convey realities in the domain of sound. So, there would, after all, be the possibility of successful communication whenever onomatopoeia can be employed. Moreover, Sextus’ “vice versa” would encourage one to speculate that an ideographic supplement to speech would go at least some way toward solving Gorgias’ problem of communication. Even more disappointingly, Sextus recurs to this weak reading of the argument in the remark “the visible can be grasped by one organ, and logos by another,” which is immediately followed by the jejune gloss: “many of the external realities logos fails to show” (VII.86). 7 It is fair to assume that Gorgias’ elenchus was far more radical. A syntactic feature common to the two versions gives the relevant clue: The doctrine of distinctness of sensory domains provides not an explanation of the communicative inadequacy of logos; it provides an analogy. Interestingly, the wording “just as . . . so too” appears in both our versions (KǀVSHUKRXWǀ in MXG 980b1; kathaper . . . aQDSDOLQ    KRXWǀV in Sextus VII.84 ad fin.). 8 Gorgias is arguing on the basis of, what we would call, categorial distinctions. Here is how the argument proceeds in MXG: The speaker speaks discourse (OHJHLKROHJǀQORJRQ, b2-3, cf. b6). He can no more speak the thing itself, to pragma (b3, cf. a20 touto eipoi) than vision can hear sounds, or hearing can see colors. The logos is something quite other than the thing itself (cf. b4 KHWHUǀL

6

Cf. Loenen 1959, p. 197. Kerferd 1981a, p. 99, endorses this weak reading of Gorgias’ argument. Cf. Kerferd 1984, p. 219: “the majority of such externally existing objects are incapable of being communicated by words” (emphasis mine). 8 The MXG text explicitly compares vision-and-its-objects with hearing-and-its-objects, then implicitly both of these with logos-and-its-objects. In Sextus the analogy between the two sensory domains is marked by enallax (83) and anapalin (84); but the kathaper-KRXWǀV construction develops a point that has no exact counterpart in MXG. More on this below. 7

186

Alexander P. D. Mourelatos

tou pragmatos). For convenience, let me refer to the preceding as Gorgias’ categorial argument. The argument, as I understand it, draws on ordinary intuitions. It has been thought 9 that the argument draws, rather, on a specific philosophical theory, viz., Empedocles’ theory of poroi (“channels,” or “pores” – on which topic see above Chapter 4, Section IX). The present context of Gorgias’ treatise is one of several in which modern scholars have deployed the strategy of importing doctrines of other Presocratics as premises or considerations for Gorgias’ argument. The strategy is in all cases ill-conceived. Given that Gorgias has highly paradoxical conclusions to defend, he would have been foolish to draw on speculative theses that are in themselves controversial. The argument will have force only if it exploits ordinary intuitions or commonly held beliefs. The corresponding part of the argument in Sextus is more fully developed: For something to be communicated to another it must be in the form of logos (VII.84 ad init. and 85 ad init., where I adopt Bekker’s emendation PƝRQ). 10 But, then, logos, not things, is all that we ever communicate, since “logos is not the same as the underlying realities or the actual things” (84 ta hypokeimena kai onta); logos “is quite other than the underlying realities” (84 KHWHURVWǀQK\SRNHLPHQǀQ). Unmistakably present in Sextus is the analogy with the distinctness of sensory domains. Worth notice is that “just as . . . so too” clause makes a certain point in terms of one thing “becoming” another that has no counterpart in MXG. I translate: “And just as the visible cannot become audible or vice versa, so too, since reality obtains outside of us, it could not become our own logos” (84). One may, of course, add the obvious corollary, that our own logos could not become the actual reality when this logos is conveyed to another person. But what is remarkable here is that Sextus has singled out the converse: There is no way for things to be transformed into our logos. If this does reproduce an argument of the original treatise, then Gorgias did not only maintain the inability of logos to communicate reality to another person, he also argued that even for the solitary individual in meditative monologue logos cannot furnish, constitute, or represent the external reality. Astonishing though this extension of the doctrine might be, it is by no means incredible. If the categorial argument is taken seriously, it must apply not only when I use words to convey reality to another but equally when I use words to conjure up a reality that is not immediately present to me. Indeed, we shall shortly discuss another argument, this one unique to MXG, in which problems of perceptual sameness are raised in intra-subjective as well as in 9

10

Gigon 1936, p. 110; cf. Kerferd 1984, pp. 323-24; Mazzara 1982, p. 137 and n. 79. For criticism of Gigon, see Sicking 1976, pp. 395-96. I presume this is also the construe of Bury 1935, who translates “and not being [subject ‘the existent,’ to on] speech . . . .,” even though he prints PƝǀQ[subject would have to be logos]. Cf. George Kennedy in Sprague 1972, p. 46; Robinson 1968, p. 298; Barnes 1979a, II, p. 169; and Kerferd 1981b, p. 216. See also Newiger 1973, p. 163 n. 45.

8. Gorgias on the Function of Language

187

intersubjective situations (980b14-17). A Gorgias who was alert to these issues may well have also realized that the categorial argument would tend to undercut both communication and the possibility of a private language. This is a fascinating issue to which we shall return. The next six lines in MXG, 980b3-8, are riddled with problems. In the MSS there are three lacunae of between three and five letters, and there are word fragments in the vicinity of the lacunae. The most convincing reconstruction remains essentially that in Diels 1900, who draws on Cook Wilson 1892-1893 for filling out the lacunae: Indeed (oun), that which a man is not aware of (ennoei), how – Gorgias demands (aitei) – will such a man come to grasp it mentally (HQQRƝVHL) from someone else by means of discourse (ORJǀL), or by means of a sign, something other than (VƝPHLǀLKHWHUǀL) the thing itself – except (DOO¶Ɲ) by virtue of having seen it, if [it is] a color, or having heard [it, if it is] a sound. For (gar) it is decidedly not (DUFKƝQRX) that the speaker speaks, nor color, but discourse (logon). It follows (KǀVWH) that the possibility is not one of having the thought of (dianoeisthai) a color, rather of seeing it; and not one of having the thought of a sound, rather of hearing it. 11

Quite apart from the textual uncertainties, Gorgias’ argument seems to have been compressed in the paraphrase, and any fleshing out of it is bound to be speculative. But the genre of the argument is clear enough. As the connecting particles indicate, what we have here is a continuation of, or variation on, the categorial argument. The theme of “awareness” transposes the categorial problem to the hearer’s predicament. If speaker A is to communicate to listener B some entity X that belongs to the real, extra-mental world but is unknown to, or unperceived by, B, there must eventually come a stage at which B forms at least some notion of X. But this would be of no help; for the notion would still be something quite other than the thing X itself. Borrowing a formulation by George Kerferd, we might speak of a second categorial “gulf,” one between things and thoughts. 12

11

I read aitei with the MSS – Diels 1900 adopts the emendation auto, proposed by Apelt. Cassin 1980, whose preserving of aitei prompted me to reconsider whether emendation was necessary in this instance, does not treat aitei as parenthetical; she argues for having aitei govern par’ allou, which results in a construction I find rather choppy (pp. 541, 543-44). Her construction is again unlikely at 980b4, where she rejects the emendation KHWHUǀL in favor of the attested reading heterou. Kerferd 1984, p. 217, makes a more plausible proposal for staying close to the MSS: heterou tou (the tou here being the enclitic equivalent of tinos). But the natural construction in that case would be: “sign of some other thing,” which is unsatisfactory (only one “thing” is at issue). The translation Kerferd actually offers is: “some sign (taken) from something which is different.” 12 Kerferd has Gorgias propounding six “gulfs”: (A) one sense-faculty and another; (B) logos and things; (C) logos and sense-impression; (D) sense-impressions and thoughts; (E) one man’s sense impression and those of another man; (F) one man’s thoughts and another man’s thoughts. See Kerferd 1981b, pp. 324-25; cf. Kerferd 1984, pp. 218- 21.

188

Alexander P. D. Mourelatos

The sense of the verb ennoein obviously shifts slightly, from its first occurrence (“imperfective” present-tense infinitive), where the sense is that of a state, to the second (aoristic, or “perfective,” subjunctive), where the sense is that of an achievement; hence my two different translations of the same verb. In the final sentence, a carelessly literal translation of the modal construction might mislead one into concluding that Gorgias is arguing that it is generally impossible to have the thought (or notion or mental image) of a color or a sound, which would contradict the manifest implication of the initial sentence, that the speaker does have a certain noetic awareness (cf. ennoei), one which the hearer, in turn, is being invited to achieve. Yet even though the genre and thrust of the argument seem obvious, there is a confusing turn at the middle of our passage. The appended “except” (DOO¶Ɲ) clause might well seem to undercut the thesis of incommunicability by allowing for exceptions. On the face of it, the “except” clause appears to be saying that if both speaker and listener have seen (or heard, as the case may be), the thing to which the speaker’s words refer, communication should be perfectly possible after all. 13 Now considering that a large part – perhaps the largest part – of language use is empirically grounded, it would appear that the appended clause makes concessions of truly massive scope. 14 What is more, the clause would go against the grain of Gorgias’ antirealism. It is philosophers like Locke, Berkeley, Hume, or the late nineteenth- and early twentieth-century positivists who have argued that words convey meanings only when they refer to items respectively experienced by the interlocutors. We cannot seriously entertain the possibility that Gorgias endorses the empiricist theory of meaning. 15 That theory makes preeminently the strongest possible claim concerning the success of language in making contact with reality. It holds that when language is properly used, words and things are in a one-toone correspondence; reality is mirrored in language. 16

13

Kerferd tries to account for the exception clause by proposing that Gorgias envisages a situation of simultaneous, or nearly simultaneous, perception of the object at issue by speaker and listener. See Kerferd 1981b, p. 324: “[an unintelligible] shout by one man may alert the attention of the other towards an object, provided it is present to his sight” [emphasis in the original]. But our text reads LGǀQ. . usas (Cook Wilson’s restoration, adopted in Diels 1900), “having seen . . . having heard”; not KRUǀQ . . . DNRXǀQ, “as he currently sees . . . as he currently hears.” 14 This serious difficulty is overlooked by Loenen 1959, pp. 196-99, and Newiger 1973, pp. 155, 164, 175-80. Kerferd is aware of it, but chooses to play it down; cf. preceding note. 15 Yet this is precisely the position of Newiger 1973, p. 179: “given that the experience is grounded in a perception, it is also possible to convey a thought to, and to share it with, another [subject] (auf der Grundlage der auf der Wahrnehmung beruhenden Erfahrung ist es auch möglich, einen Gedanken weiterzugeben und mit einem anderen zu teilen).” Cf. p. 176: “One is tempted to say, with John Locke, ‘There is nothing in the intellect that has not pre-existed in the senses’ (Man ist versucht mit John Locke zu sagen: nihil est in intellectu, quod non fuerit in sensu).” Cf. Mazzara (who adds some qualifications), pp. 136-38, 151-54. Mansfeld 1985, too, appears to assume that Gorgias endorses an empiricist theory of meaning: pp. 252-53. 16 That for many empiricists reality is a construction out of sense data is, of course, a separate issue.

8. Gorgias on the Function of Language

189

To be sure, the concessions made by the “except” clause would be eventually withdrawn; for the upshot of the final three sections in MXG is that speaker and listener can never mentally grasp the same thing. 17 But this consideration would only heighten our difficulties in interpreting b3-8. The thesis of incommunicability is suspended as premise in the sequel, though, of course, it swings back as conclusion. The suspension, signaled by the juridical “and even if” (ei de kai, b8) which is characteristic of Gorgias, would be robbed of much of its logical and rhetorical force by the undercutting of the thesis of incommunicability in the preceding “except” clause. It is possible, of course, that the author of MXG has misstated Gorgias’ argument. But an interpretative move of this sort on our part would be gratuitous. What might be more proper, methodologically, is to allow for a mild anacoluthon in our text. 18 Accordingly, the DOO¶ Ɲ clause does not represent a condition which, when added to the circumstances of the imparting from speaker to listener of a “word or sign,” would lead us to assume that communication has in fact taken place. Rather the DOO¶Ɲ offers a reminder of what is involved in the only case in which there is, for Gorgias, a mental grasp of a pragma: the case of first-hand experience. Indeed, the rhetoric of the passage shows a sharp contrast between derived comprehension (via language) and first-hand experience. The correct syntactic relationship between the main question and the appended clause is, in effect, conveyed by this paraphrase: How would a man come to grasp something mentally? Only after seeing it, if the object at issue is a color, or by hearing it, if it is a sound. How could he, then, grasp it from another by means of language? In no way!

This reinforcement of the categorial argument contributes a poignant paradox: Words are totally superfluous when I have the mental grasp of a thing; and they are totally ineffective when I do not have that grasp. The considerations Gorgias would seem to be bringing forth here have affinities with the sophistic paradox of Plato’s Meno: If you do not know the object after which you inquire, you cannot pursue the inquiry; if you do know it, then the inquiry is pointless. The passages from Sextus that correspond to the MXG section discussed so far are Adv. math. VII.83-84 and 86. Section 85 has no counterpart in MXG, 19 and it makes a strikingly different point, viz., that the relation between logos and thing(s) is the reverse of what it is generally taken to be. The passage is, so 17

18

19

This, too, is overlooked by Loenen and Newiger (see above, nn. 14 and 15). For criticism, see Graeser 1978, p. 69. Here I am developing a suggestion made by my former colleagues Richard McKim and Sarah Waterlow Broadie in the course of a discussion of an early version of this paper. It has been claimed that there is parallelism between MXG 980b3-8 and Adv. math. VII.85. See Newiger 1973, pp. 163-64. But a striking feature of the MXG passage, the repeated occurrence of verbs of mental awareness (ennoei, ennRƝVHLGLDQRHLVWKDL), is conspicuously absent in Adv. math. VII.85.

190

Alexander P. D. Mourelatos

far as I can tell, unique in either the Sextus or the MXG paraphrase in propounding a positive thesis concerning the function of language. In Part VI of this Chapter I shall be in a position to offer a new interpretation of this fascinating passage. The order of business at this stage, however, is for us to examine the later passages of MXG, the ones that have no counterpart in Sextus. II. MXG ON THE PROBLEMS OF SAMENESS IN PERCEPTION: THE CASE OF TWO SUBJECTS In a closing recapitulation of the third part of Gorgias’ treatise, MXG identifies two main arguments in support of the thesis of incommunicability: viz., “that things are not logoi; and that no one mentally grasps the same thing as anyone else” (980b18-19). 20 The first argument is, of course, the categorial one. The second, which encompasses three distinct sections of the MXG text (a primary argument, and then two fall-back arguments), makes an important assumption: communication is possible only to the extent that two subjects can “mentally grasp the same (thing).” The three sections address this issue of intersubjective mental grasp directly, without dwelling on the point – which is, however, brought out in the recapitulation – that sameness of intersubjective mental grasp is the condition posited for communication. 21 Here are the three sections: And if it is possible (for one) to know and to speak that which he happens to know, 22 still how will the listener mentally grasp (HQQRƝVHL) the same thing (to auto)? For it is not possible for the same (thing) to be simultaneously in (subjects) that are more than one (pleiosi) and distinct (FKǀULV); the one would then be two. (980b8-11) Yet even if it should be – he says – in more than a single (subject) and be the same, nothing prevents (RXGHQ NǀO\HL) that it should not appear (phainesthai) alike (homoion) to them, inasmuch as they are not in every respect alike and in the same (place, HQWǀLDXWǀL). For if in the same place, they would be one and not two. (980b11-14) 23 20

One manuscript reads lekta, rather than logous. In the second colon the reading is Ƞ੝įİ੿Ȣ ਪIJİȡȠȞ ਦIJȑȡ૳ IJĮ੝IJઁȞ ਥȞȞȠİ૙, which is preserved by Cassin (1980) and then given a rather willful translation (pp. 554 and 559). Diels (1900), sensibly, follows the early nineteenth century editor Foss in deleting heteron. 21 Presumably Gorgias, or the anonymous author of MXG, avoids dwelling on the obvious. The absence of explicitness has prompted Newiger (1973) to make the rather misleading comment: “in the third part, the one devoted to the impossibility of communication, there is rather more talk of knowledge than there is of speech (fast mehr von Erkennen als vom Reden [die Rede ist])” (p. 170). 22 “Happens to know” renders the subjunctive JLJQǀVNƝL 23 At b13-14 I read HQ WǀL DXWǀL, , with Newiger 1973, p. 156 n. 22. The restoration in Diels (1900), HQ WǀL DXWǀL HLƝ [masculine plural subject], , is grammatically unacceptable, as Calogero (1977) has pointed out, p. 256 n. 60.

8. Gorgias on the Function of Language

191

And manifestly not even the same subject in relation to himself perceives at the same time things that are alike (phainetai . . . homoia aisthanomenos), 24 but other by hearing and other by sight; moreover, he perceives them differently now and previously. So, one would hardly perceive the same thing as another. (980b14-17)

Note the structure of the passage. Gorgias forswears use of the thesis of incommunicability as premise; yet the thesis is virtually reinstated at the end of each of the three component sections. Thus, the form of the overall argument is: “even if p, then not-p” (where p stands for the proposition that knowledge can be communicated). A clear contrast between the first and the second section is marked by the play of phainesthai, “to appear,” against HLƝ “would be,” at b11-12. The first section raises a purely metaphysical issue – whereas in the second epistemological considerations seem to come in. To understand the issues in each of these two sections we need to resolve a pattern of ambiguity that runs through the key terms “more than one,” “two,” “one thing,” as well as the two terms I shall first cite without translation, homoion and tauto. In ordinary Greek, the latter two terms can both mean either “numerically the same” or “qualitatively the same.” Of course, “same” has exactly that ambiguity in ordinary English. At bl4, as we shall see, Gorgias may well be exploiting both meanings of homoia. But in the first and second sections this ambiguity interacts with corresponding contextual ambiguities of the three terms cited in translation to produce two systematically different readings of Gorgias’ argument. Let me dwell on these two possible readings. First, what might be called, a “realist” reading: If there is a “thing,” a pragma, external to the two subjects, the issue raised in the first section is: How can this one numerically identical pragma nevertheless be “in” entities that are distinct and “more than one'” (pleiosi)? This version of the metaphysical puzzle would be a cognate of the sophistic one-in-many puzzle commonly known as “The Sail” (or in more apt translation of istion, “The Awning”) from Plato’s Parmenides 131B. 25 The epistemological puzzle in the second section would be this: Even if we should allow that the same external thing should somehow also be “in” two knowing subjects, it need not appear the same to them, because the two subjects are differently constituted and differently placed. This realist reading has a serious drawback. In connection with the metaphysical one-in-many puzzle, the reference to two subjects seems gratuitous. The puzzle would arise even when a single subject contemplates the external thing: How could the thing be outside itself, in a knowing subject? We could, of course, say that Gorgias is reasoning a fortiori: There is a problem for the 24

While the sense “alike” is quite apt for homoios at b12, at b14 the translation “same” works better: see below, this Section and Section III. 25 Cf. Mansfeld 1985, p. 255.

192

Alexander P. D. Mourelatos

existence of the thing in the mind of the speaker, and (a fortiori) for its existence in the mind of the listener. But the wording of the text establishes that specifically “two” (dyo) entities are at issue, not three. To be sure, he refers to them as “more than one,” but the Greek makes it clear that the reference is not to the thing itself plus either one or two knowing subjects; it is strictly to the two subjects, the speaker and the listener. 26 If the problem is one of the existence of a single entity in two knowing subjects, a subjectivist or phenomenological reading seems more apt. The “one” and “same” which cannot become “two” and “different” is not an external third thing; it is simply the perception or experience or thought. We could, in fact, translate b9: “How will the listener have the same perception?” The puzzle gains poignancy in this formulation. We are not dealing with a special case of the istion-argument; we are dealing with a fundamental problem in the philosophy of mind, viz., How can two minds have the same perception? The issue is not, as yet, one of skepticism concerning knowledge of other minds. It is, rather, the more rudimentary metaphysical issue of ownership of perception: My perceptions are mine; your perceptions are yours; how is it that you and I can have the same perception? Unfortunately, the phenomenological formulation appears to run into difficulties in the second section of our text. If the entity at issue is a perception, it is awkwardly pleonastic for Gorgias to have said “nothing prevents that it [the perception] should not appear alike” (b11-12). Indeed, the problem is not just one of pleonasm. The second section opens with another of those Gorgianic concessions: “Yet even if it should be in more than a single (subject), nothing prevents . . .” (b11). In accordance with the phenomenological formulation, that concession specifically allows that the same perception exists in two minds. Can we, then, suppose that the concession is withdrawn in the very next sentence, which muses that the two subjects may not have similar (homoion) perceptions? Here it will not do to point to the form of Gorgias’ overall argument – “even if p then not-p.” For in the overall argument “not-p” emerges eventually, as the unexpected upshot of considerations that did not at the start, or prima facie, import “not-p”; whereas in b11-12 “not-p” would follow immediately upon “p.” The problem of incipient contradiction can, nonetheless, be solved if we rehearse a certain natural dialectic of arguments involving the issue of intersubjective perception. Consider the following dialogue: How can two separate subjects have the same perception? – Well, the perceptions are not numerically the same; they are virtually the same in the

26

That pleiosi at b10 is masculine is made clear when its reiteration at b11 is picked up at b12-13 by autois . . . ekeinois, which unmistakably refer to speaker and listener.

8. Gorgias on the Function of Language

193

sense of being similar. 27 – But how similar? – I should say exactly similar. – But how likely is it that two subjects who are differently constituted and differently placed should have exactly similar perceptions?

My suggestion is that the concessive statement, “even if it should be in more than a single (subject) and be the same,” represents the familiar strategy of substituting “similar” for “numerically the same.” And since similarity admits of degrees, the questioner in the above dialogue – or, correspondingly, Gorgias – can easily force his interlocutor to admit that the perceptions may not be exactly similar, after all. So “p” is not immediately contradicted by “not-p”; rather it is gradually transformed into “not-p” through philosophical analysis. Here, then, is how the two sections are related in accordance with the phenomenological reading. The metaphysical problem of ownership of perception raised in the first section is solved at the beginning of the second through the metaphysical device of parceling out similar perceptions to the separate subjects. This, however, leads to epistemological doubts concerning the degree of similarity. The difficulty of pleonasm may now be regarded a minor wrinkle. The residual difficulty of the realist reading – failure to do justice to Gorgias’ repeated and exclusive reference to the two subjects – is, by comparison, more serious. Assuming that the phenomenological reading is the correct one, we must next briefly explore certain ambiguities that are internal to it. The term “perception” is obviously global in scope. To borrow Hume’s famous classification, 28 “perception” covers both sensory impressions and ideas (the latter being mental images). Which is the relevant meaning in Gorgias? Here are the possibilities: (a) Only sensory impressions are at issue. If so, the situation envisaged is one in which speaker and listener are simultaneously observing a certain object, e.g., a horse. The speaker says, “The horse is brown,” and the listener nods assent. And here is the question raised by Gorgias: Is what the listener sees sufficiently similar to what the speaker sees to warrant the judgment that the listener has understood the speaker’s statement? (b) Only mental images are at issue. The speaker, picturing a horse he saw in the past, says, “That horse was brown.” The listener acknowledges the remark. Gorgias’ question: Are the respective mental images sufficiently similar to warrant, . . . [etc.]. 27

28

Compare the question at the start of this exchange with the remark by Hylas in the third of Berkeley’s Three Dialogues (Adams 1979, p. 80): “But the same idea which is in my mind, cannot be in yours, or in any other mind.” And for a parallel to the answer in my exchange, compare Philonous’ answer: “. . . men are used to apply the same where no distinction or variety is perceived.” An Inquiry Concerning Human Understanding, sec. II; A Treatise of Human Nature, book I, part I, sec. I.

194

Alexander P. D. Mourelatos

(c) Both impressions and mental images are involved. The speaker is looking at a horse that is hidden from the listener’s view. The speaker says, “The horse is brown,” and·the listener acknowledges the remark. Gorgias’ question: Is what the listener pictures sufficiently similar to what the speaker sees to warrant, . . . [etc.].

The phrasing in the MXG text does not seem to select one of these to the exclusion of the others. At the start of the first section, the verb HQQRƝVHL (b9) would seem to favor situation (b). But in the sequel, phainesthai (b11), aisthanomenos (b14), and aisthoito (b17) appear to tip the case in favor of (a). Of course, we must not forget that aisthanesthai has itself a broad use, comparable to that of “perceive.” The crucial consideration, rather, is the explication of [ouch] homoia aisthanomenos, “[not] perceiving things that are alike,” by the phrase “other by hearing and other by sight” (b15). So, when we again read ennoei in the recapitulation (b19), we may by then have been persuaded to gloss both this occurrence of the verb and, retrospectively, its earlier occurrence (HQQRƝVHL, b9) as generic references to sensory impressions. And yet, just as b15 provides unequivocal support for (a), the remark made at b16, concerning intra-subjective perceptual sameness over different times, decisively, if indirectly, supports (c). The question here is whether what I see now is same with, or similar to, what I remember seeing yesterday. The governing verb is still aisthanesthai. But the comparison is, clearly, between one DLVWKƝVLV that constitutes an impression and another DLVWKƝVLs that constitutes a memory or mental image – which is the predicament of the listener in (c). When all the evidence is considered, the reasonable conclusion is that all three of the situations distinguished above are implicit in Gorgias’ argument. For if (c) is involved, it seems pointless to exclude (b). The verbs ennoein and aisthanesthai shift meaning in our reading of the text as one or the other of the situations comes to the fore. The global English verb “perceive” is, after all, the proper one to capture this effect of semantic indifference. That noein and aisthanesthai should be allowed to shift back and forth between “mentally picture” and “have a sensory impression of” reveals an assumption which is common to Gorgias and (as he sets things up) to those he addresses in his elenchus. It is the assumption familiar to us from classical empiricism, that ideas or thoughts have their origin and basis in sensory impressions. In other words, thoughts are nothing but sensory impressions preserved in memory and projected in imagination. In terms of situations (a)-(c), the assumption means that unless communication is problem-free in situation (a), there is little hope that it could be problem-free in situations (c) or (b). And it certainly is Gorgias’ argument that communication fails in situation (a). This confirms an observation made earlier, that Gorgias does not endorse the empiricist theory of meaning. And it confirms our judgment that we must not interpret the DOO¶Ɲ clause of b5 as implying just such an endorsement.

8. Gorgias on the Function of Language

195

III. MXG ON THE PROBLEMS OF SAMENESS IN PERCEPTION: THE CASE OF A SINGLE SUBJECT I turn now to the last of the three final sections, b14-17. Just as the second section began by reconsidering the consequence deemed absurd in the first, the third section rhetorically – if not logically – grows out of the second by adapting HQWǀLDXWǀL, , “in the same, one,” into KDXWRVKDXWǀL, “the same, in relation to himself.” It is as though the problem Gorgias raises is now to be addressed to the single subject that has emerged after, in the absurd consequence of the second section, the two subjects are thought as having collapsed into one. The thrust of the argument is clear. Puzzles of perceptual identity and similarity arise even for the solitary individual. All the less credible, then, that different subjects should have the same perception. It is also clear that the puzzles involve two cases of variation: same time but different faculties; same faculty but different times. I shall call these the “synchronic” and “diachronic” cases, respectively. Not clear at all, however, is the basis and scope of Gorgias’ argument. We have, again, the problem of ambiguity of homoia and tauto. While it is reasonable to presume that the phenomenological reading, and with it the sense “exactly similar,” should be continued into the third section, I rather think that both the latter sense and that of “numerically one” are involved, since the argument admits of reconstruction on either semantic basis. Let us dwell first on the synchronic case. The opening clause, taken by itself, seems to be making an extraordinary claim: “not even the same subject in relation to himself perceives the same (or “exactly similar”) things (homoia) at the same time.” This sounds like the pronouncement of an extreme, nightmarish, phenomenological atomism – present consciousness pulverized into bits that are absolutely unique. But the clause that follows corrects that impression. The only differences at issue are those between faculties. Even so, does this not allow another mischievous suggestion to obtrude, viz., that everything that is synchronically perceived by a single faculty constitutes a single thing? Gorgias could not have intended this doctrine of faculty monism as a basis for his argument. Surely he would have allowed that one whose field of vision is filled by the view of, say, a surface of colored square tiles sees: (a) many distinct colors and many distinct squares; (b) the same shape repeated throughout the surface; (c) the same color appearing in different squares. Indeed, there is no need to speculate whether Gorgias “would” have allowed this. Logically, in order for him to be able to raise puzzles as to whether two subjects “perceive the same” he must be able to explain what it is for one subject to perceive things that are different and things that are the same. A complex of synchronic visual perceptions (as in the case of staring at a wall of colored square tiles) would be the obvious paradigm.

196

Alexander P. D. Mourelatos

Now, according to Gorgias, a claim of sameness is illegitimate in cases of intersubjective perception. What would be the corresponding illegitimacy in the case of synchronic intra-subjective perception? The answer is obvious: We say “I see and hear the helicopter,” implying that it is the same entity we see and hear. Yet, strictly speaking, we saw a flying object and we heard a whirring-rumbling noise. A slight variation of this interpretation also appears apt. My sense of vision perceives a waterfall as a curtain-like pattern of glistening white foam; my sense of hearing perceives the waterfall as a deafening roar. There is no third object, a waterfall-in-itself, commonly perceived by my two senses. So too in the case of intersubjective perception, where only one faculty is involved: The waterfall I see is the waterfall-as-seen-by-me; the waterfall you see is the waterfall-as-seen-by-you. There is no view-of-the-waterfall-in-itself commonly perceived by the two of us. In both these interpretations the relevant sense of “same” (same helicopter, same waterfall) is that of numerical identity. But it is also possible to give “same” the sense of exact similarity. The argument would be this: The waterfall perceived through vision is quite unlike the waterfall perceived through hearing. This is understandable given the quite different natures of the eye and the ear. Besides, my constitution is different from yours. Specifically, my faculty of vision is differently constituted from yours. It is likely, then, that just as my visual and aural images of the waterfall are qualitatively dissimilar, my visual and your visual image of the waterfall will be qualitatively dissimilar as well. The two reconstructions that draw on the sense of numerical identity, as well as the one that draws on the sense of similarity, are all germane in explicating Gorgias’ argument. What I do not consider germane, however, are Presocratic theories of the physiology of perception. Again in this connection, as earlier apropos 980b1-2, it has been claimed that Empedocles’ theory of poroi, “(perceptual) channels,” provides the basis for Gorgias’ argument. 29 The methodological objection voiced earlier applies with equal force: Gorgias must exploit considerations that are ready to hand and intuitively compelling. Let us now turn to the diachronic case. The argument is that one does not perceive the same things at different times. I elaborate with an example: a patch that looked purple yesterday looks blue today. To elicit the basis of this denial of perceptual sameness we need to ask three questions: Does it happen always, i.e., may we never speak of two items experienced at different times as constituting or representing the same perception? Is this failure of trans-temporal sameness attributable to a metaphysical principle – either universal flux or, at least, widespread mutability? Or does the failure simply reflect some epistemological shortcomings and limitations on our part? 29

Gigon 1936, p. 212. On the Empedoclean theory of poroi, see above, Chapter IV, Section IX.

8. Gorgias on the Function of Language

197

Strikingly apposite here is a passage from Plato, Theaetetus 154A: 30 Does anything ever appear the same (homoion) to another man as it appears to you? Are you strongly confident of this? Or would you much rather hold that it does not appear the same (tauton) even to you yourself inasmuch as you are never (PƝGHSRWH) in the same condition (KRPRLǀVHFKHLQ) in relation to yourself? – I would much rather accept the latter.

The two interlocutors in the Theaetetus text confidently agree in proclaiming “never the same,” both in the case of different individuals and in the case of different times for the same individual. That proclamation reflects, however, the interlocutors’ sympathetic consideration at that particular stage of the dialogue of the Heraclitean doctrine of flux. Gorgias had better not invoke a flux doctrine, for the same reason that he had better not borrow premises from Empedoclean physiology. 31 Clearly the best strategy for Gorgias would be to stimulate skepticism concerning numerical identity or similarity of temporally separate perceptions. He would allow that there is no difficulty in speaking of numerically the same perception if the perceived object is, say, a protracted steady tone. Similarly, he might allow, for example, that a series of sounds of identical pitch, separated from one another by short intervals, are perceived as exactly similar. But, he may well have asked, How can we be sure that the perceptions are even similar (let alone numerically identical, or exactly similar) when the temporal intervals become long – yesterday and today, last month and this month, last year and this year? 32 One does not have to subscribe to a flux doctrine to recognize that things do change, that human minds as well as human bodies change, and that – in the final analysis, and perhaps most importantly for an elenchus that exploits doubt – memory is notoriously fallible. We are now in a position to understand better the logical connection between the final two sections in the MXG text (b11-14 and b14-17). The problems of intra-subjective sameness of perception discussed in the treatise’s third section are not the cause or source of the problems of intersubjective sameness discussed in the second. The intersubjective and intra-subjective cases are related by analogy. In the intersubjective case we have two different perceiving subjects, and we postulate that the two subjects have the same perception. In the intra-subjective case we have either two different faculties or two different times of perception, and again we postulate that the perceptual entity involved is the same. Yet whether in the case of two subjects, or two

30

Cited by Newiger 1973, p. 157. Cf. Mansfeld 1985, p. 256 and 269 n. 43. It would similarly be poor strategy for Gorgias to rely on a thesis of radical subjectivism, like that of Protagoras: cf. Kerferd 1984, p. 220. In this connection, Calogero’s thesis of the ideological affinity between Gorgias and Protagoras needs to be modified. 32 Note palai, “of old,” at b16. 31

198

Alexander P. D. Mourelatos

different faculties, or two different times, there is sufficient variance in the conditions of perception to make the postulate of sameness highly suspect. I am not arguing that Gorgias is a skeptic generally; merely that a skeptical strategy is sufficient for the elenchus in these two final sections of MXG. This characterization of Gorgias’ strategy is borne out by a certain detail of wording. At b12 the text does not say that the entity at issue cannot appear exactly similar to the two subjects; it says, rather circumspectly, that “nothing prevents (RXGHQNǀlyei) that it should not appear alike,” which is another way of saying that there is no guarantee, no assurance, no certitude, that it will appear alike. Taken as a whole, the argument in the latter half of the third part of Gorgias’ treatise is not skeptical. The metaphysical one-in-two argument was suspended only temporarily, to allow consideration of the suggestion that two subjects can have the same perception, specifically in the sense of having exactly similar perceptions. Once that suggestion has now been set aside on skeptical grounds, the metaphysical argument can be reinstated with full force. Appropriately enough, that argument did not use the politely persuasive locutions of skepticism; it invoked rather the hard modalities of dogmatic metaphysics, ou(ch) hoion, “it is not possible” (b9). And so, when the thesis of incommunicability is summarized, the upshot of the latter half (all three sections) is given in a universally quantified statement: “no one (oudeis) mentally grasps the same as another” (b19). The MXG text fails to draw out an important implication. The diachronic phase of the argument concerning intra-subjective perception is not, as we saw, so extreme as to deny either that the hearing of a continuous steady tone constitutes numerically the same perception or that the hearing of a rapid succession of tones of the same pitch constitutes a series of exactly similar perceptions. But once we get away from the sort of immediacy envisaged in these two examples, the skeptical pangs become crippling. The implication I am about to elicit involves our theme of logos. Even my own words will be impotent to conjure up for me a reality that is not immediately present. Suppose I say to myself: “But when I heard that tone yesterday I said, ‘That’s middle C (or “middle Do”)’ ” or “When I saw that color yesterday I said, ‘Lavender blue.’ ” Of no avail: what I called “X” yesterday and what I am inclined to call “X” today may be quite different entities. In effect – to reverse the point Gorgias makes – I no more share a language with my past self than I do with another subject. Even for the solitary individual in communion with himself, language is either otiose or impotent. We saw earlier a hint, in Sextus’ paraphrase, that Gorgias held that even for the solitary individual logos cannot represent the real thing (things cannot be transformed into logos). 33 The implication just drawn from MXG 980b14-17 bears the hint out. 33

Cf. above, p. 184.

8. Gorgias on the Function of Language

199

Interestingly, Sextus’ remark appeared in the context of Gorgias’ other main argument in the third part of the treatise, the categorial argument. Thus, both the categorial argument and the argument concerning perceptual sameness have a doubly discomfitting upshot: neither a public nor a private language is possible. IV. TARGETS OF THE ELENCHUS: THE REFERENTIAL AND IDEATIONAL CONCEPTIONS OF MEANING Many sophistic puzzles have a double identity. They may have arisen in a purely eristic context; but thinkers who did not dismiss them out of hand often came to realize that profound philosophical issues lay at the logical source of such puzzles. The puzzle concerning the possibility of inquiry, discussed in Plato’s Meno, and the puzzle of false statement, discussed in the Sophist, are famous cases in point. A deep philosophical layer lies also at the source of the puzzles of communication raised in the third part of Gorgias’ Non-Being. 34 The categorial argument is especially well-suited to counter a certain perennially attractive, even though ultimately misguided, assumption concerning the function of language, viz., that if words are to have meaning, they must refer to things in the real (at least extra-linguistic, and perhaps also extra-mental) world. The assumption is that the meaning of all words is constituted by their reference. Accordingly, we could say that a referential conception of meaning is, in effect, the target of the elenchus pressed in the categorial argument. 35 The puzzles of perceptual sameness are correspondingly well-suited to discredit a mentalist or ideational conception of meaning. The target assumption would be that words have some sort of a tie or pairing with perceptions (sensory impressions or mental images or thoughts). In its simplest and purest version, the assumption is that the hearing of a word “W” brings to the mind of each speaker of a certain language the same mental image or thought, and that, conversely, when either that same mental image or thought or the corresponding sense impression should occur to a speaker, the perception can be conveyed to others through use of the word “W.” 36 34

In the paragraphs that follow, I draw on the discussion of alternative “theories of meaning” in Alston 1964, pp. 10-25. 35 Cf. Kerferd 1984, p. 218. 36 Kerferd reaches a markedly different conclusion. According to him, Gorgias actually endorses the ideational conception of meaning. Cf. Kerferd 1984, p. 221: “although words do represent thoughts, it is not the case that they can transfer thoughts . . . [Gorgias is] moving towards a view that meaning is to be understood because words are associated with thoughts or ideas in the mind.” This interpretation postulates an unlikely disparity between the categorial argument and the arguments from perceptual identity. The former, accordingly, would indeed constitute a reduction to absurdity of the referential assumption; but the latter would not constitute a reduction to absurdity of the ideational assumption. At the end of the article, Kerferd makes the further suggestion that Gorgias took “the vital first step, the rejection of a referential theory of meaning, along the path that was eventually to separate meaning from both objects referred and thoughts and to lead to the concept of Sinn in

200

Alexander P. D. Mourelatos

The referential and ideational conceptions might, of course, be combined; and quite often they have been combined. A proponent of the ideational conception may have recourse to the referential assumption in order to explain the nature of the tie or the pairing between words and perceptions. The explanation ready to hand is that words stand for perceptions, i.e., they have perceptions as their referents. Correspondingly, a proponent of the referential conception of meaning may stipulate that the referents of words are not extra-mental things but mental objects – sense data and the like. Moreover, when challenged to explain the nature of reference, the referentialist may have recourse to the theme of perceptual sameness – the theme that is so crucial to the ideational conception. For, in answer to that challenge, it could be said that reference is based on a tacit agreement by all speakers of the language: “We shall all call this by the name ‘W.’ ” But it is obvious that linguistic agreement presupposes some shared perceptions of the thing initially just pointed to by the pronoun “this.” Gorgias’ elenchus is especially devastating vis-à-vis those who would opt to combine the two assumptions in the manner just indicated. The categorial argument and the argument from perceptual sameness would, in this case, be not distinct attacks on alternate targets but well-staged phases in a single attack. For if one sought to evade the force of the categorial argument by countering that words refer to things by virtue of a certain nomos, the conventions of usage, Gorgias could come back with the observation that in order for nomos to take hold, speakers of the language must have convergent experiences. If what I call “W” and what you call “W” are not identical, or at least exactly similar, the nomos we have adopted is bound to be abrogated. It would be relevant here to review some of the standard criticisms urged by modem philosophers against the referential and ideational conceptions of meaning, respectively. This should bring out more clearly the force, scope, and importance of the puzzles raised by Gorgias. A representative modern exponent of the referential conception is the early Russell: “Words all have meaning, in the simple sense that they are symbols which stand for something other than themselves.” 37 An MXG passage makes a very similar remark: ȜȩȖ૳ ਲ਼ ıȘȞİȓ૳ IJȚȞ੿ ਦIJȑȡ૳ IJȠ૨ ʌȡȐȖȝĮIJȠȢ, “by means of a certain word or sign which [in either case] is something other than the thing

37

Frege’s terminology” (p. 221). Kerferd does recognize that Fregean Sinn is something quite other than “thought,” in the psychological sense. Indeed, Sinn is essentially a public entity. But then, once “meaning” is understood as Sinn, skeptical arguments concerning the identity of mental contents in different minds are completely undercut. One might even say that a Gorgianic QRƝPD or ennoia would be the antithesis of Fregean Sinn. Besides, the twin of Sinn in Frege is Bedeutung, “reference”; Frege is certainly no sworn enemy of the referential theory. So, I see no point at all in placing Gorgias at the start of a path that ultimately leads to Frege. This astonishingly sweeping remark comes from Russell 1903, p. 47; italics in the original omitted here. For a discussion of the referential theory of meaning, see Alston 1964, pp. 12-22.

8. Gorgias on the Function of Language

201

[at issue]” (980b4). Here now are four standard objections to the classical Russellian version of the referential conception of meaning: (a) A distinction must be drawn between sense and reference. Meaning cannot be identified with reference, since there are words that have the same referent but different senses, or (as with pronouns and indexical terms) same sense but different referents. (b) The conception commits us to the view that all words are either proper names or referring expressions, which is plain false. (c) The conception must necessarily postulate existing or subsisting things as the referents of expressions of nonentity, unreality, or fictive being. (d) What exactly is this relation of reference between words and things? It is not identity; and it is not resemblance. If, however, it is a sui generis semantic relation, viz., the relation that obtains between proper names and their bearers, objection (b) applies with full force.

We find no hint of objections (a) and (b) in Gorgias – not in MXG, nor in any other fragment or report. This is hardly a disappointment, since these objections presuppose a degree of sophistication in the philosophy of language not achieved before the Stoics. Objection (c), on the other hand, is one that lies within Gorgias’ intellectual horizon. True, it is nowhere implied in the third part of Non-Being. But in the second part of the treatise, Gorgias argues that “if the objects of thought (phronoumena) are existing things (onta),” then, there would actually exist “a flying man or chariots racing at sea” (Sextus Adv. math. VII.79; DK82B3). That the thesis of the second part of the treatise is suspended in the third part does not bar us from considering this piece of evidence. The crucial point here is that the argument of the second part concerning fictions provides not premises but rather a pattern for an argument Gorgias could have developed – though, if the MXG paraphrase is in this respect complete, did not actually develop – viz., that if the meanings of words are existing things, then there must exist things corresponding to, say, “flying man” or “chariot racing at sea.” Argument (d) has something in common with Gorgias’ categorial argument. Certainly, the idea that the relation between words and things is neither one of identity nor one of resemblance is strongly implied in the relevant passages both in MXG and in Sextus. But the specifically categorial strategy of Gorgias’ argument is not adequately brought out through comparison with (d). We have a more suggestive modern analogue in the following argument, which was a commonplace in twentieth-century philosophy of language, and remains so today: If the meaning of a word were the word’s referent, then the meaning of “cake” would be edible, the meaning of “hardness” would be a characteristic of

202

Alexander P. D. Mourelatos

flint – both of which consequences strike us as category mistakes. 38 Gorgias put the matter bluntly: The speaker speaks not a color nor a sound, nor any other thing; he speaks logos (combining lines 980b2-3 and b6). Blunt, even simplistic though the formulation may be, the argument is by no means trivial. As an elenchus of the referential conception, the argument has fully as much force as the refined modern version of it: we do not eat the meaning of “cake.” Let us now explore modern objections to the ideational conception. A representative exponent of the pure version is Stephen Ullmann: “If I hear the name ‘table’, I shall think of a table; if I think of a table, I shall articulate the name if required.” 39 Here are three standard objections: 40 (a´) It is simply false that our use of words (or, for that matter, of any other linguistic unit) is invariably associated with mental imagery. What image accompanies our use of “when,” “in,” or “until”? (b´) Introspection is notoriously unreliable. If we had to depend on mental images in order to use words meaningfully, linguistic expression would be more hesitant and clumsier than in fact it is. (c´) If meanings were mental images, there would always be doubt as to whether a given word has the same meaning when used by different speakers, or even when used by the same speaker at different times.

Nothing in the Gorgias materials lends itself for comparison with arguments (a´) and (b´). It is argument (c´) that immediately commands our attention, since it is virtually a precis of Gorgias’ arguments concerning perceptual sameness in intersubjective and intra-subjective situations. The resemblance is perhaps strong enough to give the reader cause to suspect tendentiousness in my exposition. Let me allay this suspicion by quoting statements of (c´) by twentieth-century philosophers. Here is the intra-subjective case in a statement by William P. Alston: “A little introspection should be sufficient to convince the reader that insofar as his use of the word ‘dog’ is accompanied by mental imagery, it is by no means the case that the mental image is the same on each occasion the word is used in the same way.” 41 And here is the intersubjective case in a statement by A. J. Ayer: “I have no means of knowing that other people have sensations or feelings 38

See Stevenson 1944, p. 42. Cf. Alston 1964, pp. 20-21: “For example, the phrase ‘the father of pragmatism’ refers to C. S. Peirce. If the meaning of that phrase were identical with its reference, we would have to be able to say, both intelligibly and truly, that the meaning of ‘the father of pragmatism’ was married twice and that the meaning often wrote reviews for the Nation. Meanings, however, do not get married and they do not write reviews.” 39 Ullmann 1957, pp. 69-70. The most famous exponent of the ideational conception is, of course, John Locke. His, however, is a combined ideational-referential version: “Words, in their primary or immediate signification, stand for nothing but the ideas in the mind of him that uses them” (An Essay Concerning Human Understanding, book III, ch. 2, sec. 2; cf. ch. 2, passim). 40 I draw not only on Alston 1964, pp. 22-25, but also on Lyons 1977, pp. 96-99, 113-14. 41 Alston 1964, p. 25.

8. Gorgias on the Function of Language

203

which are in any way like my own. I cannot tell even that they mean the same by the words which they use to refer to physical objects, since the perceptions which they take as establishing the existence of these objects may be utterly different from any that I have ever had myself.” 42 The text from Ayer brings to mind the larger context within which this sort of argument came to be deployed by philosophers in the twentieth century. Russell, in his “Logical Atomism” phase, as well as the early positivists of the Vienna Circle, had argued that the meanings of words are, in the final analysis, sense data. Critics were soon to point out that such a view led inevitably to a predicament of linguistic solipsism: only a private language would have the proper semantic grounding; communication would be a hit-and-miss affair drawing on devices of analogy and psychological suggestion. 43 Russell had anticipated this difficulty, but found a way of making virtue out of necessity: When one person uses a word, he does not mean by it the same thing as another person means by it. I have often heard it said that that is a misfortune. That is a mistake . . . [S]ince different people are acquainted with different objects, they would not be able to talk to each other unless they attached quite different meanings to their words . . . It would be altogether incredibly inconvenient to have an unambiguous language, and therefore mercifully we have not got one. 44

It remained for the later Wittgenstein to undercut the ideational conception of meaning that led to linguistic solipsism. The tables were turned on the positivists: it is precisely a private language that is not possible. Here is how Wittgenstein handles the Gorgianic puzzle of sameness of perception in the intersubjective case: We said that there were cases in which we should say that the person sees green what I see red. Now the question suggests itself: if this can be so at all, why should it not always be the case? . . . But then it is clear that the very idea of seeing red loses its use if we can never know if the other does not see something utterly different. 45

And here is how Wittgenstein handles the puzzle in the intra-subjective case: What if someone asked: “How do I know that what I call seeing red isn't an entirely different experience every time? and that I am not deluded into

42

Ayer 1956, p. 206. Ayer here presents not his own view but a view that “was current, at one time, in the Vienna Circle.” 43 See Weinberg 1936, pp. 200-07 and passim. 44 “The Philosophy of Logical Atomism” (1918), in Russell 1956, pp. 195-96. 45 From “Notes for Lectures on ‘Private Experience’ and ‘Sense Data’,” in Jones 1971, p. 272.

204

Alexander P. D. Mourelatos

thinking that it is the same or nearly the same?” Here again the answer “I can't know” and the subsequent removal of the question. 46

V. THE BEHAVIORAL CONCEPTION OF MEANING AND GORGIAS’ HELEN A standard alternative to the referential and ideational conceptions of meaning is the behavioral conception. 47 In John B. Watson’s classical statement, “The words function in the matter of calling out responses exactly as did the objects for which the words serve as substitutes.” 48 Thus the effect of the utterance “Stop!” on the person hearing it would be the same as that of a physical barrier or some form of blocking action. Conversely, words are substitute responses. 49 Instead of jumping for joy, I say “Hurrah!”; instead of writhing and thrashing from pain, I say “It hurts!”. These substitute responses may in turn become substitute stimuli for other persons who may be within hearing. 50 Words, accordingly, are devices of pretense and make-believe in a game humans are inextricably and universally drawn into, from childhood on. There may be no one or nothing physically blocking my path, but the word “Stop!” arrests my movement. My friend may be calm and smiling, but his saying “It hurts!” prompts me to treat him as though he had been writhing and thrashing from pain. The right question to ask about a word is not “What does it stand for?” or “What image does it bring to mind?”, but rather “What stimuli call it forth?” and “What effects does it have on other speakers of the language?” I hasten to point out that the phrase “behavioral conception of meaning” might be understood to cover a wide variety of assumptions, proposals, and theories. Certainly included within its scope would be the “causal theory of meaning,” also the “dispositional theory of meaning” and (for a narrowed domain) the “emotivist theory of meaning.” Moreover, just as the referential conception has sometimes been combined with the ideational conception, there are sophisticated versions of the behavioral conception that accommodate not only the referential function but also such concepts as logical entailment and truth conditions. For purposes of my comparison with Gorgias, however, the relevant version is the one suggested by my quotation from Watson. It is the version associated with the heyday of behaviorism in psychology and linguistics (Watson, A. P. Weiss, Clark Hull, Leonard Bloomfield) and with the flourish of emotivism in philosophy (A. J. Ayer, C. L. Stevenson). I shall call it the “classical” version; for, even though the conception is now considered

46

Jones 1971, p. 236 See Alston 1964, pp. 25-31; Lyons 1977, pp. 120-37. 48 Watson 1930, p. 233. 49 Lyons 1977, pp. 125-27. 50 Loc. cit. 47

8. Gorgias on the Function of Language

205

extremist and simplistic, it had in fact dominated the field of semantics for a good part of the middle of the twentieth century. Something very like the classical version of the behavioral conception of meaning is implied in Gorgias’ celebration of the powers of logos in Helen. 51 Gorgias starts with the famous remark, “logos is a grand potentate [or ‘a powerful executive,’ G\QDVWƝVPHgas] who can, by means of an extremely tenuous and altogether invisible body, accomplish effects that are utterly divine” (B11(8)). The contrast between the diminutive substance of logos and the magnitude of its effects reaffirms the categorial gulf between words and things – an echo of the third part of Non-Being. But the contrast also brings out the character of logos as substitute stimulus. Gorgias develops his thesis by drawing on a variety of contexts. First case in point: the use of words in the theater. What is important about his choice of that particular context is that the events played out on the stage are of no real concern to us: “in response to the happy and unhappy occurrences affecting things and bodies that are not one’s own, the soul comes itself (idion) to experience a certain emotion, through the logoi” (B11(9)). One could hardly ask for a better illustration of the conception of words as substitute stimuli. Gorgias’ second case in point is the use of words in temple medicine: “divinely inspired spells (HSǀLGDL) by means of logoi take away pain and bring on pleasure” (B11(10)).52 The modern analogue would be the use of words in psychotherapy. A critic of Gorgias might have objected at this point that the behavioral conception of meaning can hardly do justice to the use of logos in contexts of scientific inquiry or generally in contexts of argument. Gorgias was ready with a rejoinder. At B11(13) he takes up the three contexts that might be considered least favorable to his view: “first, the logoi of cosmologists (PHWHǀURORJRXV) 53 who, exchanging one opinion for another – removing one and instilling another – make what is hidden (DGƝOD) and hardly credible (ta apista) appear evident (phainesthai) to the eyes of opinion.” Here Gorgias touches on the thorny problem of the semantics of theoretical terms. Since theoretical entities are not accessible to observation, the language we employ in introducing and describing such entities cannot be based directly either on the things of the familiar and manifest world or on our repertoire of experiences. We resort to analogies, metonymy, heuristic devices, make-believe, psychological prompting. So, the context of scientific inquiry, argues Gorgias, far from discrediting the conception of words as substitute stimuli rather serves to uphold that conception. 51

The text is most easily accessible in Diels-Kranz and in Untersteiner 1961, pp. 88-113. But there is also an excellent critical edition with introduction, translation, and notes: Donadi 1982 (in Italian). See also the very useful student edition by MacDowell 1982, in the Bristol Classical Press series. For another English translation (by George Kennedy), see Sprague 1972, pp. 50-54. 52 For the use of incantations in temple medicine, see Lloyd 1979, p. 42. 53 Not “astronomers”; meteorologoi are those who speculate about the nature of things in the heavens. See Aristophanes Nub. 228, 1284; Plato Ap. 18B, 23B. Cf. Lloyd 1979, p. 84 n. 129.

206

Alexander P. D. Mourelatos

As for the use of logoi in argument, Gorgias has ready rejoinders in the second half of B11(13): “second, forensic contests, 54 in which a particular logos excites and persuades a huge crowd – a written logos prepared with oratorical skill, not one spoken in truth. Third, the moots of philosophical logoi, in which merely (kai) the quickness of thought [of one party] manifestly makes confidence [by another party] in an opinion appear easily changeable.” What Gorgias is saying here is that the crucial function of logos in argument – whether in a court of law or in disinterested friendly debate – is not to represent reality and not to establish or defend truth, but simply to bring about a change of opinion. The celebration of logos concludes with the famous comparison of the power of logos with that of drugs: “Just as different drugs draw different humors from the body, and some put an end to disease while others put an end to life, so too with logoi: on those who hear them, some cause pain, others pleasure, some cause fear, others instill courage, and still others poison and bewitch the soul through some sort of ruinous persuasion” (B11(14)). If only we changed the archaic expression of “drawing out humors” to the behaviorist idiom of “eliciting a physiological reaction” this sentence could just as well have been written by such advocates of the stimulus-response conception of meaning as Leonard Bloomfield, or B. F. Skinner, or C. L. Stevenson. VI. WORDS AS RESPONSES TO PHYSICAL STIMULI: THE TESTIMONY OF SEXTUS There is an intriguing possibility that the behavioral conception of meaning was hinted at, in a brief aside, in Non-Being. This occurs in a section of Sextus’ paraphrase that I have not so far discussed: Indeed, discourse (logos), he says, comes to be formed for us (KƝPLQ . . . synistatai) from things that make an impact (SURVSLSWRQWǀQ KƝPLQ) on us from the outside, which is to say, from the objects of sensation (WǀQ DLVWKƝWǀQ). Thus, from the encounter (HQN\UƝVHǀV) with a flavor, there arises within us (HQJLQHWDLKƝPin) the discourse that is elicited by that (NDWDWDXWƝV ekpheromenos) quality, and from the impinging (K\SRSWǀVHǀV) of a color, [the discourse elicited by] color. And if this is so, it is not the case that discourse is something that represents (parastatikos estin) external reality (tou ektos); rather, it is the external reality that comes to be PƝQ\WLNRQof the discourse. (Adv. math. VII.85) 55

54

DQDQNDLRXVDJǀQDV, that is, contests that are under the “constraints” of political or juridical process, in contrast to free philosophical hamillai. See Immisch 1927, p. 32; Sykutris 1928, p. 14 n. 3. 55 The phrase HQJLQHWDLKƝPLQ suggests that the earlier KƝPLQwhich is quite naturally taken as complement of SURVSLSWRQWǀQ must also be taken in an apo koinou construction with synistatai. For discussion of other points of translation, see below.

8. Gorgias on the Function of Language

207

The passage bristles with the technical terms of Hellenistic epistemological debates. The paraphrase has been cast in what Sextus, or his source, considers philosophically respectable vocabulary. But it is perhaps not impossible to recover behind that Hellenistic encrustation a recognizably Gorgianic thought. It is important to notice at the outset that Sextus presents Gorgias as making a positive statement concerning the origin and the function of language. Since so much of the argument of Non-Being is negative and dialectical, this positive tenor of the passage may be reason enough for us to discount the testimony the text offers. But such a ruling would be rash. The argument in the third part of Non-Being already presupposes two major positive theses – Gorgias’ own concessions – viz., that, contrary to the argument of the first part, there is some external reality, and that, contrary to the argument of the second part, external reality can be known. Inevitably, the argument that linguistic communication is impossible does imply some alternative view concerning the function of language, and it is not out of the question that Gorgias should have indulged in some hints of that alternative. Does the Sextus passage contain some such hint or hints? Much hinges on the adjective I have left untranslated: PƝQ\WLNRQ The cognate verb PƝQ\ǀ standardly means “communicate (a message).” This verb is especially apt in contexts in which the communication involves an intermediary, such as a messenger or a transmitted message or a sign. Sextus himself uses the verb three times in this sense when he presents his version of Gorgias’ categorial argument: How can these [perceptible things] be communicated (PƝQ\HVWKDL) to someone else? That by means of which we communicate (PƝQ\RPHQ) is discourse; and discourse is not the underlying realities or the actual things (ta hypokeimena kai onta). It follows that we communicate (PƝQ\RPHQ) to our interlocutors not the actual things but rather discourse, which is quite other than the underlying realities. (Adv. math. VII.83-84; cf. 86)

In accordance with this use, the adjective PƝQ\WLNRV -Ɲ -on conveys the “communicatory” function performed by the messenger, the message, or the vehicle of the message. It does not appear, however, that this sense is at all suitable for the PƝQ\WLNRQof section 85. Applying that sense yields a statement which pushes sophistic paradox to the point of absurdity: “It is not the case that discourse represents external reality; rather it is external reality that is communicatory of discourse.” What could this mean? Do stones speak? Are messages being spoken by the roaring waves, or by the whispering pines? Modern scholars have tried to steer away from absurdity with the help of a clever gloss. They modulate the sense of “communicatory” slightly, so as to give it the nuance of “explanatory” or “explicatory.” Gorgias would thus come across as proposing that it is the external things that “explain,” “interpret,” or

208

Alexander P. D. Mourelatos

“make intelligible” the words we use.56 The mechanism by virtue of which things explain words is, on one elaboration of this reading of Gorgias, that of onomatopoeia; 57 on another elaboration, it is that of reference (in the semantic sense) to items (whether external things or sense data) that are immediately furnished in experience. 58 Now it should be obvious to any reader of the present Chapter that this modulated rendering of “communicatory,” in either of the elaborations advocated by modern scholars, must be judged in conflict with the rest of the argument in the third part of Non-Being. For if words have an onomatopoetic origin, there ought to be more than a glimmer of hope that intersubjective communication is possible after all – at least in those cases in which language has not diverged widely from the original onomatopoeia. This reading, as I have already objected, leads to trivialization of Gorgias’ argument. Nor is it at all plausible that Gorgias should be endorsing and propounding the thesis that words refer to items directly experienced. That is no alternative espoused by Gorgias; rather it is – as we saw in Part IV of the present Chapter – the target most directly vulnerable to an assault that combines the categorial argument with the argument from perceptual sameness. Indeed, both these readings make even Sextus’ text incoherent. Assuming that words either imitate external things through onomatopoeia or merely function to label for us items directly experienced, why should Gorgias have denied the proposition that language has the function of “representing” or “exhibiting” or “setting forth for us” (parastatikos) something that is extra-linguistic (first half of the concluding statement in section 85)? If words do imitate things, then words necessarily give us some representation of reality. Or, if the relation between words and things is one of reference, i.e., if words perform the sort of job commonly performed by a signpost or label or pointing hand, why should Gorgias have denied that words “exhibit” the objects to which they point? Advocates of the modulated translation of PƝQ\WLNRQ “communicatory,” have not been unaware of the problems of incoherence entailed by that translation. They have, however, chosen to see in these problems not evidence that the translation may be faulty but rather evidence of confusion on Sextus’ part, and, for that matter, evidence of the superiority of the paraphrase in MXG over that in Sextus. 59 I am not wholly unsympathetic to this sentiment; for I do think that Sextus must have missed the radical tenor of Gorgias’ argument at least 56

Here is the rendering in Calogero 1977 of the concluding remark of section 85: “. . . it is not discourse that exhibits (esibisce) the external reality, rather it is the external reality that makes intelligible (fa capire) the discourse” (pp. 256-57). And here is the rendering by Untersteiner (1961): “. . . it is not words that represent outer experience, rather it is the outer experience that gives meaning to words (dà un senso alla parola)” (pp. 55, 57). Similar translations are adopted by Loenen 1959, pp. 197-98; Bury 1935, ad loc.; Newiger 1973, p. 163; and Kerferd 1981a, p. 98. 57 So Kerferd 1981a, p. 98. 58 So Loenen 1959, pp. 198-99; Newiger 1973, pp. 164, 167-68. 59 See Calogero 1977, p. 260; cf. Newiger 1973, pp. 161-70.

8. Gorgias on the Function of Language

209

once – when, in section 86, he offers the jejune gloss: “many of the external realities logos fails to show.” But in the case of section 85 we do have, as I shall presently show, the possibility of an interpretation that is both charitable and philosophically engaging. VII. RELEVANCE OF THE STOIC CONCEPT OF “COMMEMORATIVE SIGNS” It has been overlooked that the home context of PƝQ\WLNRQ is in Sextus’ critique of the Stoic theory of signs. 60 The Stoics had distinguished between “indicative” (endeiktikon) and “commemorative” (K\SRPQƝVWLNRQ) signs. The former have the power to disclose the character or presence of something that is “naturally non-evident” (SK\VHL DGƝlon). At Adv. math. VIII.154, Sextus, who is our source for this doctrine of the Stoics, attributes to them the quite extraordinary claim that the indicative sign “manifestly, out of its proper nature and constitution, reveals (VƝPDLQHL) that of which it is indicative, almost as though it were voicing (it) in speech (ȝȩȞȠȞ Ƞ੝Ȥ੿ ijȦȞ੽Ȟ ਕijȓİȞ).” An example would be bodily motions as an indicative sign of the existence and operation of soul (VIII.155). Commemorative signs, by contrast, do not have this intrinsically revelatory power. No disclosure of something that lies beyond observation is involved. Rather, when two quite distinct events have been repeatedly observed to occur either in a set pattern of succession or simultaneously, then the observed occurrence of the one event becomes a commemorative sign either of the simultaneous (but temporarily non-obvious), or prior, or subsequent occurrence of the other event. Examples from VIII.152-153: smoke as a sign of fire (simultaneous occurrence); the scar as a sign of a prior wound; puncture of the heart as the sign of future death. Sextus has a long brief against indicative signs. But he has no quarrel with commemorative signs, which he finds quite in the spirit of Pyrrhonian skepticism. (In the Stoics’ “commemorative signs,” we, of course, recognize the ancestor notion to the “natural signs” of early modem philosophy; and note in particular the Stoic-Skeptical origin of Berkeley’s doctrine that the nexus involved in “laws of nature” is that of “signification.”) This review of Sextus’ critique of the Stoic theory of signs puts the final remark in VII.85 in quite a different light. Unlike Gorgias, the Stoics did not fundamentally question the ability of language to reveal the nature of things; so much so that they did not hesitate to use the metaphor of “things voicing speech” (cf. ȝȩȞȠȞ Ƞ੝Ȥ੿ ijȦȞ੽Ȟ ਕijȓİȞ) when they sought to characterize the function of indicative signs. In interpreting Gorgias’ position, Sextus comes 60

The main text is VIII.141-299, esp. 145-158. For a helpful survey of all the relevant texts, see Stough 1969, pp. 125-37. – Here is the break-up of the thirteen occurrences of PƝQ\WLNRV + genitive UHFRUGHG LQ WKH LQGH[ RI -DQiþHN  RQH LV RXU 9,, SDVVDJH ;, LV PLVFODVVLILHG LW LQvolves a different construction, with the dative); and, out of the remaining eleven, nine involve the PƝQ\WLNRVfunction of the sort of natural signs that are taken up in the Stoic theory.

210

Alexander P. D. Mourelatos

very close to saying – and perhaps even Sextus would have balked at the blatant anachronism of this formulation of Gorgias’ position – that discourse does not function as an indicative sign of reality; it functions at best as a commemorative sign. To say that the relation between words and things is that of commemorative signification is to say that there is a firmly established regularity, one that connects linguistic events with events in the natural environment. The original sequence could not possibly be words-things; human behavior is much too labile and unpredictable for such laws of linguisticophysical correlation to be established; it would have to be things-words. The correct way to interpret PƝQ\WLNRQ, then, is as alluding to the regularity with which certain non-linguistic stimuli are followed by certain linguistic responses. 61 Here, then, is my completed translation of the concluding remark of VII.85: . . . it is not true that discourse is something that represents some external reality; rather it is the external reality that comes to foretoken (PƝQ\WLNRQ ginetai tou) the discourse. 62

Examples of the foretokening envisaged are given by Sextus in the sentences that precede his concluding remark: different flavors elicit different verbal responses; and so too with different colors. The doctrine of the commemorative sign can then fill in the rest of a stimulus-response account of the function of language, along more or less the lines that were retraced by the behaviorists of the twentieth cantury. The story is as familiar as it is simplistic: 63 Having heard ourselves and others repeatedly make the utterance “sweet!” after tasting honey, we come to expect the presence of honey upon merely hearing the utterance “sweet!” We also learn that we can instill in others the expectation that there is some honey around just by uttering the word “sweet!” And, once we realize that we can manipulate others by uttering “sweet!” even when there is no honey to be had, we are already beyond the invention of language; we have invented Gorgias’ own special art, the art of rhetoric. The analysis I am proposing here is in a position to give an especially satisfactory explanation of the prominence given to flavors in our Sextus pas61

Especially noteworthy as a parallel for the concluding remark of VII.85 is a passage from Book V of Adv. math. Sextus adapts his critique of the Stoic theory of signs to a challenge of the claims of Chaldean astrology. Citing the stock example of puncture of the heart leading to death, he questions whether astrologers can make this sort of claim: “This configuration of the stars has been established as PƝQ\WLNRVof this sort of life” (V.104). 62 Gigon’s translation has anticipated mine: “Accordingly, logos is the result of the causal action from outside (das Resultat der Einwirkung von aussen) and not a sign (Zeichen) for what lies outside” (1936, p. 94). But Gigon failed to exploit the conceptual possibilities of a purely causal theory of meaning implicit in this rendering. Instead, he interpreted the text as an adaptation of the Empedoclean theory of efflux in perception. 63 And it is not that much more simplistic in its Gorgias-Sextus version than it is in its twentiethcentury revival. See Alston 1964, pp. 25-31; Lyons 1977, pp. 133-37.

8. Gorgias on the Function of Language

211

sage. The more immediate and primitive response to a flavor is characteristically an interjection, something like “Yum!” or “Yuk!” Many of the Greek words for flavors – and this holds correspondingly of English words – have a strong emotive component. It is significant in this respect that the adapting of chymos, “juice,” as the generic term for “flavor,” is not attested before Democritus; for generic reference, the early philosophers and physicians of Greece appear to have made do with the opposites KƝGRQƝ “pleasure,” and DƝGLD, “unpleasantness, disgust.” 64 Evidently, the case of flavors is mentioned first by Sextus because it is flavors that provide the relevant paradigm: the nature of the stimulus is such as to trigger an immediate linguistic response, one that is expressive of delight or disgust. Colors are not as immediately compelling; the response in this case may be delayed or suppressed. The case of colors is being assimilated to that of flavors; but it can in turn serve as the bridge to the grand generalization that all language is a play of stimulus and response. Here one might object that if Gorgias were propounding (or if Sextus, or his source, were attributing to Gorgias) a stimulus-response conception of meaning, the choice of flavors and colors as cases in point is incongruous. The point would have been made more effectively by the citing of the tendency of stimuli either of pain or of pleasure to call forth, respectively, utterances of anguish or of delight. But in a compendious remark the best strategy is to omit the obvious cases. For it is obvious enough that the stimulus-response account works rather well with interjections. So, it is best to go directly to the ones that can serve as a bridge to the less obvious ones. Relevant to recall in this connection is Gorgias’ virtuoso deployment of Slippery Slope argument in Helen: “Follow me as I shift from one argument to another” (B11(9)). 65 The shift, as we saw, was from the use of logos, first, in the theater, also in magical incantation, and in psychotherapy (the three cases that lend themselves most directly to Gorgias’ thesis); and then to the harder cases of the use of logos, scil., in scientific inquiry, in the courts, and in disinterested philosophical debate. It would seem that the same argumentative strategy motivates the shift from flavors to colors in Adv. math. VII.85. It is rather uncanny how closely the vocabulary of section 85 resembles the vocabulary of modem behaviorist theory. External objects are not said to be noticed, attended to, or experienced; they “fall upon us” or “make an impact on us” or “impinge upon us” (SURVSLSWRQWǀQ K\SRSWǀVHǀV); the event is not described as one of perception or awareness; it is a “running into” or “encounter” (HQN\UƝVHǀV). The discourse spoken is not a contemplative acknowledgment of the quality present; it is “drawn-out-from us and projected-toward it [the quality]” (NDWDWDXWƝVHNSKHURPHQRV). If, using Sextus as our basis, we had to 64 65

See LSJ and index to Diels-Kranz, s. vv. ijȑȡİ į੻ ʌȡઁȢ ਙȜȜȠȞ ਕʌૃ ਙȜȜȠȣ ȝİIJĮıIJ૵ ȜȩȖȠȞ. This has often been misunderstood as a mere filler: “I shall now shift to another topic.”

212

Alexander P. D. Mourelatos

devise Greek equivalents for the two key terms of modern behaviorist theory, the logical formation for “stimulus” would surely be SURVSWǀVLVor K\SRSWǀVis (cf. German Einfall), and for “response,” ekphora (“something drawn out,” an “ex-pression”). 66 The analysis and translation offered here avoids the difficulties entailed by translations that have things “explaining” words. The categorial argument is not compromised. Gorgias’ doctrine, according to Sextus, is that there is absolutely no affinity between words and things. Onomatopoeia, where it may arise, is at best a superfluous feature and standardly a misleading one. A wordthing connection will come about only if there is a physical routine that pairs one particular stimulus with a particular response. If there should also be some onomatopoetic resemblance, it is irrelevant – a mere epiphenomenon. The tie is established not by the onomatopoeia but simply and wholly by the routine of paired occurrence. Besides, onomatopoeia is more misleading than helpful. For it intimates that there is, after all, some kind of affinity between words and things. Nor does my analysis envisage any softening of the elenchus of the referential conception of meaning. According to that conception, the referent provides some sort of basis, ground, or legitimation of the word that does the referring. The stimulus-response conception makes no such assumption. The paraphrase of Gorgias given in Adv. math. VII.85 may, of course, be a mirage. Perhaps the remarks attributed to Gorgias in that passage are only reflections of attempts by Sextus or by his Hellenistic predecessors to read the philosophical subconscious of the famous Sophist. Perhaps Non-Being offered no positive hint of a conception of meaning that could serve as an alternative to the referential and ideational conceptions that are demolished in that treatise. In that event, Adv. math. VII.85 loses the status of testimony; but it gains stature as a piece of incisive philosophical historiography. For we have seen good evidence that a stimulus-response conception is hinted at – not in a brief aside but quite broadly – in Helen. Sextus or his sources should at least get the credit for whatever measure of acceptance readers of the present Chapter may choose to accord to its thesis of reconciliation of the treatise on Non-Being with Gorgias’ career and output as orator, teacher, and diplomat. 67 66

67

In Modern Greek, the English term “stimulus” is often translated, erethisma, which is standardly “irritation” or “rash”; “response,” often rendered antidrasis, “reaction,” can bear political coloring (cf. antidhrastikos, “reactionary”). The Sextus terminology is obviously superior. Early drafts of this Chapter have received the benefit of criticism at presentations as follows: the 1982 Symposium on the Sophistic Movement held in Athens; and the 1983 Gorgias conference held at Lentini and Catania (for details for both, see unnumbered note at the start of the present Chapter). Subsequent intermediate versions were presented at: the University of Bern; the summer 1986 N.E.H. seminar on “Sophists and Skeptics” (Paul Woodruff, director), held at my home university; and at Stanford University in fall 1986. For critical comments and for other direct assistance, I acknowledge debts to: Geoffrey Lloyd, Sarah Waterlow Broadie, Rick McKim, Rick Benitez, and Larry Schrenk. – The bulk of the research for this study was done in 1982, while I held a Fellowship for Independent Study and Research awarded by the National Endowment for the Humanities. – Inevitably, the bibliography reflects, by and large, the status quaestionum before the mid-1980s

Bibliography

Adam, J. 1963. The Republic of Plato. Edited with Critical Notes, Commentary and Appendices. 2nd ed., 2 vols. Cambridge: Cambridge University Press. Adams, R. M. (ed.) 1979. George Berkeley, Three Dialogues between Hylas and Philonous [orig. publ. 1713]. Indianapolis: Hackett Publishing. Allen, R. E. 1983. Plato’s Parmenides: Translation and Analysis. Minneapolis: University of Minnesota Press. Allen, R. E. and Furley, D. J. 1970-1975. Studies in Presocratic Philosophy. 2 vols. London: Routledge & Kegan Paul. Alston, W. P. 1964. Philosophy of Language. Englewood Cliffs, NJ: Prentice-Hall. Annas, J. 1976. Aristotle’s Metaphysics: Books Mu and Nu. Oxford: Clarendon Press. Armstrong, D. M. 1989. Universals: an Opinionated Introduction. Boulder/San Francisco/London: Westview Press. Audi, R. (ed.) 1995. The Cambridge Dictionary of Philosophy. Cambridge: Cambridge University Press. Austin, S. 1979. “The Method of Parmenides,” Ph.D. dissertation, University of Texas at Austin. Ayer, A. J. 1956. The Problem of Knowledge. Harmondsworth: Penguin. Bächli, A. and Graeser, A. 2000. Grundbegriffe der antiken Philosophie: Ein Lexikon. Stuttgart: Philipp Reclam. Bächli, A. and Petrus, K. (eds.) 2003. Monism. [Andreas Graeser Festschrift] Philosophische Analyse 9. Frankfurt: Ontos Verlag. Barnes, J. 1979a. The Presocratic Philosophers. 2 vols. London: Routledge & Kegan Paul. Barnes, J. 1979b. “Parmenides and the Eleatic One,” Archiv für Geschichte der Philosophie, 61, pp. 1-21.

214

Alexander P. D. Mourelatos

Barnes, J. 1981. “Reply to Professor Mourelatos,” Philosophical Books, 22, pp. 7780. Beare, J. I. 1906. Greek Theories of Elementary Cognition from Alcmaeon to Aristotle. Oxford: Clarendon Press. Berryman, S. 2003. “Democritus and the Explanatory Power of the Void.” In V. Caston and D. W. Graham (eds.), Presocratic Philosophy: Essays in Honour of Alexander Mourelatos. Aldershot: Ashgate Publishing Company, pp. 183-91. Bollack, J 1969. Empédocle. Vol. II: Les Origines, édition et traduction des fragments et des témoignages. Vol. III: Les Origines, commentaire. Paris: Éditions de Minuit. Bollack, J. 1965. Empédocle. Vol. I.: Introduction à l’ancienne physique. Paris: Éditions de Minuit. Bonitz, H. 1870. Index Aristotelicus. Berlin: Georg Reimer. Brentlinger, J. 1972. “Incomplete Predicates and the Two-World Theory of the Phaedo,” Phronesis, 17, pp. 61-79. %ULQN % WHQ  ³'HPRFULWL OLEHU ʌİȡ੿ ਕȞșȡઆʌȠȣ ij઄ıȚȠȢ´ Philologus, 8, pp. 414-24. Broadie, S. 2012. Nature and Divinity in Plato’s Timaeus. Cambridge: Cambridge University Press. Bruno, V. J. 1977. Form and Color in Greek Painting. New York: W.W. Norton & Company. Burkert, W. 1962. Weisheit und Wissenschaft: Studien zu Pythagoras, Philolaos und Platon. Nürnberg: Verlag Hans Carl. Burkert, W. 1972 (translation of the above). Lore and Science in Ancient Pythagoreanism. Transl. E. L. Minar Jr. Cambridge, MA: Harvard University Press. Burnet, J. 1957. Early Greek Philosophy. 4th ed. New York: Meridian Books. Bury, R. G. (transl.) 1935. Sextus Empiricus with an English Translation, 4 vols., Loeb Classical Library. Cambridge, MA: Harvard University Press. [the paraphrase of Gorgias’ 3HULWRXPƝRQWRV is in Adv. math. VII.65-87, which in Bury’s edition is found in vol. 2, “Against the Logicians,” I. 65-87]. Calogero, G. 1977. Studi sull’Eleatismo. Rev. ed. Florence: La Nuova Italia. First edition, Rome: Tipografia del Senato, 1932. Cartledge, P. 1997. Democritus. New York: Routledge.

Bibliography

215

Casertano, G. 1978. Parmenide: il metodo, la scienza, l’esperienza. Naples: Guida. Cassin, B. 1980. Si Parménide: le traité anonyme De Melisso Xenophane Gorgia. Édition critique et commentaire. Cahiers de philologie, 4. Lille: Presses Universitaires. Churchland, P. M. 1984. Matter and Consciousness. Cambridge, MA: MIT Press. Cook Wilson, J. 1892-1893. “Apelt’s Pseudo-Aristotelian Treatises” [= critical notice of Otto Apelt’s Aristotelis quae feruntur De Plantis, De Mirabilibus Auscultationibus, Ventorum Situs et Nomina, De Melisso Xenophane Gorgia], The Classical Review, 6 (1892), pp. 16-19, 100-07, 156-62, 20914, 441-46; and 7 (1893), pp. 33-39. Curd, P. 2004. The Legacy of Parmenides: Eleatic Monism and Later Presocratic Thought. Las Vegas: Parmenides Publishing. First edition, Princeton: Princeton University Press, 1998. Curd, P. 2007. Anaxagoras of Clazomenae: Fragments. Text and Translation with Notes and Essays. Toronto: University of Toronto Press. Curd, P. and Graham, D. W. (eds.) 2008. The Oxford Handbook of Presocratic Philosophy. Oxford and New York: Oxford University Press. De Ley, H. 1980. “Pangenesis versus panspermia: Democritean notes on Aristotle’s Generation of Animals”, Hermes, 108, pp. 129-53. ǻȘȝȘIJȡȐțȠȢǻHG ȂȑȖĮȜİȟȚțઁȞIJોȢਦȜȜȘȞȚțોȢȖȜȫııȘȢYROV$WKHQV Dimitrakos. Diels, H. 1900. “Aristotelis qui fertur de Melisso Xenophane Gorgia libellus.” Philosophische und historische Abhandlungen der königlichen Akademie der Wissenschaften zu Berlin. Berlin, pp. 1-40. Diels, H. and Kranz, W. 1952. Die Fragmente der Vorsokratiker. 6th ed. revised by W. Kranz, 3 vols. Berlin: Weidmann. Referred to by the standard abbreviation “DK.” Donadi, F. 1982. Gorgia: Encomio di Elena. – Bollettino dell’Istituto di filologia greca, suppl. 7, Università di Padova. Fränkel, H. 1938. “A Thought Pattern in Heraclitus,” American Journal of Philology, 59, pp. 309-337. Repr. in Mourelatos 1974, pp. 214-28. Fränkel, H. 1962. Dichtung und Philosophie des frühen Griechentums: Eine Geschichte der griechischen Epik, Lyrik und Prosa bis zur Mitte des fünften Jahrhunderts. Munich: Beck.

216

Alexander P. D. Mourelatos

Fritz, K. von 1966. Philosophie und sprachlicher Ausdruck bei Demokrit, Plato und Aristoteles. Darmstadt: Wissenschaftliche Buchgesellschaft. First edition, New York: G. E. Stechert & Company, 1938. Furley, D. J. 1967. Two Studies in the Greek Atomists. Princeton: Princeton University Press. Furley, D. J. 1973. “Notes on Parmenides.” In Lee–Mourelatos–Rorty 1973, pp. 115. Furley, D. J. 1976. “Anaxagoras in Response to Parmenides.” In R. A. Shiner and J. King-Farlow (eds.), New Essays on Plato and the Pre-Socratics. – Canadian Journal of Philosophy, suppl. vol. 2, pp. 61-85. Furley, D. J. 1987. The Greek Cosmologists: the Formation of the Atomic Theory and Its Earliest Critics. Cambridge: Cambridge University Press. Gallop, D. A. 1981. “Ex Nihilo Nihil, In Nihilum Nil: A Replay to Mourelatos,” Journal of Philosophy, 78, pp. 666-67. Gigon, O. 1935. Untersuchungen zu Heraklit. Leipzig: Dieterich. Gigon, O. 1936, “Gorgias ‘Über das Nichtsein’,” Hermes, 71, pp. 186-213. Repr. in his Studien zur antiken Philosophie. Berlin: Walter de Gruyter, 1972, pp. 69-97. Gill, M. L. and Pellegrin, P. (eds.) 2006. A Companion to Ancient Philosophy. Blackwell Companions to Philosophy. Oxford: Blackwell Publishing. Gomperz, T. 1901. The Greek Thinkers. 4 vols. London: John Murray. Goudge, T. A. 1967. “Emergent Evolutionism.” In P. Edwards (ed.), The Encyclopedia of Philosophy, 8 vols. New York: Macmillan & Free Press, vol. 2, pp. 474-77. Goudge, T. A. 1973. “Evolutionism.” In P. P. Wiener (ed.), Dictionary of the History of Ideas, 4 vols. New York: Charles Scribner’s Sons, vol. 2, pp. 174-89. Graeser, A. 1978. “Vier Bücher zur Eleatik,” Göttingische Gelehrte Anzeigen, 230, pp. 37-69. Guthrie, W. K. C. 1950. The Greek Philosophers, from Thales to Aristotle. London: Methuen & Company. Guthrie, W. K. C. 1962-1982. A History of Greek Philosophy. 6 vols. Cambridge: Cambridge University Press. Guthrie, W. K. C. 1975. Review of Newiger 1973, Gnomon, 45, pp. 705-08.

Bibliography

217

Hamilton, H. A. 1899. The Negative Compounds in Greek. Baltimore: John Murphy & Co. Hankinson, R. J. 1991. Galen, On the Therapeutic Method, Books I and II: Translated with an introduction and commentary. Oxford: Clarendon Press. Hankinson, R. J. 1998. Cause and Explanation in Ancient Greek Thought. Oxford: Clarendon Press. Hasper, P. S. 1999. “The Foundations of Presocratic Atomism,” Oxford Studies in Ancient Philosophy, 17, pp. 1-14. Hayduck, M. (ed.) 1904. Michaelis Ephesii in libros De partibus animalium, De animalium motione, De animalium incessu commentaria. Commentaria in Aristotelem graeca, 22. Berlin: Georg Reimer. Heidel, W. A. 1906. “Qualitative Change in pre-Socratic Philosophy,” Archiv für Geschichte der Philosophie, 19, pp. 333-79. Repr. with the original pagination in W. A. Heidel, Selected Papers, ed. L. Tarán. New York/London: Garland Publishing, 1980. In abridged version repr. in Mourelatos 1974, pp. 86-95. Hett, W. S. (transl.) 1936. Aristotle, Minor Works. Loeb Classical Library. Cambridge, MA: Harvard University Press. Huffman, C. A. 1993. Philolaus of Croton: Pythagorean and Presocratic: A Commentary on the Fragments and Testimonia with Interpretive Essays. Cambridge: Cambridge University Press. Immisch, O. 1927. Gorgiae Helena. Berlin and Leipzig: Walter de Gruyter. -DQiþHN . >,QGH[HV WR WKH 7HXEQHU HGLWLRQ RI 6H[WXV (PSLULFXV@ In Sexti Empirici Opera, 3 vols, vol. 3. Leipzig: B.G. Teubner. [The index is printed as a separate vol. 4 in later editions.] Joachim, H. H. 1926. Aristotle on Coming-to-Be and Passing-Away. Oxford: Clarendon Press. Jones, O. R. (ed.) 1971. The Private Language Argument. London: Macmillan. Kahn, C. H. 1960. Anaximander and the Origins of Greek Cosmology. New York/London: Columbia University Press. Repr. Indianapolis and Cambridge: Hackett Publishing Company, 1994. Kahn, C. H. 1973. The Verb ‘Be’ in Ancient Greek. – Foundations of Language, suppl. ser., 16. Dordrecht and Boston: D. Reidel. Repr. with a new introductory essay, Indianapolis: Hackett Publishing Company, 2003. Kahn, C. H. 1976. “Why Existence Does Not Emerge as a Distinct Concept in Greek Philosophy,” Archiv für Geschichte der Philosophie, 58, pp. 323-34.

218

Alexander P. D. Mourelatos

Kerferd, G. B. 1981a. The Sophistic Movement. Cambridge: Cambridge University Press. Kerferd, G. B. 1981b. “The Interpretation of Gorgias’ Treatise 3HULWRXPƝRQWRVƝ SHULSK\VHǀV,” Deucalion, 36, pp. 319-27. Kerferd, G. B. 1984. “Meaning and Reference: Gorgias and the Relation Between Language and Reality.” In [Greek Philosophical Society], ‫ܻ ݠ‬ȡȤĮȓĮ ıȠijȚıIJȚțȒ: The Sophistic Movement, Papers Read at the First International Symposium on the Sophistic Movement Organised by the Greek Philosophical Society, 27-29 Sept. 1982. Athens, 1984, pp. 215-22. Keuls, E. C. 1975. “Skiagraphia Once Again,” American Journal of Archaeology, 79, pp. 1-16. Keuls, E. C. 1978. Plato and Greek Painting. Columbia Studies in the Classical Tradition, 5. Leiden: E.J. Brill. Kirk, G. S. 1954. Heraclitus: The Cosmic Fragments. Cambridge: Cambridge University Press. Kirk, G. S. and Raven, J. E. 1957. The Presocratic Philosophers. Cambridge: Cambridge University Press. Kirk, G. S., Raven, J. E. and Schofield, M. 1983. The Presocratic Philosophers. 2nd ed. Cambridge: Cambridge University Press. Klowski, J. 1967. “Zum Entstehen der Begriffe Sein und Nichts und der Weltentstehungs- und Weltschöpfungstheorien im strengen Sinne” (I. Teil), Archiv für Geschichte der Philosophie, 49, pp. 121-48. Kroll, G. (ed.) 1902. Syriani in Metaphysica Commentaria. Commentaria in Aristotelem graeca, 6. Berlin: Georg Reimer. Kühn, C. G. (ed.) 1825. Claudii Galeni Opera omnia. Medicorum Graecorum opera quae extant. Vol. 10: De methodo medendi. Repr. Hildesheim: Olms, 1965. Kühn, C. G. (ed.) 1830. Claudii Galeni Opera omnia. Medicorum Graecorum opera quae extant. Vol 19: Pseudo-Galen, Definitiones medicae. Repr. Hildesheim: Olms, 1965. Laks, A. and Louguet, C. (eds.) 2002. Qu’est ce que la philosophie présocratique? Lille: Presses Universitaires du Septentrion. Laks, A. and Most, G. W. (eds.) 2016. Early Greek Philosophy, 9 volumes, Loeb Classical Library. Cambridge, MA, and London: Harvard University Press. Lattimore, R. 1947. The Odes of Pindar. Chicago: The University of Chicago Press.

Bibliography

219

Lee, E. N., Mourelatos, A. P. D., and Rorty, R. M. (eds.) 1973. Exegesis and Argument: Studies in Greek Philosophy Presented to Gregory Vlastos. Phronesis, suppl. vol. 1. Assen: Royal van Gorcum. Also New York: Humanities Press. Lesher, J. H. 1992. Xenophanes of Colophon: Fragments, a Text and Translation with a Commentary. The Phoenix Presocratics, 4. Toronto, Buffalo and London: University of Toronto Press. Liddell, H. G. and R. Scott 1940. A Greek-English Lexicon. 9th ed., rev. H. S. Jones and R. McKenzie. Oxford: Oxford University Press. [Cited as LSJ.] Lloyd, G. E. R. 1979. Magic, Reason and Experience. Cambridge: Cambridge University Press. Lobel, E. and Page, D. 1955. Poetarum Lesbiorum Fragmenta. Oxford: Clarendon Press. Löbl, R. 1989. Demokrit: Texte zu seiner Philosophie, ausgewählt, übersetzt, kommentiert und interpretiert. Elementa-Text, 4. Amsterdam/Atlanta: Rodopi. Loenen, J. H. M. M. 1959. Parmenides, Melissus, Gorgias: A Reinterpretation of Eleatic Philosophy. Assen: Van Gorcum. Loveday, T. and Forster E. S. (transl.) 1913. De Melisso, Xenophane, Gorgia. In W. D. Ross (ed.), The Works of Aristotle. Vol. VI, Opuscula. Oxford: Clarendon Press. Lovejoy, A. O. 1936. The Great Chain of Being. Cambridge, MA: Harvard University Press (repr. 1964). Lur’e [Luria], S. J. 1970. Demokrit: teksty, perevod, issledovanija. Leningrad: Nauka. Lyons, J. 1977. Semantics. 2 vols. Cambridge: Cambridge University Press. MacDowell, D. M. 1982. Gorgias: Encomium of Helen. Edited with Introduction, Notes and Translation. Bristol: Bristol University Press. Mansfeld, J. 1972. “Ambiguity in Empedocles B17,3-5: A Suggestion,” Phronesis, 17, pp. 17-37. Mansfeld, J. 1979-1980. “The Chronology of Anaxagoras’ Athenian Period and the Date of His Trial,” part I, Mnemosyne, 32 (1979), pp. 39-69; part II, Mnemosyne 33 (1980), pp. 17-95. Mansfeld, J. 1985. “Historical and Philosophical Aspects of Gorgias’ ‘On What Is Not’.” In Montoneri–Romano 1985, pp. 243-71.

220

Alexander P. D. Mourelatos

Mazzara, G. 1982. Gorgia: Ontologo e metafisico. Palermo: Ila Palma. McKirahan, R. D., Jr. 1994. Philosophy Before Socrates: an Introduction with Texts and Commentary. Indianapolis/Cambridge: Hackett Publishing Co. Meinwald, C. C. 1992. “Good-Bye to the Third Man.” In R. Kraut (ed.), The Cambridge Companion to Plato. Cambridge: Cambridge University Press, pp. 365-96. Moline, J. 1981. Plato’s Theory of Understanding. Madison: University of Wisconsin Press. Montoneri, L. and Romano, F. (eds) 1985. Gorgia e la sofistica: Atti del convegno internazionale (Lentini-Catania, 12-15 dic[embre] 1983), Siculorum Gymnasium, N.S. a. XXXVIII nos. 1-2. Catania: Fac. di Lettere e Filosofia, Univ. di Catania. Moorhouse, A. C. 1959. Studies in the Greek Negatives. Cardiff: University of Wales Press. Moorhouse, A. C. 1965. “A Use of Ƞ੝įİȓȢ DQG ȝȘįİȓȢ´ Classical Quarterly, n.s. 15, pp. 31-40. Morel, P.-M. 1996. Démocrite et la recherche des causes. Paris: Klincksieck. Mourelatos, A. P. D. 1984c. “૽ȅīȠȡȖȓĮȢȖȚ੹ IJ੽ ȜİȚIJȠȣȡȖȓĮIJોȢȖȜȫııĮȢ” In ‫ݠ‬ ܻȡȤĮȓĮ ıȠijȚıIJȚțȒ/The Sophistic Movement. Athens: Greek Philosophical Society, pp. 223-31. Mourelatos, A. P. D. (ed.) 1974. The Pre-Socratics: A Collection of Critical Essays. Garden City, NY: Anchor Press/Doubleday. Repr. Princeton: Princeton University Press 1993. Mourelatos, A. P. D. 1970/2008a, The Route of Parmenides: A Study of Word, Image, and Argument in the Fragments. 1970, New Haven, Conn.: Yale University Press. Revised and expanded ed., 2008, Las Vegas: Parmenides Publishing. [Part I of the 2008 edition is identical in its pagination to that of the main text of the original edition. Additional material in the 2008 edition: new introduction, in Roman-numbered pagination; reprinting of three previously published essays by the author; and a previously unpublished essay by Gregory Vlastos.] Mourelatos, A. P. D. 1973. “Heraclitus, Parmenides, and the Naïve Metaphysics of Things.” In Lee–Mourelatos–Rorty 1973, pp. 16-48. Repr. in Mourelatos 2008a, pp. 299-332. Mourelatos, A. P. D. 1976. “Determinacy and Indeterminacy, Being and NonBeing in the Fragments of Parmenides.” In R. A. Shiner and J. King-Farlow

Bibliography

221

(eds.), New Essays on Plato and the Pre-Socratics. Canadian Journal of Philosophy, suppl. vol. 2, pp. 45-60. Repr. in Mourelatos 2008a, pp. 333-49. Mourelatos, A. P. D. 1979. “‘Nothing’ as ‘Not-Being’: Some Literary Contexts That Bear on Plato.” In Arktouros, Hellenic Studies Presented to Bernard M. W. Knox, ed. by G. W. Bowersock, W. Burkert, and M. C. J. Putnam. Berlin and New York: Walter De Gruyter, pp. 319-329. Repr. in J. P. Anton and A. Preus (eds.). Essays in Ancient Greek Philosophy, Volume Two. Albany: State University of New York Press, 1983, pp. 59-69. Repr. in this volume pp. 35-44. Mourelatos, A. P. D. 1981a. “Saving the Presocratics,” Philosophical Books, 22, pp. 65-77. Mourelatos, A. P. D. 1981b. “਺ȡȐțȜİȚIJȠȢ, ȆĮȡȝİȞȓįȘȢ țĮȚ ਲ ਖʌȜȠȧț੽ ȝİIJĮijȣıȚț੽ IJ૵Ȟ ʌȡĮȖȝȐIJȦȞ,” ǻİȣțĮȜȓȦȞ, nos. 33/34, pp. 57-75. Mourelatos, A. P. D. 1984a. “Aristotle’s Rationalist Account of Qualitative Interaction,” Phronesis, 29, pp. 1-16. Mourelatos, A. P. D. 1984b. “ǻȘȝȩțȡȚIJȠȢ: ijȚȜȩıȠijȠȢ IJોȢ ȝȠȡijોȢ” [Democritus: Philosopher of Form]. In Proceedings of the First International Congress on Democritus, 2 vols. Xanthi: International Democritean Foundation (vol. 1, pp. 109-19). [English summary, pp. 118-19.] Mourelatos, A. P. D. 1987. “Gorgias on the Function of Language,” Philosophical Topics, 15, pp. 135-70. Repr. in this volume pp. 180-209. Mourelatos, A. P. D. 2000. “ȅ੿ ਕʌĮȡȤȑȢ IJોȢ ijȚȜȠıȠijȚțોȢ ਩ȞȞȠȚĮȢ IJȠ૨ țĮșȩȜȠȣ” [Beginnings of the philosophical concept of the universal], ȆȡĮțIJȚț‫ ޟ‬IJ߱Ȣ ݃țĮįȘȝȓĮȢ ݃șȘȞࠛȞ [Proceedings of the Academy of Athens], 75, pp. 50925 [English summary, p. 526.] Mourelatos, A. P. D. 2002a. “Xenophanes’ Contribution to the Explanation of the Moon’s Light,” Philosophia (Athens), 32, pp. 47-59. Mourelatos, A. P. D. 2002b. “La terre et les étoiles dans la cosmologie de Xénophane.” transl. C. Louguet. In Laks–Louguet 2002, pp. 331-50. Mourelatos, A. P. D. 2003. “Democritus on the Distinction between Universals and Particulars.” In Bächli–Petrus 2003, pp. 43-56. Repr. in this volume pp. 143-155. Mourelatos, A. P. D. 2005. “Intrinsic and Relational Properties in the Democritean Ontology,” in Salles 2005, pp. 39-63. Repr. in this volume pp. 156-177. Mourelatos, A. P. D. 2008a. See Mourelatos 1970/2008a.

222

Alexander P. D. Mourelatos

Mourelatos, A. P. D. 2008b. “The Cloud-Astrophysics of Xenophanes and Ionian Material Monism.” In Curd–Graham 2008, pp. 134-68. Mourelatos, A. P. D. 2013. “Xénophane et son ‘astro-néphologie’ dans les Nuées,” transl. M. Gondicas and A. Laks. In A. Laks and R. Saetta Cottone (eds.), Comédie et Philosophie: Socrate et les ‘Présocratiques’ dans les Nuées d’Aristophane. Études de littérature ancienne, 21. Paris: Éditions Rue d’Ulm, pp. 31-60. Mourelatos, A. P. D. 2016. “‘Limitless’ and ‘Limit’ in Xenophanes’ Cosmology and in His Doctrine of Epistemic ‘Construction’ (dokos),” In K. Ierodiakonou and P. S. Hasper (eds.), Ancient Epistemology. Logical Analysis and History of Philosophy, 19, pp. 16-37. Mourelatos, A. P. D. 2020. “All the Texts for Xenophanes of Colophon. Critical Discussion of [Strobel–Wöhrle 2018],” Rhizomata, 8, pp. 1-17. Mourelatos, A. P. D. and Pulpito, M. 2018. “Parmenides and the Principle of Sufficient Reason.” In Pulpito–Spangenberg 2018, pp. 121-41. Mourelatos, A.P.D. 1981c. “Pre-Socratic Origins of the Principle That There Are No Origins from Nothing,” The Journal of Philosophy, 78, pp. 649-65. Repr. in this volume pp. 20-34. Nehamas, A. 1975. “Confusing Universals and Particulars in Plato’s Early Dialogues,” Review of Metaphysics, 29, pp. 287-306. Repr. in Nehamas 1999, ch. 8, pp. 159-75. Nehamas, A. 1999. Virtues of authenticity: Essays on Plato and Socrates. Princeton: Princeton University Press. Newiger, H.-J. 1973. Untersuchungen zu Gorgias’ Schrift Über das Nichtseiende. Berlin: Walter de Gruyter. Nikolaou, S.-M. 1998. Die Atomlehre Demokrits und Platons Timaios: eine vergleichende Untersuchung. Beiträge zur Altertumskunde, 112. Stuttgart: B.G. Teubner. Nussbaum, M. 1979. “Eleatic Conventionalism and Philolaus on the Conditions of Thought,” Harvard Studies in Classical Philology, 83, pp. 63-108. Nussbaum, M. 1987. “Commentary on Mourelatos,” Proceedings of the Boston Area Colloquium in Ancient Philosophy (ed. by J. J. Cleary), 2, pp. 195207. O’Brien, D. 1968. “The Relation of Anaxagoras and Empedocles,” The Journal of Hellenic Studies, 88, pp. 93-113.

Bibliography

223

Owen, G. E. L. 1966, “Plato and Parmenides on the Timeless Present,” The Monist, 50, pp. 317-340. Repr. in Mourelatos 1974, pp. 271-92. Owen, G. E. L. 1971. “Plato on Not-Being.” In G. Vlastos (ed.), Plato, I, Metaphysics and Epistemology. New York: Doubleday and Co., pp. 223-67. ȆĮȞȑȡȘȢ, I. P. 1984. ‫ ݠ‬țȠıȝȠȜȠȖȓĮ IJȠࠎ ȆȜȐIJȦȞĮ ı‫ ޡ‬ıȤȑıȘ ȝ‫ ޡ‬IJ‫ޣ‬Ȟ ܻIJȠȝȚț‫ ޣ‬șİȦȡȓĮ IJȠȣ ǻȘȝȩțȡȚIJȠȣ [Plato’s cosmology in relation to the atomic theory of Democritus]. Xanthi: International Democritean Foundation. Peirce, C. S. 1931-1960. Collected Papers. Ed. by C. Hartshorne and P. Weiss, 8 vols. Cambridge, MA: Harvard University Press. Peirce, C. S. 1998. The Essential Peirce. Ed. by the Peirce Project. 2 vols. Bloomington and Indianapolis: Indiana University Press. Prauss, G. 1966. Platon und der Logische Eleatismus. Berlin: Walter de Gruyter. Pulpito, M., and Spangenberg, P. (eds.) 2018. ‫ݸ‬įȠ‫ ޥ‬ȞȠ߱ıĮȚ, Ways to Think. Essays in Honour of Néstor-Luis Cordero. Bologna: Diogene. Rau, R. 1923. “Demokritos fr. 124 Diels,” Philologische Wochenschrift, 43, issue 35, column 838-40. Reale, G. 1970. Melisso: Testimonianze e frammenti. Florence: La Nuova Italia. Reinhardt, K. 1916. Parmenides und die Geschichte der griechischen Philosophie. Bonn: Friedrich Cohen. Riepe, D. 1961. The Naturalistic Tradition in Indian Thought. Seattle: University of Washington Press. Robinson, J. M. 1968. An Introduction to Early Greek Philosophy. Boston: Houghton Mifflin. Ross, W. D. 1953. Aristotle’s Metaphysics: A Revised Text with Introduction and Commentary. 2 vols. Oxford: Clarendon Press. Rossetti, L. 2015. La filosofia non nasce con Talete – e nemmeno con Socrate. Bologna: Diogene Multimedia. Russell, B. 1903. The Principles of Mathematics. Cambridge: Cambridge University Press. Russell, B. 1956. Logic and Knowledge: Essays 1901-1950. Ed. by R. C. Marsh. London: George Allen & Unwin. Sale, W. 1973. Electra, by Sophocles: A Translation with Commentary. Englewood Cliffs, N.J.: Prentice-Hall.

224

Alexander P. D. Mourelatos

Salem, J. 1996. Démocrite: grains de poussière dans un rayon de soleil. Paris: Vrin. Salles, R. (ed.) 2005. Metaphysics, Soul and Ethics: Themes from the Work of Richard Sorabji. Oxford: Oxford University Press. Santillana, G. de 1967. “Prologue to Parmenides.” In W. D. Bradeen et al. (eds.), Lectures in Memory of Louise Taft Semple, First Series, 1961-1965. Princeton: Princeton University Press, pp. 45-93. Schofield, M. 1980. An Essay on Anaxagoras. Cambridge: Cambridge University Press. Schwabe, W. 1980. ‘Mischung’ und ‘Element’ im Griechischen bis Platon. – Archiv fur Begriffsgeschichte, suppl. vol. 3. Bonn: Bouvier Verlag. Segalerba, G. 2001. Note su Ousia: volume primo. Florence: Edizioni ETS. Segalerba, G. 2003. “Numerische Einheit als ontologisches Kriterium: zur Unterscheidung der Entitäten bei Aristoteles,” Wiener Jahrbuch für Philosophie, 35, pp. 59-96. Sicking, C. M. J. 1976. “Gorgias und die Philosophen,” In C. J. Classen (ed.), Sophistik. Wege der Forschung, 187. Darmstadt: Wissenschaftliche Buchgesellschaft, pp. 384-407. Previously in Mnemosyne, ser. 4, no. 17, 1964, pp. 225-47. Sider, D. 1981. The Fragments of Anaxagoras. Beiträge zur klassischen Philologie, 118. Meisenheim am Glan: Anton Hain. Solmsen, F. 1950. “Tissues and the Soul,” Philosophical Review, 59, pp. 435-68. Repr. in Solmsen 1968, I, pp. 502-35. Solmsen, F. 1968. Kleine Schriften. 2 vols. Hildesheim: Olms. [A third volume was published in 1982.] Solmsen, F. 1969. “The ‘Eleatic One’ in Melissus,” Medelingen der Koninklijke Nederlandse Akademie van Wetenschappen, Afd. Letterkunde, nieuwe reeks, 32, pp. 221-33. Solmsen, F. 1971. “The Tradition about Zeno of Elea Re-examined,” Phronesis, 16, pp. 116-141. Repr. in Mourelatos 1974, pp. 368-93. Solmsen, F. 1975. “Eternal and Temporary Beings in Empedocles’ Physical Poem,” Archiv für Geschichte der Philosophie, 57, pp. 123-45. Solmsen, F. 1977. “Epicurus on Void, Matter and genesis: Some Historical Observations,” Phronesis, 23, pp. 263-81.

Bibliography

225

Sorabji, R. 1983. Time, Creation, and the Continuum: Theories in Antiquity and the Early Middle Ages. Ithaca: Cornell University Press. Sorabji, R. 1988. Matter, Space, and Motion: Theories in Antiquity and their Sequel. Ithaca: Cornell University Press. Sprague, R. K. (ed.) 1972. The Older Sophists: A Complete Translation by Several Hands of the Fragments. Columbia, SC: University of South Carolina Press. Stevenson, C. L. 1944. Ethics and Language. New Haven: Yale University Press. Stokes, M. C. 1971. One and Many in Presocratic Philosophy. Cambridge, MA: Harvard University Press. Stough, C. 1969. Greek Skepticism: A Study in Epistemology. Berkeley: University of California Press. Strang, C. 1963. “The Physical Theory of Anaxagoras,” Archiv für Geschichte der Philosophie, 45, pp. 101-118. Repr. in Allen–Furley 1970-1975, vol. 2, pp. 361-80. Stratton, G. M. 1917. Theophrastus and the Greek Physiological Psychology Before Aristotle. London: George Allen & Unwin. Repr. Amsterdam: E. J. Bonset & P. Schippers, 1964. Strobel, B. and Wöhrle, G. 2018 (E. Wakelnig & C. Vassallo, collaborators), Xenophanes von Kolophon, Traditio Praesocratica 3. Berlin and Boston: Walter de Gruyter. Stückelberger, A. 1979. Antike Atomphysik. Munich: Heimeran Verlag. Sykutris, J. 1928. Review of O. Immisch, Gorgiae Helena. Gnomon, 4, pp. 11-18. Taylor, A. E. 1906. Aristotle on His Predecessors: Metaphysics Alpha, Translated with notes and introduction. La Salle, IL: Open Court. Taylor, C. C. W. 1999. The Atomists, Leucippus and Democritus: Fragments, a Text and Translation with a Commentary. Phoenix suppl. vol. 36 (= Phoenix Presocratics vol. 5). Toronto: University of Toronto Press. Teodorsson, S.-T. 1982. Anaxagoras’ Theory of Matter. Studia Graeca et Latina Gothoburgensia, 43. Göteborg: Acta Universitatis. Ullmann, S. 1957. The Principles of Semantics. Oxford/Glasgow: Blackwell/Jackson, Son & Co. Untersteiner M. 1961. Sofisti: Testimonianze e frammenti. 2 vols., 2nd ed. Florence: La Nuova Italia. Urmson, J. O. 1992. Simplicius: On Aristotle’s Physics 4.1-5, 10-14. Ithaca: Cornell University Press.

226

Alexander P. D. Mourelatos

Vitali, R. 1973. Melisso di Samo: Sul mondo o sull’essere. Una interpretazione dell’eleatismo. Urbino: Argalia. Vlastos, G. 1965. “Degrees of Reality in Plato,” in R. Bambrough (ed.), New Essays on Plato and Aristotle. London: Routledge & Kegan Paul, pp. 1-19. Vlastos, G. 1975. Plato’s Universe. Seattle: University of Washington Press. Repr. Las Vegas: Parmenides Publishing, 2006. Watson, J. B. 1930. Behaviorism. Revised ed. New York: Norton. Weinberg, J. R. 1936. An Examination of Logical Positivism. London: Kegan Paul. Weinberg, J. R. 1965. Abstraction, Relation, and Induction. Three Essays in the History of Thought. Madison: University of Wisconsin Press. White, S. A. 2008. “Milesian Measures: Time, Space, and Matter.” In Curd– Graham 2008, pp. 89-133. Williams, C. J. F 1982. Aristotle’s De generatione et corruptione, Translated with Notes. Oxford: Clarendon Press. Wright, M. R. 1981. Empedocles: the Extant Fragments, Edited with an Introduction, Commentary, and Concordances. New Haven, Conn. and London: Yale University Press.

Indexes

INDEX OF PROPER NAMES 1 Adam, James: 44 n. 16. Adams, Robert M.: 193 n. 27. Aeschylus: 40. Aëtius: 114, 147, 161 n. 10, 165 n. 21, 177. Alcaeus: 82 and n. 7. Alcmaeon: 95-96. Alexander, Samuel: 90. Allen, Reginald E.: 133 n.12. Alston, William P.: 199 n. 34, 200 n. 37, 202 and nn. 38, 40-41, 204 n. 47, 210 n. 63. Anaxagoras: 51, 59, 62, 63 nn. 20-21, 66-69, 73, 74 and n. 34, 79, 80 n. 2, 84, 86, 90-102, 105 and n. 47, 108-111, 118-122, 126, 134, 171 n. 34. Anaximander: 25, 50 n. 2, 51, 54-57, 83, 133-134. Anaximenes: 25, 54-55, 57, 83, 133. Annas, Julia: 131 n. 9, 132 n. 11. Anton, John P.: 36. Apollodorus: 119. Aristophanes: 40, 58, 73 n. 33, 205 n. 53. Aristotle: 21-22, 24 n. 9, 34, 51, 55, 65, 74-75, 78, 81 and n. 5, 86, 90 1

Parmenides references are assembled only in the Index of Cited Texts.

n. 18, 95, 107-110, 115, 118, 125, 127-132, 134-137, 139-141, 143144, 146-147, 149-150, 152-155, 159- 162, 163, 166, 170-172, 175178, 184. Armstrong, David M.: 128, 130. Audi, Robert: 130. Augustine: 81. Austin, Scott: 23 n. 7, 151 n. 7, 158 n. 17. Ayer, Alfred J.: 202-204. Bächli, Andreas: 127 n.1, 146. Barnes, Jonathan: 22-27, 34 and n. 30, 60 n. 18, 69, 79 n. 1, 80 n. 4, 86 n. 13, 88 n. 14, 97 and n. 32, 120 n. 78, 134, 168 n. 31, 186 n. 10. Beare, John I.: 107. Bechtle, Gerald: 132 n. 10, 145 n. 34. Bekker, A. Immanuel: 186. Benitez, Rick: 212 n. 67. Berkeley, George: 188, 193 n. 27, 209. Berryman, Sylvia: 171 n. 34, 172 n. 36. Bloomfield, Leonard: 204, 206. Bollack, Jean: 30 n. 21-22, 88 n. 14, 104 n. 54, 112 n. 65, 114 n. 68, 116 n. 73, 121 n. 79. Bonitz, Hermann: 176. Bonnet, Charles: 100. Bowersock, Glen W.: 36. Brentlinger, John: 135 n. 14.

228 Brink, B. ten: 149, 153. Broadie, Sarah W.: 189 n. 18, 212 n. 67. Bruno, Vincent J.: 89 n. 15, 94 n. 27, 119 and n. 76. Burkert, Walter: 36, 50 and n. 2, 51, 53 n. 5, 69 and n. 25, 70, 72-73, 137-138. Burnet, John: 27 n. 17, 31 n. 24, 50, 92 n. 21. Bury, Robert G.: 184 n. 5, 186 n. 10, 208 n. 56. Calogero, Guido: 183 and nn. 1-3, 184, 190 n. 23, 197 n. 31, 208 nn. 56, 59. Capriglione, Iolanda: 49. Cartledge, Paul: 179 n. 46. Casertano, Giovanni: 21 n. 2. Cassin, Barbara: 184 n. 3, 187 n. 11, 190 n. 20. Caston, Victor: 140 n. 26. Churchland, Paul M.: 104 n. 44. Cicero: 157 n. 16, 158, 167 n. 29. Cook Wilson, John: 187, 188 n. 13. Cornford, Francis M.: 50. Curd, Patricia: 60 n. 18, 168 n. 31. De Ley, Herman: 147 n.3. Dean-Jones, Lesley: 158 n. 17. Democritus: 51-52, 65, 74-79, 83-84, 101, 103-104, 109 n. 58, 123-126, 136 and nn. 15-17, 139-163, 165169, 171-180, 211. Demosthenes: 40 n. 9. Diels, Hermann: 18, 25, 27 n. 17, 82 n. 6, 116 n. 71, 129 n. 5, 147, 148, 184 nn. 3, 5, 187 and n. 11, 188 n. 13, 190 nn. 20, 23, 205 n. 151, 211 n. 64. Dimitrakos, Dimitri: 176 n. 43.

Alexander P. D. Mourelatos

Diogenes of Apollonia: 79, 81. Diogenes Laertius: 50 n. 1, 158. Donadi, Francesco: 205 n. 51. Dragona-Monachou, Myrto: 50. Empedocles: 22 and n. 4, 24, 28-35, 51, 59, 62-66, 68-69, 72-74, 79-81, 84-90, 95, 97, 102, 104-126, 134, 136-137, 171 n.34, 186, 196. Epicurus: 34, 74, 77, 140 and n. 24, 165. Eudemus: 172 n. 36. Euripides: 40 and n. 9. Forster, Edward S.: 184 n. 3. Fränkel, Hermann: 56 n. 9. Frege, Gottlob: 81 n. 5, 200 n. 36. Fritz, Kurt von: 161-162, 176 n. 44. Furley, David J.: 36 n. 2, 83 n. 10, 91 n. 19, 99 and n. 36, 101 n. 42, 105 and n. 46, 135 n. 13, 145 n. 33. Gagarin, Michael: 45 n. 23. Galen: 115, 130-131, 142-143, 147, 149, 157 n. 15. Gallop, David: 21. Garmong, Dina: 158 n. 17. Gigon, Olof: 186 n. 9, 196 n. 29, 210 n. 62. Gill, Mary Louise: 127. Goldin, Owen: 109 n. 57. Gomperz, Heinrich: 148-149. Gomperz, Theodor: 21. Gorgias: 63 n. 20, 129, 132, 183-212. Goudge, Thomas A.: 85 n. 12, 100 n. 40. Graeser, Andreas: 127 n. 1, 146, 184 n. 5, 189 n. 17. Granger, Herb: 158 n. 17. Gregory of Nazianzus: 93. Guthrie, William K. C.: 49-51, 88 n.

Indexes

14, 171 n. 34, 184 n.5. Hamilton, Hollister A.: 44 nn. 17-18, 21. Hankinson, R. J.: 131 nn. 7-8, 157 n. 15, 160 n. 8. Hasper, Pieter S.: 166 n. 23, 168 n. 32, 171 n. 34, 172 n. 37. Hayduck, Michael: 154 n. 10. Heath, Sir Thomas L.: 50. Hegel, Georg W. F.: 80. Heidegger, Martin: 81 n. 5. Heidel, William A.: 133. Heraclitus: 21, 25, 52, 55-56, 79, 80 n. 3, 83, 95, 121, 150 n. 6, 179. Hett, Walter S.: 184 n. 3. Hippocrates: 115, 147. Homer: 39 and n. 6, 40 and n. 9, 44 and n. 18, 112 and n. 66, 135, 142 n. 30. Hossaini, Ali: 90 n. 17. Huffman, Carl A.: 70 and n. 27, 72 n. 28-29, 138 and n. 20-21. Hull, Clark L.: 204. Hume, David: 188, 193. Immisch, Otto: 206 n. 54. -DQiþHNKarel: 209 n. 60. Joachim, Harold H.: 161 n. 9. Jones, Owen R.: 203 n. 45, 204 n. 46. Kahn, Charles H.: 40 n. 9, 45 and n. 22, 54 and n. 6, 83 n. 9, 90 n. 18, 142 n. 30. Kant, Immanuel: 98. Kennedy, George: 186 n. 10, 205 n. 51. Kerferd, George B.: 184 n. 5, 185 n. 7, 186 n. 9-10, 187 and nn. 11-12, 188 nn. 13-14, 197 n. 31, 199 nn. 35-36, 200 n. 36, 208 nn. 56-57. Keuls, Eva C.: 89 n. 15, 90 n. 16, 118-

229 120. Kirk, Geoffrey S.: 80 n. 2, 88 n. 14. Knox, Bernard M. W.: 36. Kranz, Walther: 25 n. 11, 27 n. 17, 82 n. 6, 129 n. 5, 147, 184 nn. 3, 5, 205 n. 51, 211 n. 64. Kroll, Guilelmus: 131 n. 10. Kronz, Fred: 130 n. 6. Kühn, Carl G.: 131 and n. 8, 142, 147, 157 n. 15. Laks, André: 63 n. 21, 74 nn. 35-36, 75 n. 37, 77 nn. 38, 40-42. Lattimore, Richmond A.: 39 n. 7. Lesher, James H.: 56. Leucippus: 51, 74-75, 78, 101, 123, 139-141, 143, 155, 159-163, 166, 169, 171 n. 34. Liddell, Henry G.: 94 n. 27. Lloyd, G. E. R.: 205 nn. 51, 53, 212 n. 67. Lobel, Edgar: 82 n. 7. Löbl, Rudolf: 171 n. 34. Locke, John: 188 and n. 15, 202 n. 39. Loenen, J. H. M. M.: 185 n. 6, 188 n. 13, 189 n. 17, 208 nn. 56, 58. Loveday, Thomas: 184 n. 3. Lovejoy, Arthur O.: 145. Lucretius: 34, 74, 140 and n. 24. Lur’e [Luria], Salomon J.: 148 n. 4, 153 nn. 8-9, 158 n. 17, 159 n. 4, 160 n. 5, 161 n. 10, 167 nn. 26-29, 171 n. 34, 172 n. 36. Lyons, John: 202 n. 40, 204 nn. 47, 49, 210 n. 63. MacDowell, Douglas M.: 205 n. 51. Mansfeld, Jaap: 90 n. 18, 113 n. 67, 184 nn. 3, 5, 188 n. 15, 191 n. 25, 197 n. 30. Mazzara, Giuseppe: 186 n. 9, 188 n.

230 15. McKim, Richard: 189 n. 18, 212 n. 67. McKirahan, Richard jr.: 136 n. 15, 139 n. 22, 158 n. 17, 159 n. 4, 160 n. 5, 165 nn. 20, 22, 169 n. 33, 173 n. 38, 177 n. 45, 179 n. 46, 180 n. 47. Meinwald, Constance C.: 135 n. 13. Melissus: 21-22, 25-28, 33-35, 51, 58, 60, 62-63, 67-70, 73, 79-80, 83, 86-97, 133. Michael of Ephesus: 154 and n. 10, 155. Michalakopoulou-Veïkou, Christina: 50. Moline, Jon: 135 nn. 13-14. Montoneri, Luciano: 183. Moorhouse, A. C.: 38-39, 41, 44 nn. 17, 20. Morel, Pierre-Marie: 147 n. 3, 174 n. 40. Moschonas, S.: 127. Most, Glenn W.: 63 n. 21, 74 nn. 3536, 75 n. 37, 77 nn. 38, 40-42. Nehamas, Alexander: 128. Newiger, Hans-Joachim: 184 and nn. 4-5, 186 n. 10, 188-190,197 n. 30, 208 nn. 56, 58-59. Nikolaou, Sousanna-Maria: 145 n. 32. Nussbaum, Martha: 69 n. 26, 78 n. 45, 79. O’Brien, Denis: 90 n. 18. Owen, G. E. L.: 23 and n. 6, 24 and n. 9, 36 and nn. 1-3, 109 n. 57. Page, Denys L.: 82 n. 7. Parmenides: see under Index of Cited Texts. Panzerbieter, Friedrich: 30-31. Peirce, Charles S.: 130-132, 144, 202 n. 38. Pellegrin, Pierre: 127.

Alexander P. D. Mourelatos

Petrus, Klaus: 146. Philolaus: 51, 59, 62, 69-75, 78-79, 84, 95, 101-103, 107, 110, 124, 126, 137-139. Philoponus: 109 n.57, 121, 141 n. 28, 172. Pindar: 39. Plato: 25, 34, 36, 39, 43-44, 49, 63, 67, 78, 83, 86, 89, 97-100, 103, 119, 121-122, 127-135, 139-140, 142, 144-147, 156-157, 166-167, 189, 191, 197, 199. Plutarch: 33, 112 n. 65, 116 n. 73, 121 n. 79. Polygnotus: 119. Prauss, Gerold: 78 n. 44. Preus, Anthony: 36. Prodicus: 129. Protagoras: 86, 143, 153 n. 9, 155, 183, 197 n. 31. Pulpito, Massimo: 23 n. 7, 24 n. 8. Putnam, Michael C. J.: 36. Pythagoras: 34, 49-51, 53, 72, 83, 137138. Rau, R.: 127, 148. Raven, John E.: 80 n. 2, 88 n. 14. Reale, Giovanni: 27 n. 17. Reinhardt, Karl L.: 80 n. 3. Riepe, Dale: 90 n. 17. Robinson, John M.: 186 n. 10. Romano, Francesco: 183. Ross, William D.: 160 n. 8. Rossetti, Livio: 53 n. 5. Russell, Bertrand: 81, 200 and n. 37, 204 and n. 44. Sale, William: 40 n. 8. Salem, Jean: 147 n. 3, 160 n. 8, 174 n. 40. Salles, Ricardo: 159, 180 n. 47.

Indexes

Santillana, Giorgio de: 23 n. 7. Schofield, Malcolm: 80 n. 2, 88 n. 14, 90 n. 18, 92 and nn. 20-22, 93 nn. 25-26, 96 n. 31, 98-99. Schrenk, Lawrence P.: 212 n. 67. Schwabe, Wilhelm: 95 nn. 29-30. Scott, Robert: 94 n. 27, 151 n. 7, 158 n. 17. Segalerba, Gianluigi: 127 n. 3, 131 n. 9, 132 n. 11, 145 n. 34. Seibt, Johanna: 145 n. 34. Seneca: 150 n. 6. Sextus Empiricus: 93, 94, 96, 143, 152-153, 155-157, 184-186, 189190, 198-199, 201, 206-212. Shakespeare, William: 58 and n. 15, 122. Sharma, Ravi: 158 n. 17. Sicking, Christiaan M. J.: 186 n. 9. Sider, David: 90 n. 18, 92 n. 21, 93 n. 26. Simplicius: 27-28, 116 n. 73, 167 n. 30. Skinner, Burrhus F.: 206. Socrates: 127-129, 132-133, 136 n. 15, 145, 152, 155-157. Solmsen, Friedrich: 33 n. 27, 34 n. 29, 25 n. 31, 45 n. 23, 83 n. 11, 105 and nn. 45-46, 107 n. 51, 109 n. 60. Sophocles: 40, 42, 44. Sorabji, Richard: 158 n. 17, 159 n. 1, 168 n. 32. Sprague, Rosamond K.: 186 n. 10, 205 n. 51. Stevenson, Charles L: 202 n. 38, 204, 206. Stokes, Michael C.: 22-25, 34, 80 n. 4. Stough, Charlotte L.: 209 n. 60.

231 Strang, Colin: 92 and n. 20. Stratton, George M.: 122 n. 80. Strobel, Benedikt: 56-57. Stückelberger, Alfred: 141 n. 29. Sykutris, Johannes: 206 n. 54. Syrianus: 131 and n. 10, 144. Tannery, Paul: 92 n. 21. Taylor, Alfred E.: 50, 159 n. 4. Taylor, Christopher C. W.: 139 n. 22, 140 n. 26, 141 n. 28, 145 n. 33, 147 n. 2, 150 n. 6, 155 n. 13, 159160, 165 n. 22, 169 n. 33, 171-172, 174-175. Teodorsson, Sven-Tage: 90 n. 18, 92 n. 21, 99-100. Thales: 21 and n. 1, 53-54, 82-83. Theophrastus: 108-109, 155, 179-180. Ullmann, Stephen: 202 and n. 39. Untersteiner, Mario: 184 n. 3, 205 n. 51, 208 n. 56. Urmson, James O.: 172 n. 36. Vitali, Renzo: 27 n. 17. Vlastos, Gregory: 41 n. 10, 43 and n. 15, 83 n. 9, 92 n. 21. Watson, John B.: 204 and n. 48. Weinberg, Julius R.: 49, 69 and n. 24, 203 n. 43. Weiss, Albert P.: 204. White, Stephen A.: 53 n. 5. Wildberg, Christian: 49. Williams, Christopher J. F.: 108 and n. 55, 110 n. 62, 161 n. 9, 166 n. 25. Wittgenstein, Ludwig J. J.: 81 n. 5, 203. Wöhrle, Georg: 56-57. Woodruff, Paul: 212 n. 67. Wordsworth, William: 90. Wright, M. Rosemary: 33 and n. 26,

232

Alexander P. D. Mourelatos

88-89, 105-109, 112-113, 115-116, 120-121, 124 n. 83. Xenophanes: 25, 56-58, 62, 73, 80 n. 3, 83, 133-134. Zeno: 34-35, 79-80, 83-84, 86. Zeuxis: 119.

INDEX OF PASSAGES 2 Alcaeus: 76 (Bergk) = 23 (Diehl) = 320 (Lobel and Page), 82 n. 7. Anaxagoras: B1, 100; B4, 67, 99; B4b, 101; B6, 91; B8, 67; B10, 93; B11, 91; B12, 66, 91, 98, 100; B20, 96; B21, 67, 91, 93; B21a, 95-97. Anaximander: A1, 25; A11, 25; A26, 25; B1, 53. Anaximenes: A5, 54; A7, 54-55; A20, 55; B2, 54 n. 8. Aristophanes: Nub. 346 ff., 73 n. 33; 346-480, 58; 228, 205 n. 53; 1284, 205 n. 53. Aristotle: De cael. 303a11, 139; Fr. 208, 139; GA 769a17-19, 147; GC 314a15, 141 n. 27; 314a21-24, 161; 314a22, 139; 315b5-6, 75; 325b25-27, 166; HA 634a9, 176 n. 42; De interpr. 17a38-17b1, 127; Metaph. 984a27-984bl, 21; 985b421, 141 n. 26, 159; 985b23-31, 136 n. 14; 987a13-22, 136 n. 14; 987b14, 128; 1070a, 78; 1078b17-19, 128; 1078b17-21, 136 n. 16; 1078b172 In references to the Presocratics, numbers preceded by the letter A refer to testimonia, numbers preceded by the letter B, to fragments, as in Diels–Kranz 1952.

23, 136 n. 14; 1078b27-31, 128; 1086b1-5, 128; 1087a7-10, 131; PA 640b29-641a21, 154; 642a17 ff., 65; 642a17-24, 136 nn. 15-16; Phys. 191a30-31, 21; 192b, 78; 194a1821, 65; 194a20-21, 136 n. 15; 250a928, 86; 265b24, 171 n. 33; Sens. 4.441a3, 107. [Aristotle]: De Melisso, Xenophane, Gorgia 980a19-21, 185; 980a19b21, 184; 980a20, 185; 980b1, 185; 980b1-2, 196; 980b2-3, 185, 202; 980b3, 185; 980b3-8, 187, 189 and n. 19; 980b4, 185, 201; 980b5, 194; 980b6, 185, 202; 980b8-11, 190; 980b9, 192, 194, 198; 980b10, 192 n. 26; 980b11, 192 and n. 26, 194; 980b11-12, 191-192; 980b11-14, 190, 197; 9980b12, 191 n. 24, 198; 80b12-13, 192 n. 26; 980b13-14, 190 n. 23; 980b14, 191 and n. 24, 194; 980b14-17, 187, 191, 195, 197-198; 980b15, 194; 980b16, 194; 980b17, 194; 980b18-19, 190; 980b19, 194, 198. Cicero: Academica II.23.73, 157 n. 16. Democritus: DK67A1, 141; A6, 77, 141 n. 26; A9, 139; A15, 139; A21, 141; DK68A1(44), 141; A33, 141, 153; A34, 139 n. 22; A37, 139, 141 n. 28; A38, 140; A43, 167 n. 28; A47, 165 n. 21, 167 n. 29, 177; A49, 180; A56, 167 n. 28; A57, 75, 167 n. 26; A58, 171 n. 33; A101-102, 141 n. 28; A125, 167; A132, 141 n. 28; A135, 141 n. 28; A141, 147; B5d, 153; B6, 75; B8, 143, 150, 157; B9, 104, 180; B10, 143, 157; B32, 148; B116, 158; B117, 154; B118, 154; B124, 142, 147-151, 156, 158; B125, 104; B141, 75; B155, 150 n.

Indexes

5; B164, 156-157; B165, 143, 145, 152, 154-156; B302a, 150 n. 6. Diogenes Laertius: VIII.59, 118. Dissoi logoi (DK90): I.8, 129. Empedocles: A1, 118; A34, 115; A41, 121; A43, 114-115; A78, 107; A94, 107; B6-B9, 63; B8, 29, 115 n. 70; B8.6-7, 29; B9, 29, 115, 137 n. 18; B9.5, 105; B11, 28, 33-34; B12, 30; B12.1-2, 28; B12.2, 29; B12.3, 34; B15, 33-34; B17, 63, 106; B17.3, 85; B17.28, 87; B17.30, 106; B17.30-33, 29; B17.31, 29, 32; B17.31-33, 29; B17.34, 66, 106; B17.34-35, 116; B17.35, 105; B20, 137 n. 18; B21, 137 n. 18; B21.3, 107; B21.6, 107; B21.6-8, 124; B21.13, 66, 106; B21.13-14, 116; B21.14, 110; B22, 108; B22.4, 108, 110; B22.5, 107; B22.7, 110, 136 n. 17; B23, 88, 118, 137 n. 18; B23.1-8, 89; B23.4, 109; B23.5, 136 n. 17; B23.9-11, 88; B23.10, 66, 116; B26, 137 n. 18; B26.3, 66, 106; B26.3-4, 116; B27a, 121; B27.1, 116; B27.2, 65; B27.3, 116; B31, 114; B33, 112; B34, 112; B35, 64, 88; B35.7, 136; B35.8, 110; B35.15, 110; B35.15-16, 65; B35.16, 88; B35.16-17, 136; B35.17, 66, 88; B60, 123; B61, 123; B71, 88; B71.3, 110, 136 n. 17; B71.4, 65; B73, 112; B73.2, 136 n. 17; B75, 112; B75.2, 65; B76, 137 n. 18; B77-80, 137 n. 18; B83.4, 109; B87, 112; B88, 116; B90.1-2, 114; B91, 108; B96, 65, 87, 89, 113, 136; B96.3, 107; B96.4, 65-66, 107, 112; B98, 65, 107, 113, 136 and n. 17; B98.1, 136; B98.4, 136; B107, 107; B107.l, 65; B111, 118; B112.45, 118; B115.7, 136 n. 17; B117, 33,

233 137 n. 18; B122.2, 65, 107; B125.1, 136 n. 17; B127, 137 n.18; B128.6, 110; B128.8, 110; B129, 34; B130, 137 n. 18. Epicurus: Ad Her. 42-43, 140 n. 23. Galen: [Def. med.] 61-74, 149; 439, 142, 147; In Hipp. nat. hom. XV 32, 115. De meth. med. II.7, 157 n. 15; II.7.8-9, 131; Gorgias: B3, 201; B11(8), 205; B11(9), 205, 211; B11(10), 205; B11(13), 205-206; B11(14), 206. Gregory of Nazianzus: XXXVI 911 Migne. Heraclitus: B1, 25, 56; B2, 56; B8, 55; B10, 55; B30, 55; B31, 55; B49, 150 and n. 6; B51, 55; B54, 55; B67, 121; B79, 56; B114, 56. Homer: Od. 9.364ff., 39 n. 6. Leucippus: see s.v. Democritus DK67. Lucretius: I.150-214, 34; I.215-264, 34; II.478-531, 140 n. 23. Melissus: B1, 26; B2, 26; B3, 27; B6, 27; B7(1)-(2), 26; B7(1), 26; B7(2), 26; B7(3), 26, 68, 70, 87; B7(7), 27; B8, 68. MXG: see s.v. [Aristotle]. Michael of Ephesus: In De part. anim. 640b29, 154. Parmenides: A1 § 23, 60; A37, 60; A44, 60; A44a, 60; B1.30, 60; B2.6, 62; B2.7, 62; B4.2-3, 61, 68; Ǻ6.2, 43; B8, 163 n. 18; B8.5-49, 59; B8.5-21, 59 n. 16; B8.6-13, 22; B8.6-21, 163 n. 18; B8.7-10, 21; Ǻ8.10, 43; B8.14, 61; B8.19, 25 n. 11; B8.21, 25; B8.22-25, 59 n. 16, 163 n. 18; B8.26, 61; B8.26-31, 163 n. 18; B8.31, 59 n. 16, 61; B8.32, 62; B8.32-33, 59 n. 16, 163

234 n. 18; B8.33-61, 60; B8.36-37, 60; B8.37, 61; B8.42, 61-62; B8.42-49, 61, 163 n. 18; B8.43, 61; B8.49, 61; B8.52, 61; B8.53, 61; B8.54, 61; B8.60, 61; Ǻ9.4, 43; B12, 60; B12.3-6, 60; B14, 60; B15, 60. Philolaus: B1, 70; B2, 70-71; B3, 70; B4, 70; B5, 139; B6, 72, 102. Philoponus: In De gener. et corr., 19, 121. Pindar: Nem. vi. 1-7, 39; Pyth. viii. 95-96, 40. Plato: Ap. 18B, 205 n. 53; 23B, 205 n. 53; Meno 71B-73C, 129; 71D-E, 132; Parm. 131B, 191; 131B-C, 133; Phaed. 74A-C, 132; 97B ff., 97; Phaedr. 230A, 155, 157 n. 16; Resp. 479B, 43; Soph. 242D-E, 121; 246A-B, 49; 246A-C, 145; Theaet. 154A, 197. Prodicus: A9, 129; A11, 129; A13A19, 129. Pseudo-Galen: see s.v. Galen. Seneca: Epist. 7.10, 150 n. 6. Sextus Empiricus: Adv. math. VII.79, 201; VII.83, 185; VII.83-84, 189; VII.83-87, 184; VII.84, 185-186; VII.85, 186, 189 and n. 19, 206212; VII.86, 185, 189, 209; VII.90, 93; VII.264-67, 152, 157 n. 16; VIII.141-299, 209 n. 60; VIII.152153, 209; VIII.154, 209; VIII.155, 209; XI.245, 209 n. 60; Pyrrh. hypot. II.22-23, 152; II.2324, 155. Shakespeare: Antony and Cleopatra, Act IV, Scene 14, 58; Hamlet, Act III, Scene 2, 58 n. 15.

Alexander P. D. Mourelatos

Simplicius: In Phys. 28.15 ff., 140; 532-534, 171 n. 33; 1108.18, 86. Sophocles: El. 969, 40; 1129, 42; 1136-45, 42; 1150-51, 42; 1152, 42; 1154, 44; 1163-67, 40; 1164, 42; 1165, 42; 1165-66, 41; 1166, 40, 42; 1170, 40. Theophrastus: De Sens. 7-24, 108; 12, 122. Xenophanes: B18, 57; B23, 57; B28, 57; B34, 57; B38, 57. Zeno: A29, 86.

INDEX OF TOPICS DISCUSSED Abderites: 103. aien: 31 nn. 24-25. aisthanesthai: 194. aLVWKƝVLV: 194. All (the): 29. alloiousthai: 167 n. 26. aPHLSVLUK\VPLƝ: 77. Anachronism: 21, 151, 210. annihilation: 25-34. DQWKUǀSRV: 142-143, 146-148, 150151-153, 155. antitypia: 172. apatheia: 170. apeiron: 25, 27, 54, 70, 133. apo- compounds: 42 and n. 13, 44. apyston: 29 and n. 20. aUHWƝ: 129. Aspect/aspectual: see s.v. Verb aspect. DVSHWDSƝJƝQ: 88. Assimilation of “what is not” to “nothing”: 36. atoma (also atomoi, atomos, atomon): 75.

Indexes

Atomism (also Atomists): 51, 55, 69, 74-75, 78, 125, 139 n. 21, 140, 159, 163. Bald Man (the): 86. Behaviorism: 204. Boundary (also Bounds): 57, 61. Characterizing use of “nobody” or “nothing”: 38, 40-41, 43-44. Categorial argument: 186-187, 189190, 199-201. cKUƝ: 23. FKUHǀQHVWL(also chreon esti): 23-24. chreos: 23-24. chroia: 64, 88. Cloud-astrophysics: 57-58. Colors (mixing of): 89 n. 15, 93, 96, 118 n. 74. Communication: 190. Comparativism: 57 n. 14. Conjunctivity: 164. conturbatio: 95 n. 30. Creationism: 123. Cyclops: 39 n. 6. Degrees of existence: 41. Degrees of reality (Plato’s doctrine of): 43-44. Determinacy/indeterminacy: 62. diakosmein: 66. diakosmos: 60-61; megas diakosmos: 77; mikros diakosmos: 78. diamperes: 29 and n. 19. dLDWKLJƝ: 76, 159, 161 and n. 14, 177. Dissoi logoi: 129. DNA: 122. doxa: 60, 134. dynamis (also dynameis): 60, 84, 134. ekphora: 212. eidos: 52, 54, 64-65, 127, 136, 139.

235 Eleatic crisis: 58-59, 62, 72. Eleatic elenchus: 81-82, 84. Eleatics: 21, 21-22, 27, 34-35, 63-64, 67-68, 71, 73, 78, 80, 83, 86, 163. Eleatization of Xenophanes: see s.v. Xenophanes (Eleatization of). Elements: 165-166. Eliminative materialism: see s.v. Materialism. Emboitement: 100. Emergence: 84-85, 87, 94-95, 102, 104; quality e.: 86, 90, 92, 126; e. of structure: 97. enarthmios: 112. ENN: see s.v. ex nihilo nihil. ennoein: 194. epamphoterizein: 43. Epicureans: 140 n. 24. Epigenesis: 106. ereidein: 30, 31 n. 24. Essence: 65. Evening Star: 60. Evolution: 123. Existential quantifier: 36. Existential use of “nobody” and “nothing”: 37, 41. ex nihilo nihil (ENN): 21-35, 80, 85, 92. Fifth-century painting: 119-120. Finite: 57. Flux (Heraclitean doctrine of): 197. Folk riddle: 43-44. Formal Cause: 69. Form and Matter: 49, 59, 62, 70, 166 n. 24. Form as pattern-of-relations: 65. Form as resulting-structure: 65. Foretokening: 210.

236 Friends of Form: 49 Friends of Matter: 49. genesis: 26. Geometric proportion: 55. God: 57. Greek painting: 119. harmonia: 55, 65, 71-72, 74, 87, 89, 95, 102, 110, 120, 126. harmozein: 65, 95. Heap (Megarian puzzles of the): 86. hekaston: 157. hen DULWKPǀL: 130. KHQHLGHLORJǀLJHQHL: 130. heteroiousthai: 26. Homoeomerous/Non-Homoeomerous: 66-67, 101. homoion: 26, 192; homoion vs. tauto: 191, 195. homo-mensura: 153 n. 9. Hybrid category of thing-stuff-powerquality: 134-135. Hyperbole: 41. h\SRSWǀVis: see s.v. SURVSWǀVLV. idea: 52, 54, 64-65, 77, 89, 103, 127, 136, 140 n. 25, 144-145, 152, 154, 157, 162; atomos idea: 75. Illusionism: 117-119, 126. Impassivity: 170. Incommunicability (thesis of): 190191. Indeterminacy: see s.v. Determinacy/indeterminacy. Infinite: 57 and n. 13. Intelligibility (demand for): 73-74, 8283, 97. Intrinsic Properties of Democritean Atoms: 164. Ionians: 49-50.

Alexander P. D. Mourelatos

istion,“sail”: see next. istion, “awning”: 133, 191-192. Journey: 62. kath’ hekaston (to/ta): 127. katholou (to): 127, 136, 146. kenon: 75. koinon (to): kolla (also kollDǀand NROOƝVLV): 112. kosmos: 52, 55, 60, 67-68, 70-71, 87, 123-124, 140 n. 25, 141, 145, 169; distinction between kosmos and universe (to pan): 145 n. 32. krasis: 95, 108. Latency (principle of): 98. Like to Like principle: 140 n. 25. Limiting things: 70-71, 78, 101-102. Limit(s): 57, 61-62. Linguistic solipsism: 203. logos: 56, 183-186, 198, 205-206, 209. Love: 63-65, 74, 76, 104, 116, 118, 123, 125. Marvel: 66. Materialism: 103; eliminative m.: 104, 126, 180. Material Monists: 55, 57. mathematici: 50, 72. Matter and Form: see s.v. Form and Matter. Meaning – behavioral conception of m.: 184, 204, 206; empiricist theory of m.: 188 and n. 15; mentalist or ideational conception of m.: 184, 199; referential conception of m.: 184, 199-200; stimulus-response conception of m.: 206, 210-211. mƝGHLV: 38-39, 43-44. PƝGHQ: 21 n. 3, 26-27, 36, 38-39, 41, 43-44. mƝHRQ: 21 n. 3, 36.

Indexes

Melissan rejoinder: 69, 71-72, 75. mƝQ\WLNRQ: 207, 210. mHWDNRVPƝVLV: 51, 58, 68, 123; metaNRVPƝWKƝQDL: 87. mHWDUUK\VPHǀ: 77. metaballein: 167 n. 26. mHWHǀUD: 57. Metonymy: 41, 43. metron: 55. Milesians: 56. Millet Seed (the): 86. Mind: 66-67, 74, 76, 98. mixis: 95 and n. 30. Mixture: 107, 110. Monism: 27. Morning Star: 60. mRUSKƝ: 52, 54, 61, 64, 139, 153. Motility: 76, 164, 171. Naïve Metaphysics of Things: 52, 56. Natural piety: 90. Neo-Ionians: 34, 81-83, 97, 106. Nobody: 37. noein: 194. Nominalism: 144 n. 30. nomos: 200. Non-being: 36. Not-being: 36. Nothing: 37. nous: 66, 97-99. One (the): 60, 64, 68; the Eleatic One: 105. One-in-many (metaphysical puzzle): 191. Onomatopoeia: 208, 212. Opposites: 111. Optical fusion: 118 n. 74, 120. oudama: 26.

237 oudeis: 38-39, 43. ouden: 21 n. 3, 26-27, 36, 38-39, 43. ouk eon: 21 n. 3. ou mallon: 140 and n. 25. outidanos: 39 and n. 6. outis: 39 and n. 6. Oxymoron: 44. palmos: 165 n. 21, 177. pan (to): 64, 145 n. 32; distinction between kosmos and universe (to pan): see s.v. kosmos. Pangenesis: 147. Paradox of Plato’s Meno: 189. Parmenidean requirements (or criteria): 163 n. 18, 167-168, 177-178. Particular: 157. Passages: 108. peirata: 61. pHUDLQRQWǀQ: 70. Perception – intersubjective p.: 196, 203; intra-subjective p.: 198. phtheiresthai: 29. physiologoi: 129. physis: 45, 70, 78, 115 n. 70. Piety (natural): see s.v. Natural piety. pLOƝVLV: 55. Platonic Form: 133. Platonists: 132. pOƝJƝ: 76-77. Plenitude (principle of): 145. Pluralists: 84, 86-87, 90, 125. Polyphemus: 39 n. 6. poroi: 108-111, 113-114, 117-118, 120, 122-125, 186, 196. Positivists: 188, 203. Prefiguration: 99, 101-103, 122. Preformation: 67, 97-98, 102. Principle of Plenitude: see s.v. Pleni-

238 tude (principle of). Principle of Sufficient Reason (PSR): 23-24, 31-32, 140, 169. Privative a- compounds: 42 and n. 13, 44 and n. 18. SURVSWǀVLV(or K\SRSWǀVLV): 212. PSR: see s.v. Principle of Sufficient Reason. Pyrrhonists (also Pyrrhonian skepticism): 121, 209. Pythagoreans: 49-50, 72, 102, 126. Quality: 101. Quietude: 121. Reduction: 103-104, 108. Relation (concept of): 69. Relational Properties of Democritean Atoms: 164, 169. Reverential awe: 67. rhysmos: 77-78, 159, 161-162. Sail Argument: see s.v. istion. Samos: 28. schƝma: 144, 153, 159, 161 n. 11. Secondary qualities: 103. Seeds: 99-100. VƝPDLQHLQ: 139, 209. Signs (“indicative”, endeiktikon, and “commemorative”, K\SRPQƝVWLNRQ): 209-210. Similar (substituting it for “numerically the same”): 193, 195. Simile: 61. Slippery Slope argument: 211. Sophists: 63, 129. spermata (also sperma): 67, 99. sphairos: 64-65, 104-105, 107, 114, 116, 120-121, 123, 125. Sphere: 64. St. Elmo’s fire: 57.

Alexander P. D. Mourelatos

stoicheion: 144. Stoics: 209. Strife: 63-64, 76, 104, 123. Structure: 97-98, 101, 122, 124. Submergence: 85. Substitute stimuli: 204. Supervenient entities: 85, 102, 105-106. Sweet: 57. syllapsies: 55. symmetria: 109. Symmetry: 59. Synonymy Principle of Causation: 134. taxis: 52-53, 76, 141, 144, 159 and n. 2, 161. temperatio: 95 n. 30. thauma: 66; thauma idesthai: 88. tKƝVHVWKDL: 30 and n. 21. thesis: 76, 141, 144, 159 and n. 2, 161, 170. tKHVSHVLƝWKHQ: 87, 89, 118. Token: see s.v. Types-tokens. Transmission Model of Causality: 123124. tURSƝ: 76, 159, 161, 170, 177. Types-tokens: 131-132, 135, 140-141, 157 n. 15, 166 n. 24. UM: see s.v. Universal mixture. Underworld: 41, 43. universalia: 144. Universal (concept of/criteria): 127, 129, 145-146. Universal Mixture (UM): 91-92, 94, 96, 100, 109, 126. Unlimited (also Unlimited things): 62, 70-71, 101-102. Unmoved Mover: 78. Veil of appearances: 95. Verb aspect: 29 n. 19, 31 n 24.

Indexes

Vienna Circle: 203. Vital use of “to be”: 40 n. 9. Void: 27.

239 Words (as substitute responses): 204. Xenophanes (Eleatization of): 56.