Department of The Classics, Harvard University

Department of the Classics, Harvard University
Studies in Xenophanes
Author(s): Aryeh Finkelberg
Source: Harvard Studies in Classical Philology, Vol. 93 (1990), pp. 103-167
Published by: Department of the Classics, Harvard University
Stable URL: http://www.jstor.org/stable/311284 .
Accessed: 26/12/2014 05:58
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
Department of the Classics, Harvard University is collaborating with JSTOR to digitize, preserve and extend
access to Harvard Studies in Classical Philology.
http://www.jstor.org
This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

All use subject to JSTOR Terms and Conditions
STUDIES IN XENOPHANES
ARYEH FINKELBERG
T seemsconvenientto begin our discussion of Xenophanes' teach-

ing with Aristotle's account at Metaph. 986b10-987al. Here Aristo-
tle classes Xenophanes with Parmenidesand Melissus under the com-
mon denomination of "partisansof the One," ivi*ovtE;, and credits
them with holding in common the doctrine that the universe is one
unchangeableentity, a position which, in Aristotle's view, disqualifies
them as collaboratorsin the discovery of causes.' Nevertheless, after
dismissing the "partisansof the One" collectively, he proceeds to deal
with each of them severally, specifying their individual positions. In
doing so, he makes the principal distinction between Parmenides' and
Melissus' notions of the One: the former posited unity in definition
-rv the latter, in matter (Icarax ilV jXrlqv).Without
(•arax into•6yov),
entering a long discussion, which would be inappropriatehere, of
Aristotle's notion of materialand formal unity as such, we may observe
that the criteria underlying this distinction are also operative in
Aristotle's judgment on Xenophanes' concept of God, so that the
definition of these criteria is prerequisiteto understandinghis account
of Xenophanes. What, then, are Aristotle's grounds for attributingthe
formal unity to Parmenidesand material unity to Melissus? One rea-
son, generally recognized by commentators, is the difference in the
spatialcharacteristicsof the One (which Aristotleregardsas a corollary
(8t6) of the distinction between what is one in logos and in matter)-
finite in Parmenidesand infinite in Melissus. Anotherand more impor-
tant indication (for the difference in the spatial characteristicsis noted
only parentheticallyby Aristotle, in a passing remark)2has escaped the
' Cf. Ph. 184b25.
2 Cf. W. D. Ross, Aristotle's
Metaphysics (Oxford 1924, repr. 1958) 1.154, ad
986b22; see also J. B. McDiarmid,"Theophrastuson the PresocraticCauses," HSCP 61
(1953) 117.

104 AryehFinkelberg
notice of many critics, though it is explicit in Aristotle. I mean the

relationship between the One and the Many: ".. . But Parmenides
seems in places (rtox) [sc. in the Doxa part of his poem] to speak with
more insight. For, claiming that, besides the existent, nothing non-
existent exists, he thinks that of necessity only one thing exists, viz. the
existent and nothing else ..., but being forced to follow the observed
facts, and supposingthe existence of that which is one in definition,but
more than one according to our sensations, he now posits two
(nrtatv)
causes and two principles ... fire and earth ... ."3 Parmenides' concept
of Being as he posits it in the Aletheia does not in principlediffer from
Melissus';4yet as far as the doctrineParmenidesadvances in the Doxa
is considered, his One, appearingto allow for sensible diversity, turns
out to be unity in definition. As Aristotle puts it elsewhere: "Of those
who said the universe was one, then, none succeeded in discovering a
cause of this sort [sc. efficient cause], except perhapsParmenides,and
he only inasmuch as (icartx 'roaooo ov 6oov) he supposes that there is
not only One but also in some sense (irow)two causes."5Parmenides'
transitionfrom the One to "two causes and two principles" and their
exact relation to the One is not altogether clear to Aristotle; yet
"inasmuchas" Parmenides"in places" admits thatbeside the One there
is somehow also plurality"accordingto our sensations,"his notion of
Being may be regarded as unity in definition.6Thus its compatibility
with the sensible manifold, together with its finitude (and more pri-
marily than the latter) is, in Aristotle's view, indicative of the formal
characterof the unity postulated.
3 Metaph. 986b27 (here and elsewhere Aristotle is quoted in the Oxford translation
with occasional minor changes); cf. Cael. 298b15. Renderingntot as 'in places,' 'else-
where' seems to me preferableto the reading 'with somewhat more insight': the former
construalseems to be suggested both by the fact that Aristotlerefershere to the Doxa and
by t•adtvat 986b33.
4 Cf. Aristotle's discussion in Ph. 184b15-187a10.
5 Metaph.984b2.
6 Cf. H. Bonitz, Aristotelis Metaphysica. Commentarius (Bonn 1849, repr. Hil-
desheim 1960) 72, 84; A. E. Taylor, Aristotle on His Predecessors (repr., Chicago-
London 1949) 49; Ross (above, n. 2) 1.153, ad 986b19; K. Deichgraiber,"Xenophanes
nrepi(poeco;," RhM 87 (1938) 14. H. Cherniss,Aristotle's Criticism of Presocratic Phi-
losophy (Baltimore 1935) 220 n. 15, is therefore wrong in assuming that Aristotle
deduces the formal characterof Parmenides'and the materialcharacterof Melissus' One
only from the finitudeof the formerand the infinity of the latter.

StudiesinXenophanes 105
Turning now to Aristotle's account of Xenophanes, we should note

two things at the outset. First, the appearanceof Xenophanes' name in
our passage is by no means casual: his inclusion among the "partisans
of the One" is preparedfrom the beginning. In stating the general doc-
trine of those who "spoke of the universe as if it were one entity"Aris-
totle, before referringto each of them by name, draws the preliminary
distinctions: ".. . they are not all alike in the excellence of their state-
ment or in its conformity to the facts of nature (tp6orov Se o*. trv
&Yrbv rtvtESm oitZE to- Kic• O; OE Tol'CXar" ailV iptaiv)." In what
follows, the latter distinction appearsto be between Parmenides(obvi-
ously, on approachinghis teaching as admittingthat "there is not only
One but also in some sense two causes") and Melissus, while the
"excellence in the statement"seems to be the distinction between the
Eleatics and Xenophanes who is said to have given "no clear state-
ment." Secondly, in classing Xenophanes with the Eleatics Aristotle
does not, as it is often contended,7merely follow the traditionwhich
associated Xenophanes with Parmenides. Aristotle's remark "...
Xenophanes, the first of these partisansof the One (for Parmenidesis
said8to have been his pupil) .. ." shows that what Aristotle infers from
the tradition, whatever its origin may be, is not the nature of Xeno-
phanes' concept but only his historicalpriorityover Parmenidesin put-
ting forwardthis specific form of monism. Moreover, Aristotle clearly
distinguishes the typological affinity he observes between the Eleatic
and Xenophanean notions of the One from the presumed historical
influence of Xenophanes on Parmenides. In contrastto Plato who was
7 Notably by Kirk in G. S. Kirk, J. E. Raven, M. Schofield, The Presocratic Philoso-

phers2 (Cambridge1983) 165.
8 Aristotle's yyerathas been much pressed to prove his alleged caution and even
doubt, see, e.g., Ross (above, n. 2) 1.154, ad 986b22; W. K. C. Guthrie,A History of
Greek Philosophy (Cambridge 1962) 1.369; D. Babut, "Sur la 'thdologie' de X6no-
phane,"RPhilos 99 (1974) 435; cf. W. Jaeger, The Theology of the Early Greek Philoso-
phers (Oxford 1947) 54 n. 65, and others. Yet yyeratis regularlyused in referencesto
an established,known fact, see F. Uberweg, Grundrissder Geschichteder Philosophie, 1.
Die Philosophie des Altertums12,K. Praechtered. (Berlin 1926) 81; A. Lumpe, Die Phi-
losophie des Xenophanes von Kolophon (diss. Munich 1952) 22. M. Stokes, One and
Many in Presocratic Philosophy (Washington 1971) 84 n. 55, replies that "it need not be
so used here,"but regrettablydoes not explain why in this particularcase the word need
not be used as it is regularlyelsewhere.

106 AryehFinkelberg
ready to allow Xenophanes as one of the "Eleatic tribe,"9Aristotle

coins the word ivitovtE; to cover both Xenophanes and the Eleatics,
without, however, implying the unity of 'school'; if this precautionhas
proved inadequate to save Aristotle from misinterpretationby some
modem scholars, the blame is not his.10 To conclude, Aristotle deli-
berately groups Xenophanes with Parmenides and Melissus, and in
doing so he proceeds from the nature of Xenophanes' concept rather
thanfrom historicalor otherconsiderations.
From Aristotle's counting Xenophanes with the "partisansof the
One" it follows that he considers the XenophaneanGod to be the entire
universe envisaged as one unchangeableentity. Now, as in the case of
the Eleatics, Aristotleproceeds to specify Xenophanes' individualposi-
tion: he states regardingXenophanes what he has already said regard-
ing the "partisansof the One" collectively, namely, that Xenophanes'
One is identical with the whole of existence, "with his sight on the
whole of the world he says that the One is, viz. God,"" and notes that
Xenophanes "gave no clear statement." This remarkhas been taken to
the effect that Aristotle "obviously could not understandwhat Xeno-
phanes meant by his one motionless god."12Yet this is a conclusion
hardly to be inferred from Aristotle's text. What Aristotle plainly
means in saying that Xenophanes"gave no clear statement,nor does he
seem to have grasped the nature of these [sc. formal and material
causes]" is that, having declared the universe to be One God, Xeno-
phanes did not specify, in one way or another, whether the unity he
postulatedwas formal, like that of Parmenides,or material,like that of
Melissus, and moreover, he seems to be altogether unaware of this
difference.13
9 Sophist 242D (= H. Diels, Die Fragmente der Vorsokratiker,5W. Kranzed. [Berlin

1934] 21 A 29).
10Thus, e.g., J. Burnet, Early Greek Philosophy4 (London 1930) 126, says that "the
context shows he [Aristotle]means to suggest he [Xenophanes]was the firstof the Eleat-
ics"; similarlyMcDiarmid(above, n. 2) 119.
1 Cf. Guthrie(above, n. 8) 1. 369, 380-381 n. 1;
alternatively:"thatthe One is God."
I do not see why Kirk (above, n. 7) 172, declares this phraseto be "cryptic,"the more so
in that he himself does not doubt that Aristotle's wording "clearly implies" (my italics)
that he considersXenophanes'God to be identicalwith the world.
12Kirk(above, n. 7) 171; cf. Stokes (above, n. 8) 83.
13E. Zeller, Die Philosophie der Griechen in ihrer geschichtlichen Entwicklung6
(Leipzig 1919, repr. Hildesheim, 1963) 1.631 (followed, among others, by Uberweg-
Praechter(above, n. 8) 74; Burnet (above, n. 10) 124; cf. H. Diels, Doxographi Graeci4
[Berlin 1965] 110 and n. 1) is wrong in interpretingAristotle's remark about Xeno-

Studies in Xenophanes 107
Now why could Aristotle not determine which kind of unity was
posited by Xenophanes? To answer this question we should recall that
in forming his judgment concerning the respective natures of Par-
menides' and Melissus' One, Aristotle was relying on two criteria,the
spatial characteristicsof the One and its compatibilitywith "more than
one accordingto our sensations." The lack of clear statementon Xeno-
phanes' part must therefore mean the lack of unambiguousdetermina-
tion of God in precisely these two respects. This is to say that in Xeno-
phanes Aristotle found neitherdirect statementsnor indirectindications
as to whether he assumed that God, though one entity, nevertheless
allowed internaldiversity and whetherhe conceived of God as finite or
infinite.14The former obscurity is surprising. Melissus consistently
denied the sensible manifold, but Parmenides"in places" spoke "with
more insight,"that is, supplementedhis doctrineof the One with that of
"two causes and two principles";on Aristotle's account, Xenophanes
seems to have done neither. Aristotle, however, knew that, beside the
doctrine of the One, Xenophanes, like Parmenides, advanced cosmo-
gonical and cosmological doctrines one of which, namely, that earth
stretches indefinitely dawnwards, he discusses in De Caelo;15 yet
strangely enough, Aristotle does not take Xenophanes' cosmology as
an indicationof the formal characterof his One, as he does in the case
of Parmenides. Below, in examining Theophrastus'account of Xeno-
phanes, Aristotle's reasons will become clear; here we must content
ourselves with the mere statementof this fact.
phanes' lack of clarity as referringto the lack of specificationof God's spatialcharacteris-

tics. See the criticism of this exegesis in: Cherniss (above, n. 6) 201 n. 228; McDiarmid
(above, n. 2) 117; P. Steinmetz, "Xenophanesstudien,"RhM 109 (1966) 47; Guthrie
(above, n. 8) 1.369 and n. 2; cf. Bonitz (above, n. 6) 84 ("... naturameius unitatis non
descripsit, nihil potest a hanc de generibus causarumquestionem conferre.")and Ross
(above, n. 2) 1.154, ad 986b22.
14Cf. Jaeger (above, n. 8) 43 n. 23, and esp. 53. K. von Fritz, "Xenophanes,"RE, ser.
2, 9.2 (1967) 1555, is right in that Aristotle says that Xenophanesdid not clearly indicate
whetherthe unity he postulatedwas formal or material,not that he did not clearly express
whetherGod was finite or infinite (cf. above, n. 13). Yet this is not to say, as von Fritz
does, that Aristotle's account provides no evidence on the latterpoint: the spatial charac-
teristics are taken by him as indicative of the kind of unity, and his uncertaintyas to
whether Xenophanes' God is a formal or material One is direct evidence that Xeno-
phanes, inter alia, did not determinewhetherhis God was finite or infinite.
15 294a21 (= Diels-Kranz [above, n. 9] 21 A 47).

108 AryehFinkelberg
Thus, not finding in the "theological"context evidence for whether

Xenophanes allowed the inner diversity of his One, and, for reasons we
have not yet specified, avoiding use of Xenophanes' cosmological doc-
trine as such evidence, and in the absence of indicationswhetherXeno-
phanes conceived of God as finite or infinite (and other possible indica-
tions which he perhapswould have been ready to accept as such), Aris-
totle preferredto count God with neitherkind of unity, rathersuggest-
ing that Xenophanes was unaware of the difference. It is because of
this "naivete"that he finally dismisses Xenophanes together with Mel-
issus, whose "naivet6," however, is apparentlyof another sort,-the
crude concept of the One as a material unity, which is incompatible
with "the facts of nature."16
To sum up, Aristotle maintained that Xenophanes' God was the
entire universe envisaged as one unchangeableentity, yet the precise
sense of God's oneness and unchangeabilitywas not clear to him.17In
fact, he found no way of determiningwhether the unity postulatedby
Xenophaneswas that in definition(and thus allowed inner diversity and
hence motion) or that in matter (in which case plurality and motion
were totally excluded). In these circumstances, Aristotle preferredto
leave the issue undecidedand blame Xenophaneshimself.18
16 Cf. Bonitz (above, n. 6) 84, and Ross
(above, n. 2) 1.152, ad 986b27. I do not think,
as Ross does, that Aristotle's criticism of Melissus' argumentation(Soph. El. 167b12) is
of relevancehere.
171Ican find no supportin Aristotle for Steinmetz' conclusion (above, n. 13) 48, that, in
Aristotle's opinion, "1. Die AuBerungendes Xenophanes zielen auf Gott, sind also
eigentlich Theologie. 2. In dieser Theologie steckt ein ontologischerKern ..." Nowhere
does Aristotle regard Xenophanes' concept of God as theological; on the contrary,he
clearly regardsit as ontological, as his classing of Xenophanes with the Eleatics proves.
J. Barnes, The Presocratic Philosophers2(London 1982) 84 n. 6, thinks that in doing so
Aristotle is misled by Plato (see above, n. 9); yet Timon's fr. 59 (= Diels-Kranz [above,
n. 9] 21 A 35) speaks in favor of both Plato's and Aristotle's views. The very question
whetherXenophanes' concept was theological or ontological presupposesan anachronis-
tic distinction and results from underestimatingthe "theological" purportof the early
speculation(on which see my "On the Unity of Orphicand Milesian Thought,"HThR79
[1986] 321-335 and "The Milesian Monistic Doctrine and the Development of Preso-
cratic Thought,"Hermes 117 [1989] 257-270), as well as from the misinterpretationof
Theophrastus'words quoted in Simplicius (on which see below in the text).
18McDiarmid (above, n. 2) 117, is correct in pointing out that ".. . Aristotle ... is
admitting his own lack of evidence for classing Xenophanes with either Parmenidesor
Melissus" but comes to the surprisingconclusion that "Aristotleconcludes that whatever
Xenophanesmeant, it was something differentfrom the doctrinesof Parmenidesand Me-
lissus." This idea seems to arise from the strangeclassificationfound in Cherniss(above,

Aristotle's attributionof unchangeability to God can be checked

against Xenophanes' authenticlines. The firstrelevantpiece isfr. 26:
Always he remainsin the same [place], not moving at all,

nor it is fittingfor him to go now here and now there.
It has been suggested that dvqonagis to be taken here in its "philo-

sophic" sense, i.e., as covering the whole field of change.19Quite apart
from the question in which respect Xenophanes' God is actually
unmoved, the assumptionas such is altogetherwrong: it is anachronis-
tic to read into Xenophanes' tcavo•gEvo;oi~Ev the much later techni-
cal sense of ytvrqotg.Moreover, the context leaves no doubt that what
Xenophanes means is locomotion.20This, however, does not imply, as
some scholars contend, that locomotion is the only kind of movement
Xenophanes denies his God.21Let us consider anotherrelevant line, fr.
25:
But without toil he shakes all things by the will of his mind.22
The fact that God's accomplishmentsare effortless entails his absolute

repose, which amounts to the denial of both locomotion and distur-
n. 6) 220 n. 15: "Xenophanesdoes not say anything which can show in which way he
meant the unity he championed to be understood [viz. whether it is formal or material],
and Aristotle decides that it was with a view to the whole universe that he said that the
One is God." Does Cherniss mean that the unity declared "with a view to the whole
universe"need not be either formal or material?
19Notably by H. Frankel, Dichtung und Philosophie des friihen Griechentums(New
York 1951) 428.
20 Cf. von Fritz (above, n. 14) 1546-1547; Guthrie(above, n. 8) 1. 382; Babut (above,
n. 8) 430, and some others.
21 See Guthrie (above, n. 8) 1. 382 (cf. F. M. Cornford,Principium Sapientiae [Cam-
bridge 1952] 147), who asserts that &civrlroqin Metaph. 986b14 must mean "ungen-
erated." But does not Aristotle say: ". .. some at least of those who maintainit to be one
... say the one and natureas a whole is unchangeablenot only in respect of generation
and destrucilton(for this is a primitive belief, and all agreed in it), but also of all other
change; and this view is peculiarto them"(Metaph.984a29)? This equally applies to the
suggestion of F. M. Cleve, The Giants of the Pre-Socratic Greek Philosophy (The Hague
1965) 24, that God is unchangeableonly in respect of his size.
22On v6oi cppEvisee: K. von Fritz, "NOYX,NOEIN and Their Derivatives in Preso-
cratic Philosophy 'excluding Anaxagoras)," CP 40 (1945) 229-230; see also S. M.
Darcus,"The Phren of the Noos in Xenophanes' God,"SO 53 (1978) 25-39.

110 AryehFinkelberg
bance. Xenophanes' God thus always remainsin the same place and in
the same state, which is tantamountto freedom from all kinds of move-
ment includingchange.23
God's identity with the entire universe is not only implied by group-
ing God with Eleatic Being but is also explicitly and unambiguously
pointed out by Aristotle: "with his sight on the whole of the world he
says that the One is, viz. God." Still, some critics challenge this
report.24How, we are asked, can God, if unmoved and indentical with
the world, be said, as infr. 25, to move all things? Does this not rather
suggest that unchangeableGod is distinct from the changeable inivta,
the totality of things, which he sets in motion?25To examine the vali-
dity of this objection we must consider the question from three points
of view, doxographic,logical, and historical.
In equating Xenophanes' God with the whole of existence Aristotle
concurs with Plato who, relating Xenophanes to the "Eleatic tribe,"
specifies their-and hence also Xenophanes'-teaching as the unity of
"what we call all things."26Theophrastus,as we shall see later, and the
Sceptic Timon who, being an admirer and imitator of Xenophanes'
poetry, was necessarily well acquaintedwith it,27adheredto the same
23 Xenophanes' criticism of the
popularbelief that the gods are born (fr. 14) implies
that God is also ungeneratedand a fortiori everlasting (cf. Xenophanes' apophthegms
collected in Diels-Kranz [above, n. 9] under 21 A 12). Babut (above, n. 8) 434, thinks
that "l'immobilit6de Dieu n'est en d6finitiveque la cons6quenceimm&diatede son apti-
tude A r6alisersa volont6 Adistance." Yet God's immobility is but a particularaspect of
his unchangeability(which Babut is not ready to accept in Xenophanes),the explanation
of which should be looked for in the ontological purportof the concept of God.
24On various grounds, see K. Reinhardt, Parmenides und die Geschichte der
griechischen Philosophie (Bonn 1916) 116, 122, 125, 152; Cherniss (above, n. 6) 201
n. 228; O. Gigon, Der Ursprung der griechischen Philosophie (Basel 1945) 184; Jaeger
(above, n. 8) 43; Lumpe (above, n. 8) 22-26; J. Kerschensteiner,Kosmos (Miinchen
1962) 90-93; Kirk,Raven, Schofield (above, n. 7) 171-172.
25Frainkel(above, n. 19) 428, 431-432; cf. Lumpe (above, n. 8) 23; Kirk, Raven,
Schofield (above, n. 7) 171-172.
26Sophist 242D (Diels-Kranz[above, n. 9] 21 A 29). Stokes (above, n. 8) 84, says that
"Aristotleis clearly ... EleaticizingXenophanes. But his involuntaryEleaticizationdoes
not proceed so far as to credit Xenophanes explicitly with the view ... that all things
(meaning the world) are one." Does Stokes mean that Aristotle did not count Xeno-
phanes with the "partisansof the One" or that he did not say that the "partisansof the
One" spoke of the universe"as if it were one entity"?
27Fr. 59. On Timon as an independentsource on Xenophanes see: Steinmetz (above,
n. 13) 35-37; cf. Barnes(above, n. 17) 98.

view of God. The consensus is indeed impressive: including all our

best and most ancient authorities, it comprises the most diverse
sources, ranging from Plato to the Peripatetics and the Sceptics. The
doctrine is thus one of the best attested Presocratic conceptions, and
anyone who seeks to dismiss it must be preparedto face this fact. Yet,
to my knowledge, no really convincing explanationof it has been ever
proposed.28Now the Xenophanean line known to us as fr. 25 was
known to the ancients as well, yet it did not prevent them from equat-
ing God with the world. Must we thereforeconclude thatPlato, Aristo-
tle, Theophrastus,and Timon totally lost their acumen in their stubborn
determinationto impose on Xenophanes, by their joint efforts, an idea
he never even dreamed of? Or would it be wiser to admit that they
knew what they were talking about and that, had modern critics known
the full context of this saying, they would never have arguedin the way
they do?
But is God's immobility indeed incompatiblewith his being identi-
28The often repeated claim that Aristotle
misrepresents Xenophanes because he
wrongly regards him as a "primitiveEleatic" is the bold petitio principii. Besides, why
should he do this? With slight variations,the popularexplanationruns along the follow-
ing lines: in relatingXenophanesto the "Eleatictribe"Plato was possibly joking; Aristo-
tle probablytook this remarktoo seriously, while Theophrastus,of course, only repeated
his master's opinion. As simply as that,-see: Burnet (above, n. 10) 127; Cherniss
(above, n. 6) 353, cf. 201 n. 228; Jaeger (above, n. 8) 53-54 and esp. n. 65; Fr'inkel
(above, n. 19) 432 n. 13; McDiarmid (above, n. 2) 119-120 (approved by H. Schwabl,
"Der Forschungsbericht. Die Eleaten," Anzeigerf. d. Altertumswissenschaft10 [1957]
211); Kirk, Raven, Schofield (above, n. 7) 165; cf. Ross (above, n. 2) 1.154, ad 986b22;
the list can easily be lengthened. But what of the possibility that one of the actors in this
comedy of doxographicerrorsmisreadhis part,-for example, Plato was serious, Aristo-
tle was not a credulous simpleton, or Theophrastus disappointingly consulted Xeno-
phanes' verses? (For a balanced view of Theophrastusas a historical source on Pre-
socratic philosophy see: C. Kahn, Anaximanderand the Origins of Greek Cosmology
[New York 1960] 17-24.) To demonstratethat Plato's remarkwas not intendedas a seri-
ous historicaljudgmentit is usual to point out that to the statementthat the "Eleatictribe"
starts from Xenophanes he adds the phrase "and even before" which most likely refers
first and foremost to the Milesians (cf. Jaeger [above, n. 8] 54 n. 65). The presumption
thus seems to be that anyone who speaks of Eleatic monism as evolving from Milesian
monism must be joking. Apart from the intrinsic merits of this presumption,it may be
noted that Aristotle says precisely that at Metaph. 984a27 and he is not joking. But even
supposing that Plato was playful, while the Peripatetics,totally lacking in humor, were
ready to turnall Plato's joking into historicalfact, what aboutTimon's evidence? Was he
also guided by Plato's remarks? Or perhapshe was reading Xenophanes' verses in an
edition with a Peripateticcommentary?

112 AryehFinkelberg
cal with the world? Obviously not, provided that God is conceived as
an intelligible essence unifying the sensible manifold,-in Aristotle's
terms, the One in definition. Now it seems thatfr. 25 suggests just this
idea of unity. From the more general point of view, this understanding
of God is requiredby the fact that, beside the concept of the unchange-
able One, Xenophanes posited a cosmological doctrine which in its
very essence assumes plurality and change. If in trying to understand
Xenophanes' position we wish to spare ourselves the difficulties which
generationsof Parmenideanscholars have experienced in adopting the
formal approach to Parmenides' system, with the inevitable result of
viewing his mind as split between the real and the fantasy world,29we
will have to grantXenophanes a more or less unified outlook, that is, to
assume the "formal"characterof his One. I am aware of the objection
that logically necessary is not the same as historically true or even
plausible, and I shall thereforeturn now to the historical aspect of the
issue.
Do we indeed possess historical evidence that allows us to credit
Xenophanes with such a sophisticatedidea of unity? Fortunately,we
do. The idea in fact antedatesXenophanes, being attested, at least in a
rudimentaryform, in Anaximander. I mean Diogenes' report that,
according to Anaximander, "the parts change, while the whole is
unchangeable."30"The whole" referredto here can hardly be anything
other than the Apeiron which is therefore said to remain unchangeable
notwithstandingthe changes of its "parts,"i.e., the various components
of the developed world. If my interpretationis correct, the Apeiron
appearshere as a selfsame intelligible essence unifying the changeable
manifold.31I do not intend to claim that this concept of the Apeiron is
predominantin Anaximander;all I wish to show is that Anaximander
29 "A doctrine
which, taken literally, might seem to be either madness or sheer sophis-
try" (C. Kahn, "The Thesis of Parmenides,"Review of Metaphysics 22 [1969] 715); "As
to whetherParmenideshimself accepted these conclusions ... we can merely guess ...
perhapshe believed it all the time, and was mad"(M. Furth,"Elementsof Eleatic Ontol-
ogy," in A. P. D. Mourelatos ed., The Pre-Socratics [New York 1974] 296). There is
something decidedly wrong with the interpretationof a philosophic doctrinethat requires
one to assume the insanity of its author.
30 Diog. 2. 1 (= Diels-Kranz[above, n. 9] 12 A 1).
31It is not impossible that it was Anaximander,and precisely for this reason, whom
Plato primarily had in mind when he said that the "Eleatic tribe" started even before
Xenophanes(cf. above, n. 28).

occasionally touches upon the idea of the intelligible unity of the mani-
fold. But occasional as it may be, the idea is not incidental: it arose, as
I argue elsewhere, in response to the inadequacy of the material form
of monism.32Consequently, if truly understood, the idea had to pro-
duce a profoundeffect on the monistic thinkeraware, as alreadyAnax-
imanderevidently was, of the failure of material monism to provide a
consistently monistic picture. Now who is a better candidate for the
role of such a thinker than Ionian-born and Ionian-educated Xeno-
phanes who in his youth, as Theophrastus tells us, was personally
acquaintedwith Anaximander33and who, as Aristotle puts it, was the
first of the "partisansof the One," that is, the champions of the new
kind of monism?34
I conclude that the Xenophaneanconcept of God as the single and
unchangeable, intelligible essence unifying the manifold, an essence
endowed with divine powers and causing and controlling all that goes
on in the world, must be the development of one of the facets of the
Anaximandreannotion of the Apeiron, the divine substanceunderlying
the entire universe and governing it.35
32"The Milesian Monistic Doctrine" (above, n. 17). The

insuperable difficulty of
materialmonism is the logical impossibility to reduce the sensible manifold to one of its
members or, to put it otherwise, both to maintain the existence of the manifold and to
claim it to be a self-identical definite body.
33Diog. 9. 21 (= Theophr.fr. 6a = Diel-Kranz [above, n. 9] 21 A 2); cf. Diels-Kranz
(above, n. 9) 28 A 1. See also von Fritz (above, n. 14) 1543. McDiarmid(above, n. 2)
119, declares that "... to complete the Eleatic line of descent Theophrastusadds that
Xenophanes was the pupil of Anaximander-apparently on no other grounds than
Aristotle's questionable supposition that the Eleatic doctrine was an outgrowth and a
reaction against the Ionian monism." Apart from the implied accusation that Theo-
phrastus committed deliberate historical falsification, we must ask whether what fits
Aristotle's notions must necessarilybe dismissed as unhistorical?
34On the dependence of Xenophanes' meteorology on Anaximander's see Kahn
(above, n. 28) 204.
35So understoodthe Xenophaneanconcept is the directforerunnerof Parmenides'idea
of Being as the intelligible unity underlyingthe world composed of fire and night,-see
my "Parmenides:Between Materialand Logical Monism,"AGPh 70 (1988) 1-14.

114 AryehFinkelberg
Turningnow to Theophrastus,let us begin with his general approach

to Xenophanes' concept of God which is outlined in the statement
quoted by Simplicius:36
Theophrastussays that Xenophanesof Colophon, the teacherof

Parmenides,assumes that the principleis one, or that being and the
whole is one, and he agrees that the mention of Xenophanes'
opinion rather belongs to a study other than that concerned with
naturalphilosophy.
Some critics take these words to mean that Theophrastus,presum-

ably following Aristotle in the passage discussed above, dismisses
Xenophanes' idea of the One as too "naive" to be included within the
frameworkof his inquiry.37The Greek, however, does not mean this;
what it says is simply that Theophrastusadjudgesthe account of Xeno-
phanes' One to be more appropriateelsewhere. Where, then, if not in
the study of natural philosophy, should, in Theophrastus' view, the
Xenophanean notion of God be considered?38The clue is found in
Aristotleat De Caelo 298b15:
Some removed generationand destructionfrom the world alto-

gether. ... So maintainedthe school of Melissus and Parmenides.
But however excellent their theories may otherwise be, anyhow
they cannot be held to speak as students of nature. There may be
things not subject to generationor any kind of movement, but if so
36Phys. 22.22 (= Diels-Kranz[above, n. 9] 21 A 31).
37E.g., Jaeger (above, n. 8) 40 and n. 2; Guthrie (above, n. 8) 1.367; Babut (above,
n. 8) 436. The wide currencyof this misinterpretationis the more surprisingin that it not
only has no textual supportin the Greek of Simplicius' quotation,but is also at variance
with the indisputablefact that Theophrastusdoes account for Xenophanes' doctrine, cf.
Diels (above, n. 13) 113.
38It has been suggested that this domain must be theology. This is the view of Diels
(in Diels-Kranz [above, n. 9] ad loc., retained by Kranz: "Xen. gehire eigentlich zur
Theologie";but see Diels (above, n. 13) 109-110) which has become standardin Xeno-
phanean scholarship, see, among others, Reinhardt (above, n. 24) 102; Deichgrliber
(above, n. 6) 13 and n. 23; M. Untersteiner,Senofane, testimonianzeeframmenti (Firenze
1956) cxciv n. 94; Barnes (above, n. 17) 86 n. 6.

Studiesin Xenophanes 115
they belong to anotherand higher inquirythan the study of nature.

They [the Eleatics], however, had no idea of any form of being
other than the substance of things perceived; and when they saw,
what no one previously had seen, thatthere could be no knowledge
or wisdom without some such unchanging entities, they naturally
transferredwhat was true of them to things perceived.
In criticizing the Eleatics Aristotle proceeds here from his principaldis-

tinction between changeable, viz. corporeal being, the study of which
falls in the domain of naturalphilosophy, and unchangeable,viz. incor-
poreal being, to inquire into which is the business of first philosophy,
the highest kind of philosophic investigation. The Eleatics' fault lies in
their confounding the two provinces. Insofar as they assumed the
unchangeable,they were speaking as true metaphysicians,39but insofar
as they attributedunchangeableness to what was corporeal, i.e., to
essentially changeable being or nature,their teaching appearedto be a
naturalphilosophy and a mistakenone. If then the Eleatics have philo-
sophic merits, these are to be found in the province of first philosophy
rather than in philosophy of nature. As Aristotle says elsewhere,
"while inquiringinto truthconcerning being, they assumed being to be
the perceptiblealone."40
The locution Theophrastususes in his judgment on Xenophanes'
God, -'rpa;qevatt t0XXov Tfi;nrepi poGEA(o iaoropia•, is the same as
Aristotle's t&^XX6v oEatv -'~pat Kait rporpa;q 1i qniuotfjiG oCfIVEWo;
in his judgment on the Eleatics, and Theophrastus' -Trpaioropia can
hardlymean other than his master's -TrpaKa 7rpot rpa oGiVt;, that is,
first philosophy. What Theophrastusis saying is that the account of
Xenophanes' idea of God as the unchangeable One belongs not so
much to the study of naturalphilosophy with which he is preoccupied,
as to that of first philosophy.41It is thus an applicationto Xenophanes
of Aristotle's approach to the Eleatics. Theophrastusthus shares his
master's view of Xenophanes' God as a concept kindred to Eleatic
39Cf. Ph. 184b25; 185a12.

40Metaph. 1010al.
41 Cf. Diels (above, n. 13) 109-110 (cf. above, n. 38): "nimirumquae primaephiloso-
phiae sunt a naturaliumrerumcognitione arcet."

116 AryehFinkelberg
Being,42but so far it is still unclear with which kind of unity he counts

the XenophaneanOne.43
The synoptic picture of the Xenophanean doxography concerning
his "theological" teaching (Table 1) shows that the unity of "the
whole" (2) and its equationwith God (4) are two statementswhich con-
stitute the solid core of the original account: they are found in almost
all our sources ultimately depending on Theophrastusand in almost
identical formulation. The divine whole is said to be unchangeablein
Hippolytusand Cicero and eternallyselfsame in Ps.-Plutarch(3), which
is, in effect the same thing.
Now let us comparethis presumablyTheophrasteandescriptionwith
those found in Aristotle and Timon (Table 2). The resemblance
between the three descriptionsadduced in Table 2, which cannot stem
from each other (the fact that Theophrastus'description goes beyond
Aristotle's shows that, even if he took his master's formulation into
consideration,he definitely consulted the original, as the parallel with
Timon also indicates), confirmsthat all three authorsparaphrasecertain
Xenophanean lines." (piaot 6ooirl in Timon's paraphrasespeaks in
favor of Ps.-Plutarch's &ei with Hippolytus'"•o pzrEOT3oXfi
and Cicero's neque mutabilei0t1otov,
as a gloss. The fact that we have here the
paraphraseof Xenophanes' lines is of the utmost importance,for we
are now in a position to verify the doxographicequation of God with
the world and to assess its correctnessfinally and completely.
In Hippolytus and Ps.-Plutarch the paraphrase(2)-(3)-(4) is pre-
ceded by similar statements(1). Taking into accountthat in Hippolytus
42Cf. McDiarmid(above, n. 2) 116.
43 Barnes (above, n. 17) 84 n. 6, is correct in that "Theophrastusdid not say (pace
Jaeger ... [cf. above, n. 37]) that Xenophaneswas not a physiologos; ratherhe said that
Xenophanes' alleged monism was not a 'physical opinion."' Yet Barnes believes that
Theophrastusrelates the Xenophaneanmonistic doctrine to theology (cf. above, n. 38)
and in consequence misinterpretsthe intendedmeaningof Theophrastus'remark: "Xeno-
phanes' theological monism was lightheartedlyconstruedby Plato as an ontological mon-
ism; Theophrastussolemnly indicatesthatPlato is romancing." Setting aside the assump-
tion thatPlato misconstruedXenophanes'doctrine,Theophrastusmeans nothing like this;
he, like Aristotle, sharesPlato's notion of Xenophanes'God as an ontological concept.
44Steinmetz (above, n. 13) 37 and n. 69, 47-48, assesses the paraphraseon the strength
of the resemblance between Aristotle's and Timon's wordings; as does also Barnes
(above, n. 17) 99. Stokes (above, n. 8) 71, in his turn,observes the resemblancein word-
ing between Aristotle and the doxographies stemming from Theophrastusbut makes no
constructiveconclusions.

oi03 icvEi^zatseems to be superfluousin view of 50 tRetapooXfig in the

next phrase and that all other sources which refer to God's immobility
mention it considerably later, in (11), we should regard Hippolytus'
version as a loose paraphraseand preferPs.-Plutarch'swording.
The paraphrasewhich defines the essence of Xenophanes' concept
of God is followed by statementsof his characteristics. Three possibil-
ities are given in (5): ungenerated45(representedby the argument in
Ps.-Plutarch and by another such argument in Simplicius where the
order of (5) and (6) is reversed); 'ungeneratedand eternal'46in Cicero
and Theodoretus, in the latter (10); and 'eternal' in Hippolytus. The
question whether 'eternal' is a gloss cannot be answeredwith certainty,
but 'eternal' alone in Hippolytusseems to be wrong.47
Though God's oneness appears only in two sources, in Hippolytus
(6) and Simplicius (5) in whom it is again representedby the argument,
this attributionis nevertheless highly plausible, as the paraphrase(2)-
(3)-(4) strongly suggests. As distinct from oneness, Hippolytus' and
Alexander's predicationof God's homogeneity (7) is wrong. True,fr.
24 implies that God is somehow homogeneous qua single common sen-
sorium.48Still, homogeneity is not here God's attributeon its own mer-
its but the implicit corollary of his perceiving as a whole. Indeed, had
Aristotle found in Xenophanes the explicit statement of God's homo-
geneity, he need not have deliberated whether or not God somehow
allows internal diversity and, since he ignored Xenophanes' cosmol-
ogy, he would definitely have taken the XenophaneanOne as unity in
matter. Homogeneity thus was not in Xenophanes, and this, together
with the fact that it is found only in two sources, strongly suggests that
it was not in Theophrastuseither; ratherit is to be regardedas a later
45 Cf. fr. 14.

46 Cf. Arist. Rh. 1399b5; 1400b5 (= Diels-Kranz
[above, n. 9] 21 A 12, 13).
47 Stokes (above, n. 8) 70, is perhapsright in taking Adtius' attributionof ungenerated-
ness and eternity to the cosmos as supportingthe Theophrasteanorigin of 'ungenerated'
and 'eternal'as applied to God (AMt.2.4.11 = Diels-Kranz [above, n. 9] 21 A 37). But his
tentative conclusion that being found in Cicero (Acad. 2.118 = Diels-Kranz [above, n. 9]
21 A 34) they justify regardingCicero's omnia (plural) as also Theophrasteanis unwar-
ranted. In Theodoretus(4.5 = Diels-Kranz [above, n. 9] 21 A 36) who also says that God
is ungeneratedand eternalthe subject is in the singular,see Table 1.
48 On God as the common sensorium see von Fritz (above, n. 22) 228-230; cf. Gigon
(above, n. 24) 185; Jaeger (above, n. 8) 44; Lumpe (above, n. 8) 18; Guthrie(above, n. 8)
1.375.

118 AryehFinkelberg
Table 1
Hippol. Ps.-Plut. Simpl. Cic. Ps.-Galen
Haer. 1.14 Strom.4 Phys. 22.22 Acad. 2.118 Hist. Phil. 7
t oiTe y~veatv
oMboy
o& yive•~ ov?s
0peiperat Topiav
o-V&KtVEdat aLtLt,
arO
,
2 KaiiOn Ev a ETvat gav &Miv unumesse Tvat
T6
siRv ontv fYVttr6 n&v Ev omnia ndzvEa
Tv
ap•rv
Toov Katirav.
rIlot
3 i( neque id
LeTaxo•i;. daE
8iiotov" esse mutabile
4 TplTYoip y p Fv et id Kal oo
Kat ?Tv iai
Tou3To esse deum ot
IOnpX•etv
OEOov
y•EOv
dvat In&vOv
OEv 'AeyEv,
5 datov [dIyvriltov: 8vFva CvV neque natum

argument] i~iKcvUotv umquamet
[argument] sempiternum,
6 Iai tva dy~vi1Tov
[argument]
7 Katiiaotov
nvcivT1
8 Kai oiiTe
& conglobata
REvov Kait
ni•eepao- netpov o)tE figura. ivov,
nen•epao-
oqtatpoe~tm nerepaoL~dvov
eFvat . ..
10
11 oute
UE
vov oIUre ;OYtKov,
gLEVOVOUtFE
'tvoV-
lpegofvV...
12 Kati taot adeta-

TotSLoptot; OlTrlov.
atoOrlonK6v.
13 Kaizirdvra
voiv 5U
prlotv "&zaX'
X~Myov aabxb
... •Kpacat-
vet" [B 25].
14

Table 1 (continued)
Sext. Emp. Theodoret. Diogenes Alexander Nicolaus
Pyr. 1.224 4.5 9.19 ap. Simpl. ap. Simpl.
2 v Evat v dvat
7b L&v, 7r L&v
iE~pioe
3
4 mKa v olaitv
Oebvo•upvfi Oeoe
rot; naotv,
7 navraXdOeV
o8otov
8 edvat8o'zttpoet&- i,
opatpoetm nEnepaogpvov
ompatpoetL8 Ka) Kat •ratpov
Kat
nEnEpacLovOV, octpatpOetm
9 ptyuiv 4otov
EXo-oav
10 0o YEVtIo)V
aXStov
11 m
tnaOfi mti d tvar7ov
ri
Kat an•av
artvtrlov.
Xl9Trov
12 Ka'i 8Pov 8E
oyLKdv. Opav Kati
0 Kov duoiEtV,
g11
~lEVTOt
dvvatvoi3v
pcp6vlvotv
13
14 Kati itov.

Table 2
AristotleMetaph.986b24 Theophrastus
... ei 6 kov o-pavbv

Xov v tb cai avCont
vv
[=tb ^v149 Ol9oX1Jf;
t6bev dvai prlat
i0
g•taloXok/e__
o•otvov
tbvOE6v. icairouro060E6;
cott.
49 & )o;opav6; doesnotheremean'sky' (in which W. P6tscher,"ZuXenophanes,Frgm. 23,"

pace Diels-Kranz[above, n. 9] adfr. 23; cf. also Bonitz [above, n. 6] 84; Cleve [above, n. 21]
Kirk, Raven, Schofield (above, n. 7) 171-172, and is synonymous with 6 0kXovicai nx&vin Cae
1.154, ad 986b24.

doxographicadditionafter the patternof Parmenides'Being.5o

All our sources, Ps.-Plutarchbeing an honorable exception, ascribe
one or another spatial characteristic to Xenophanes' God (8). The
overwhelming majority are in favor of finitude and/or sphericity;
Cicero, who in Acad. 2.118 adheres to this attribution,surprisingly
changes his mind in Nat. D. 1.11.28, where he ascribes infinity to God;
Nicolaus Damascenus is also in favor of infinity, as Simplicius reports;
he himself proposes the apparentlywrong 'neither finite nor infinite.'
Diels takes the majorityopinion as authoritativeand deduces the Theo-
phrasteanorigin of 'finite' and 'spherical.'51Yet we saw that Aristotle
could find nothing in Xenophanes to indicate whether God is finite or
infinite.52Where then did Theophrastusobtain the information which
had been unavailableto Aristotle? The suggestion which first comes to
50 Some critics (see Burnet [above, n. 10] 125 n. 1; Deichgriiber [above, n. 6] 27 and
n. 44; Barnes [above, n. 17] 98-99) take Timon's fr. 60 (... OEbv... Toov &ncvrT
(d&rpefii)d&acr0i voespotepovin as an evidence of God's homogeneity. Here,
vorl•ta)
however, as distinct fromfr. 59, Timon obviously goes far beyond Xenophanes' wording
and even Xenophanes' very idea, for it is of course needless to argue that Xenophanes
could neither speak nor conceive of God as voep•kepog;i v6rlLa. Timon here accom-
modates Xenophanes' notion to later conceptual patterns, so that we need not, while
rejecting i7v6rlega, take Toov &nainv as more authentic, and the fact that
Aristotle voeptrepov
did not know about God's homogeneity indeed proves that it is not authentic.
This, however, is not to say that the word i8goto; did not occur in Xenophanes: Table 2
shows that Deichgrdiber(above, n. 6) is correct in suggesting that the word can be traced
back to Xenophanes. But we must distinguish between gLoto; without qualificationor
qualified with &eli(as in Ps.-Plut. Strom.4 = Diels-Kranz[above, n. 9] 21 A 32; cf. Timon
fr. 59.6: 6pion;6jooir) which gives 'the same,' 'selfsame,' and ~igoto; qualified with
adverbialslike or riavral (Hippol. Haer. 1.14; Simpl. Phys. 22.22;
Timonfr. 60 =ixcvrl,, cavax~•68Ev,n. 21 A
Diels-Kranz [above, 9] 33, 31, 35) which means 'everywherethe
same,' viz. 'homogeneous.' The appearanceof 'homogeneous' in some doxographersis
thus easily explicable, especially in view of Parmenides'use of i"Lotov.
51Diels (above, n. 13) 111 and n. 3, 112 and n. 2, 113; cf. Burnet (above, n. 10) 125;
McDiarmid(above, n. 2) 117; Steinmetz (above, n. 13) 54.
52Diels (above, n. 13) 110 n. 1, quoting with approvalSusemihl's (wrong, see above,
n. 13) exegesis of Metaph. 986b22, does not explain why Theophrastusallegedly goes
against his master's opinion here. The fact that Aristotle unambiguously testifies that
Xenophanes did not provide God with spatial characteristicsdoes not prevent many
Xenophanean scholars from asserting the sphericity of God, a temptation their ancient
colleagues also could not resist, see, e.g., Reinhardt(above, n. 24) 115-116; DeichgrAiber
(above, n. 6) 27 and n. 45; Gigon (above, n. 24) 183; B. Snell, The Discovery of the Mind,
trans. T. G. Rosenmeyer (New York 1982) 142; Kahn (above, n. 28) 80; Guthrie(above,
n. 8) 1.376-379; Cleve (above, n. 21) 10-11, 14; Barnes (above, n. 17) 98-99; Darcus
(above, n. 22) esp. 30-31.

122 AryehFinkelberg
mind is that Theophrastus,who, as distinct from Aristotle, took into

account not only the 'theological' but also the physical part of Xeno-
phanes' teaching, found finitude and/or sphericity in the latter and,
assuming the identitybetween the world and God, transferredit to God.
Yet apartfrom the lack of any evidence that Xenophanes asserted the
finitude and/or sphericityof the world, this possibility is precludedby
Theophrastus' view of Xenophanes' physical and 'theological' doc-
trines as issuing from alternativecognitive approaches--aistheseis and
logos (see below), which makes it impossible for him to come to con-
clusions about one of these doctrines on the basis of the other. Must
we therefore with Diels, who indicates as TheophrasteanREltEpaaotI-
vov 6•Kicai opatpoEti a&r6 [sc. Ti Gn&v= tbv OEv] 6t& 16 ravta-
X60EViotIov VCyEtv in Simplicius, suggest the Theophrastean origin
of the inference of God's finitudeand sphericityfrom his homogeneity
as found by Simplicius in Alexander?53Hardly, for Theophrastus,as
we concluded, did not attributehomogeneity to Xenophanes' God, nor
did Alexander, as Simplicius' reportmakes clear, ever claim that this
inference had been drawnby Theophrastus.From a more generalpoint
of view, it is highly improbablethat such an invention could belong to
Theophrastus:the inference is historicallyunwarranted,for while Par-
menides derived the finitude of his Being from its homogeneity (fr.
8.42-49), another 'partisanof the One,' Melissus, conceived of Being
as homogeneous and yet infinite. Could Theophrastusindeed have
overlooked this? And to saddle Theophrastuswith such a historically
crude inference one ought to be able, at least, to suggest some reason
for his making it. Yet there was no such reason: as we shall see, Theo-
phrastusneitherinterpretedXenophanes' God as formal unity, in which
case he might have sought supportfor his approachin God's finitude,
nor tried to find similaritieseverywherebetween Xenophanes' and Par-
menides' teachings.
On no account thereforecan the finitudeand/or sphericityof Xeno-
phanes' God be plausibly attributedto Theophrastus. The fact that all
our sources, with the exception of Ps.-Plutarch,provide God with one
or other spatial characteristiconly proves the relatively early (at least
as early as Cicero) corruptionof the Theophrasteanaccount, while the
consensus of the majorityof these sources on God's finitudeand spher-
icity shows that the assimilationof God to Parmenides'Being was the
53Diels (above, n. 13) 481 and n. ad loc.; cf. above, n. 51.

predominanttendency.54When homogeneity and sphericity penetrated

into the Theophrasteanreport, it is only naturalthat they should have
been connected in the Parmenideanmanner,i.e., the latterwas taken as
entailed by the former.55
We merely mention (9) which is a paraphraseof fr. 23, and (10) the
appearance of which here is obviously the result of some rearrange-
ment of the original order, and come to (11). Here four of our authori-
ties (in Ps.-Galen (11) and (12) are reversed) are concerned with God's
relation to motion. Of these four, three, Ps.-Galen, Sextus, and Theo-
doretus, to which one should add the reported opinions of Alexander
and Nicolaus, say that God is unmoved. This is undoubtedly Theo-
phrastus' view: God's immobility was implied in Aristotle's statement
of the common doctrine of the "partisansof the One" and is clearly
entailed by the descriptionof God as "eternallyselfsame" in the para-
phrase (2)-(3)-(4). On the other hand, Simplicius' minority opinion
"neithermoved nor unmoved"is patentlymistaken.
The synopsis shows that Theophrastus' description of God con-
cluded with the statementthat he perceives as a whole (12) and governs
all things with his mind, the latter being supportedby the quotationof
fr.25(13).
To sum up, what we can safely ascribe to Theophrastusbesides the
paraphrase(2)-(3)-(4), are the following statements: God is ungen-
erated and eternal, one, unmoved, thoroughlyperceiving, and govern-
ing all things by his mind. At the same time we can observe that the
doxographic sources at our disposal fall into three main groups accord-
ing to the way in which they deviate from Theophrastus'exposition.
The first group, representedby the great majority of our authorities,
reportswhat may be called the "Parmenidized"version of Xenophanes'
concept; these are the accounts of Hippolytus, Cicero (in Acad.), Ps.-
Galen, Sextus, Theodoretus,and Diogenes, as well as the view which
Simplicius found in Alexander;in all of these the characteristicallyPar-
menidean 'finite' and/or 'spherical,' and in two of them--Hippolytus
and Alexander-also the Eleatic 'homogeneous,' are mistakenly
54Cf. Jaeger (above, n. 8) 43 n. 23; Lumpe (above, n. 8) 29; cf. also Untersteiner's
arguments(above, n. 38) lxx-lxxvi.
55 As Stokes (above, n. 8) 75, puts it, "It would not have been difficult for a historianto
deduce the sphericity of the god from his being the same everywhere,and his being the
same everywherefrom his ability to see, hear,and perceive as a whole."

124 AryehFinkelberg
ascribed to Xenophanes. On the other hand, Cicero's (in Nat. D.) and
Nicolaus' (as reported by Simplicius) attributionof infinity to God
seems evidence of the existence of, in a sense, the contrary doxo-
graphic tendency, namely, the assimilationof God to Melissus' Being.
The third group is represented by Simplicius alone who ascribes to
Xenophanes the surprisingidea that God is neither finite nor infinite
and neither moved nor unmoved. As distinct from the two previously
described deviations from Theophrastus' account, that of Simplicius
has no rationalexplanationand is, in all probability,the result of some
doxographic corruptionof the original text. We must therefore try to
determinewhat exactly in Theophrastus'account could have given rise
to such a misrepresentation.
Let us repeat: Xenophanesdid not provide God with spatial charac-
teristics, nor did Theophrastusever attempt to do so. This, however,
does not necessarily mean that Theophrastuspassed the issue over in
silence. On the contrary, there is every reason to believe that Theo-
phrastussomehow had to refer to the question, for, apartfrom its gen-
eral interest, the spatial characteristicsof the One was precisely the
controversialissue among the Eleatics, and it would hardly be possible
to speak of the "firstof these partisansof the One" without mentioning
his position concerning this problem. It may be added that in his dis-
cussion of the "partisansof the One" Aristotle took the spatial charac-
teristics of the One as one of the two criteriaof whether the One was
conceived as formal or materialunity; true, Theophrastusseems not to
follow these distinctions, preferringhis master's more usual and gen-
eralized view of the Eleatics, but it seems improbablethat he merely
ignored a point to which Aristotle drew attention. Now to state Xeno-
phanes' position in the Eleatic controversyregardingthe spatialcharac-
teristicsof the One is precisely to say thathe had none, that he said nei-
ther that God is finite nor that he is infinite. We may assume thus that
it was some such Theophrasteanstatementthat was misinterpretedas
the "neitherfinite nor infinite"which we find in Simplicius.56Advanc-
ing this explanationof Simplicius' "neitherfinite nor infinite,"we must
apply the same kind of explanation to his "neither moved nor
unmoved"as well.57Let us consider whether some such explanationis
56 Cf. Jaeger(above, n. 8) 53 n. 64; McDiarmid(above, n. 2) 117-118; Guthrie(above,
n. 8) 1.369; cf. also Stokes (above, n. 8) 72 n. 15.
57To say, as Guthrie (above, n. 8) 1.369, does, that "Aristotle's negative verdict that
Xenophanes did not distinguish between material and non-material,nor (as is implied)

applicablealso in the lattercase.

As far as we can see from our material, Theophrastusadduces two
authenticlines to supporthis judgment that God is unmoved,fr. 26. It
follows that Xenophanes never made the formal statement of God's
immobility in all respects, and Theophrastus, like ourselves, had to
gather this from God's being "eternally selfsame," from the partial
descriptions like fr. 26, and from the general context of Xenophanes'
monistic doctrine.58If so it would only be naturalif Theophrastusnoted
this fact, namely, if he let his reader know that Xenophanes not only
did not state whether God is finite or infinite, but also whether he is
moved or unmoved. Nevertheless, Theophrastushad to continue, that
Xenophanes conceived of God as unmoved is clear from his calling
God eternally selfsame and from the following words, fr. 26. And
indeed, the fact that Simplicius, after reporting "neither moved nor
unmoved,"immediately proceeds to the exegesis of fr. 26 to show that
it is not incompatible with this thesis, suggests the original purportof
what became "neither moved nor unmoved." Being originally sup-
ported by the quotationof fr. 26 this could hardlyhave been something
other than the assertionof God's immobility59and in the form allowing
the doxographicmisinterpretationof the kind found in Simplicius, i.e.,
in the form I suggest.60
In Table 3 we can now restore the main points of Theophrastus'
descriptionof Xenophanes' God and compare it with the three kinds of
deviations displayed by the sources stemming from it. We can see how
the corruptionof Theophrastus'account gave rise to the whole spec-
between finite and infinite, is absurdlytwisted by the later writers into a positive state-
ment that the divine unity of Xenophaneswas both moved and unmoved, both finite and
infinite" is hardly to explain the emergence of the "neithermoved nor unmoved." It is
clear what was twisted into "neitherfinite nor infinite,"but what was twisted into "neither
moved nor unmoved"?
58 Steinmetz (above, n. 13) 53, is correctin saying that Theophrastusdid not apply the
notions "finite"and "infinite"to God, but is mistakenin adding (because of assuming that
the MXG closely reflects Theophrastus' account) that he did not apply the notions
"moved"and "unmoved"either.
59Cf. Stokes (above, n. 8) 72.
60My argumentproceeds from the assumptionthat Simplicius' reportis not a mechani-
cal compilation, see below in the text.

Table 3
Theophrastus Simplicius 'Parmenid

version
(1) ungeneratedand ungenerated ungenerateda

eternal eternal
(2) one one one
(3) it is not said neitherfinite nor finite and sph
whetherfinite or infinite
infinite
(4) it is not said neithermoved nor
whethermoved or unmoved
unmoved
(5) yet actually unmoved
unmoved
(6) thoroughly thoroughly
perceiving perceiving
(7) governingall governingall things
things by his mind by his mind

trum of the logically possible misinterpretations.61On the one hand,

garbling led to the textual misinterpretationslike those in Simplicius,
and on the other, dropping,in the course of epitomizations,what might
seem less relevant, naturallyled to excisions of "what Xenophanes did
not say"-(3) and (4), and the absence of (3) immediately opened the
way to doxographicsuggestions concerningthe issue.
The important difference between Aristotle's and Theophrastus'

reports is that while Aristotle deliberately confines himself to Xeno-
phanes' monistic doctrine, Theophrastus accounts both for Xeno-
phanes' conception of God and his cosmology. This being the case,
Theophrastushad to face the problemof the relationshipbetween these
two facets of Xenophanes' teaching.
About Parmenides and his opinion Theophrastus in the first

book of his Physical Opinions says thus: 'But Parmenides who
came after him'-he means Xenophanes-'took both ways. For
indeed, he both says that the whole is eternal and tries to account
for the coming-to-be of existing things, not however thinking
about both [ways] alike, but accordingto truthassumingthe whole
to be one, ungenerated,and spherical,while according to the opin-
ion of the many as to accounting for the coming-to-be of percepti-
ble things, positing two principles,fire and earth,etc.62
61It is odd to think, as some critics do, that
Theophrastus'(and Aristotle's) wordings
were necessarily directly responsible for the confusions in the later doxographers,see,
e.g., Jaeger (above, n. 8) 53 and n. 64; Guthrie (above, n. 8) 1.368-369; Stokes (above,
n. 8) 72 and n. 15. Theophrastuswas a conscious writerand, as far as we can judge from
De sensu, expressed himself pretty clearly (cf. G. M. Stratton, Theophrastusand the
Greek Physiological Psychology Before Aristotle [London 1917] 16 n. 1, 53; Kahn
[above, n. 28] 21 and n. 2). The later authors,on their part, were not so unintelligent as
not to understandTheophrasteanGreek; it is, of course, possible that sometimes certain
misunderstandingsmight arise because of careless reading of Theophrastus,but the real
source of the doxographic misrepresentationslies elsewhere, in inadequateabridgments
and condensations of the original account and rearrangementsmade on inadequate or
loose grounds, not to speak of copyists' errorsand mechanical corruptions. Gross slips
on the author'spartand inexplicable insensitivity or complicatedperversenesson the part
of laterreadersare scholarlyconventions.
62 Alex. in Metaph.984b3
(= Theophr.fr. 6 = Diels-Kranz[above, n. 9] 28 A 7).

128 AryehFinkelberg
Notwithstandinga slight corruptionof the text,63there can be no doubt

that this account of Parmenidesfollowed upon and was formally con-
nected with (toirc8p&',ltyev6ogevog) that of Xenophanes. The position
with which Parmenidesis contrastedhere is, then, that of Xenophanes.
Now the unpreparedemergence of &pC'p6tepat 66oi in the account of
Parmenidessuggests that they have alreadybeen defined in the preced-
ing passage, that is, in the account of Xenophanes. Are we thus to con-
clude that TheophrastusinterpretedXenophanes' teaching as a bipartite
system like Parmenides' though attributingto Parmenideshimself the
labelling of the "ways" as "accordingto truth"and "accordingto the
opinion of the many"? On the contrary:the Theophrastean"but Par-
menides who came after him ... took both ways" rathersuggests that
Xenophanes,on Theophrastus'account, took only one "way,"undoubt-
edly that of the "eternal whole."64 Ps.-Plutarch's report turns this
suggestion into fact. The opening sentence of his section on Xeno-
phanes provides the counterpartto the opening sentence of Theo-
phrastus' account of Parmenidesas quoted by Alexander;the parallel-
ism is so complete that we should assess this sentence as a quotationor,
at least, a very close paraphraseof Theophrastus'words:
[Ps.-Plut.Strom.4:] 'But Xenophanesof Colophon who pursueda

certain way of his own different from [that of] all these spoken of
beforehand [sc. the Milesians] allows neither coming-to-be nor
destruction but says that the whole is eternally selfsame ...'
[Theophr.fr. 6 ap. Alex.] 'But Parmenides who came after him
took both ways [sc. that of Xenophanes as well as that of his Mile-
sian predecessors]. For indeed, he both says that the whole is eter-
nal and tries to accountfor coming-to-be of existing things .. .'65
63 See Diels (above, n. 13) 482 and n. ad 1.8.
64Cf. Simpl. Phys. 28.4 (= Theophr.8 = Diels-Kranz[above, n. 9] 28 A 8): AESuKtRRoq

6/ ..., trlv atiiv HaPCEV i
0l&taElKcl~ REPitUv ivtoV 066v....
0 l.V MZEvocwVEFt
65One should say a few words here about the quality of the relevant portion of Ps.-
Plutarch'sStromateis. To begin with the Parmenideansection, it consists almost entirely
of more or less close paraphrasesof the text of the first book of Theophrastus'Physical
Opinions:
Ps.-Plut. Strom.5: Theophr.fr. 6 (ap. Alex.):
lapg 6• •tatpo; to~tp 6& Lv06 l-ap

JEvi&l& 0"'Eerrl;, EE~YEv61Evoq
ajia iEVK Toso lh-p7rto;'6 'EXde'"
;i'
EEvop•pvoI, tvtov
8o(o&vavteRtotMaato,, aja 6&%Kai.tiv &potpa; XOeta%`686o

Theophrastus'interpretationof Xenophanes' position is surprising:

how indeed could he assert that Xenophanes pursued only the way of
"the eternally selfsame whole" while accounting for the Xenophanean
doctrines concerned with "coming-to-beof existing things"? The solu-
tion, if there is one, must be that Xenophanes, while positing a certain
physical doctrine, somehow dismissed it or stated something about it
that could be taken by Theophrastusas such a dismissal. This being
the case, the famous Xenophaneanscepticism comes to mind, but to be
a suitable candidate, i.e., be possibly understoodas a sort of dismissal
of his own physical account, it must apply to the cosmological realm
alone. We must then examine the relevant material in order to deter-
mine the precise scope of Xenophanes' sceptical attitude.
ConcerningXenophanes' sceptical position we possess a numberof
reports,which fall into two groups. One includes Sotion as reportedin
Diogenes, and Hippolytus, both of whom depict Xenophanes as a per-
Evavdtav EXEetprlaEv
aToadav. [i.e., both adheringto Xenophanesand
diverging from him].
di8tov giv yTp O xR&v
Kci d~ivrltov KcayYap ; da6t6v oxttr6tn&v
anorcpaiveTat caxt tliv t(ov ov KcaiyVEatv
Rpaywtr &rnopawivatat
&?'ijOetavelvat yXpar6%[B &ano8t6vat netp&XatTcxiv 6vtxov,oij
8.4]"
6ooio; repiagPoTepowy
og0moV,
&U&6 gv iT
rcaT'&XijAetav
CV t6 v Kct
xyEv7jTov Kati a patpoEt8&;-j;o`aMo43civov,
i
Yti UTv Kcva'
yvEaEtv Kaj
v eu8fji
?1`MNnXtv a
0tX&av &•TxvvRoXXv E?it6
E
8ol'ovTOv dvat. yeveoiv adoIo8Ovat T-v watvo0vPov
io notuov ; ApcXdq,
x5RIp Ki%yiyv KtX.
Kattra aiatija&7
eK(kXXXcp
S ELC
eK
aXr1OEta;.ei'st
9p71o b'et rnap"T6 8v i•dppXEt, Theophr.fr. 7 (ap. Simpl.):
toUto olKcatv vT8EJO LROv Lv TOnapa Toov tv olcKOvo6iiv
6v" Toov. o•lKov"
t; ot; oIcK otxt.
To• Ot4TO0 tv ovpa v
oitv3vT%
aVyvrltov knoxdieit KcX.
We have, therefore, ground to expect-and the correspondencebetween Alexander's

quotationand the first sentence of the Strom.4 confirmsour expectation-that the section
on Xenophanesshould more or less faithfully reflect the exposition of the first book of the
Physical Opinions, the more so that in Ps.-Plutarchthe two sections constitute a continu-
ous exposition (as distinct from all other sources) reproducing that of Theophrastus.
(Close verbal parallels between the Stromateis 4 and Hippolytus' Refutatio are not per-
tinentto the issue, for they mostly occur in the materialwhich comes from other books of
Theophrastus'work.)

130 AryehFinkelberg
fect Sceptic who denied anything to be knowable;66the other group

consists of Timon and Ps.-Galen who qualify Xenophanes' position as
inconsistent and "mixed," asserting that, while applying the sceptical
attitudeelsewhere, in his monistic doctrinehe was a "dogmatic."67It is
not hard to see which of these two accounts deserves to be trusted.
Both traditions are of Sceptic origin68 and within this framework
Timon's and Ps.-Galen's presentation is, so to speak, the lectio
difficilior. Certainly, Timon who greatly esteemed Xenophanes-
obviously, because of the latter's sceptical attitude69-would have
been happier had Xenophanes been more consistent, had he been
tOXeto; &dxxpo;and not only We have thus no reason to
'xiaxvrpo;.
disbelieve Timon who, in addition, is our earliest and best qualified
source on the issue. We may therefore, following in this Diogenes,70
dismiss the rival account as a misrepresentationoriginatingin inferior
Sceptic sources.71
We can thus safely conclude that the scope of Xenophanes' scepti-
cism was limited to naturalexplanations,and hence it was the sceptical
qualificationwhich Xenophanes gave his cosmological doctrine that is
reflected in Theophrastus'statementthat Xenophanes abandoned"the
way of accountingfor coming-to-be of existing things" while pursuing
"the way of the eternal whole." Since, however, Theophrastuscould
not possibly have made the simple statementthat Xenophanes rejected
cosmology and then gone on to account for his cosmological doctrines,
we should admit that this statementwas not unqualified. The wording
66Diog. 9.20; Hippol.Haer. 1.14.2 (= Diels-Kranz[above, n. 9] 21 A 1, 33).

67Timonfr. 59 where Xenophanesis shown
repentingof his being poagpgo6ep3Xerxro;
because of positively assertingthe unity of the whole (cf. Sext. Emp. Pyr. 1.224 = Diels-
Kranz [above, n. 9] 21 A 35; Math. 7. 48-52); Ps.-Galen (Hist. phil. 7 = Diels [above,
n. 13] 604, cf. Diels-Kranz [above, n. 9] 21 A 35): toiSq
ks e vuL1ty•v ai'peatv LErel-
X-06tcat Evoqdv7Iv Rev Rvroyv
d ipxopcK6ra, ~ BECd6vov
npxdipxetv roiro •iepi ooyiatioavra rov. Cf.
Tb elvat RIadvraEv Kcai nxdpXetv Oeyv nexEpaaot(vovXoylKcbv &idprtiI
H. Diels, Poetarum Philosophorum Fragmenta (Berlin 1901) 45; id., "Uber Xeno-
phanes,"AGPh 10 (1897) 530 (= id., Kleine Schriftenzur Geschichte der antikenPhilo-
sophie [Hildesheim 1969] 53); Deichgraiber(above, n. 6) 30.
68 On Sotion's associating Xenophaneswith the Sceptic trend see Diels (above, n. 13)
146, 148.
69See Sext. Emp. Pyr. 1.223 (= Diels-Kranz[above, n. 9] 21 A 35); cf. Diog. 9.3.
704.20 (= Diels-Kranz [above, n. 9] 21 A 1): 9p01i %& oritMovtpvrtova~xbyveinriv
KaStcaximtxa E(vat n.
1r3v2a,niavootwEvos.
71 Steinmetz(above, n. 13) 23, is wrong in assumingthat Sotion follows Timon.

of fr. 34 and 35, which, as we now know, refer to the physical province
only and which are precisely the lines most probably quoted by Theo-
phrastushimself, give us an idea as to what Theophrastus'qualification
might have looked like: "the other way, that of accounting for
coming-to-be of existing things, he dismissed, declaring such accounts
to be no more than opinions deprivedof any certainty,saying thus (fr.
34); nevertheless he proposes some such opinion which he seems to
have adjudgedlooking plausible, as he says himself (fr. 35)."72
We can see now why Aristotle was reluctantto take Xenophanes'
cosmology as evidence of the formal characterof his One: it seems
that he was not certain of the precise purportof Xenophanes' scepti-
cism, whether it was restricted to explanations of the structureof the
sensible manifold or extended to its very existence. The status of
cosmology not being altogetherclear to him, Aristotle preferrednot to
base on it his conclusions regarding the kind of unity postulated by
Xenophanes; he thus neither referred to Xenophanes' cosmological
doctrine as evidence of the "formal"character of God, nor took the
sceptical qualificationof its validity as evidence of the contraryview.
The conclusion that Xenophanes' scepticism applies to natural
explanationsonly, not extending to the "theology,"naturallyraises the
question why the doctrine of the One is immune to it. To find the
answer we should first determinethe reasons for Xenophanes' physical
scepticism, i.e., his labelling naturalexplanationsas no more than more
or less plausible opinions. The main statementof Xenophanes' scepti-
cism isfr. 34:73
72In Theodoretus(4.5 = Diels-Kranz

[above, n. 9] 21 A 36) after the reporton Xeno-
phanes' "theology" we read: nityv Bi a3 t6&v(EVtB vrXy(ov E~ntlrXc
0Evo; Eti7K 1Yi q
<p9vat ~iiava Eiprl(KEv. Though somehow reflecting Theophrastus'general approach
this crudeexegesis must be, I believe, a doxographicgloss.
73 Frtinkel("Xenophanesstudien," Hermes 60 [1925], later included in his Wege und
Formen friihgriechischen Denkens2 [Miinchen 1960] 338-349; an English trans. of the
second part of the study under the title "Xenophanes'Empiricism and His Critique of
Knowledge (B 34)" is published in Mourelatos[above, n. 29] 118-131; subsequentrefer-
ences are to this translation)misconstruesthe Greek of the fragmentand misinterpretsits
purport. Frainkel'sexegesis has exercised much influence on subsequentinterpretations,
notably those of Untersteiner,Snell, von Fritz, and Guthrie. For the correctrenderingof
the Greek see: E. Heitsch, "Das Wissen des Xenophanes," RhM 109 (1966) esp.
208-216; see also Lumpe (above, n. 8) 34; Barnes (above, n. 17) 138-139.

132 AryehFinkelberg
Certain truth no man has seen nor will be anyone who knows
about the gods and concerningeverythingI speakof; for even if he
should happen to say what is true, he himself however does not
know [this];opinion is wroughtover all [men].74
To begin with, the two last lines of fr. 34 can hardlymean that precise
knowledge in the physical realm (for as we know now it is this realm
that constitutes the scope of the sceptical pronouncementof 34.1) is
unattainablebecause the objects of this knowledge are illusory, lacking
in reality, or the like. On the contrary,34.3 seems to imply that they
are perfectly knowable in themselves75(and are actually known by
God)76yet to the human being this kind of knowledge is nevertheless
unattainable. Why so? The answer that first suggests itself is that this
is due to the relativity,invalidity,or the like of our sense-perceptions.77
On a closer examination of the problem, however, we see we must
abandonthis approach. Quite apartfrom the question whether Xeno-
phanes indeed discredited sense-perceptions,we should ask ourselves,
is it because of the relativity, deceptiveness, or the like of our senses
that we lack certain knowledge in such mattersas to whether the rain-
bow is the goddess Iris or a cloud of a certain kind, whetheror not the
earth indefinitely stretches downwards, or whether the moon is
somehow useful to the world?78But let us consider the method to
which Xenophanesresortedin his physical theory:
74For the grammatical possibilities of construing 34.2 see U. von Wilamowitz-

Moellendorff, Euripides Herakles2 (Bad Homburg 1957) 3.62, ad 237; Friinkel(above,
n. 73) 127 and n. 38; Guthrie (above, n. 8) 1.395 and n. 3; Heitsch (above, n. 73),
223-224 and n. 60. In 34.4 I render i&tor as the masculine following Burnet (above,
n. 10) 121 n. 1; Untersteiner(above, n. 38) ccxxv n. 36; Barnes (above, n. 17) 138 and
n. 5. For the meaningof "gods"in 34.2 see below, n. 101.
75Cf. von Fritz (above, n. 14) 1559.
76Augustin. De civ. D. 7.17; Arius Did. ap. Stob. Ecl. 2.1.18 (= Diels-Kranz [above,
n. 9] 21 A 24); cf.fr. 23.2.
77 Cf., among others, Diels (above, n. 67) 53; Deichgr5iber(above, n. 6) 20-21; Heitsch
(above, n. 73) esp. 221-225. H. A. T. Reiche, "Empirical Aspects of Xenophanes'
'Theology'," in Essays in Ancient Greek Philosophy, J. P. Anton and G. L. Kustas, eds.
(Albany 1972) 88, stresses, in addition,"biologicaland culturalexpectancy-patterns."
78 Fr. 32, cf. Aet. 2.18.1 (= Diels-Kranz [above, n. 9] 21 A 39); fr. 28, cf. Arist. Cael.
294a21 (= Diels-Kranz [above, n. 9] 21 A 47); Aet. 2.30.8 (= Diels-Kranz [above, n. 9]
21 A 42).

... Xenophanes thinks that mingling of earth with sea takes place
and that in course of time earth is dissolved by the wet element,
claiming as proofs that shells are found in the midst of the land and
on mountains;and in the quarriesat Syracuse, he says, the impres-
sions of a fish and of seaweed have been found, on Paros the
impressions of a bay-leaf in the depth of the stone, and on Malta
flattened shapes of all sea-creatures. These, he says, were formed
when everything, long ago, was covered in mud, and the impres-
sion dried out in the mud. All men are destroyed when the earthis
carried down to the sea and turns to mud, then a new generation
begins.. .79
The basic approachis clearly empiricaland inductive. Certainfacts are

observed, their common denominator is suggested and some further
inferences are drawn. Not a word is said about the relativity or uncer-
tainty of the primarydata-the case is not that what one person takes
to be the impression of a fish another thinks to be a bird or not an
impression at all. It follows, then, that if there is something dubious
and uncertainhere, it is not the facts themselves, and hence it must be
the conclusions one draws from facts that are uncertain and unreli-
able.80Why so? For the answer we must returntofr. 34.
The fragment is a self-contained passage with a clear-cut structure:
the sceptical pronouncementis made in the firsttwo lines and the scope
of its applicationis delineated; in the last two lines, introducedby the
explicative ydp, the grounds of the sceptical pronouncementare stated
and the general conclusion is drawn.81What, then, are the grounds?
79Hippol. Haer. 1.14.5 (= Diels-Kranz [above, n. 9] 21 A 33), Guthrie's translation.

The report provides an excellent example of what Deichgriber (above, n. 6) 22, calls,
with referenceto Alcmeon'sfr. 1, Tekmerienmethode,that is, drawingconclusions by, as
Reiche (above, n. 77) 88, puts it, "collecting and collating empiricaltekmeria."
80Cf. von Fritz (above, n. 14) 1558.
81 Heitsch (above, n. 73) 222, points out that "Die Gedankenfiihrunggliedert sich in
zweimal zwei Verse; von ihnen enthiilt das erste Verspaareine Behauptung,das zweite
eine Begriindung",but immediately qualifies his own statement:"Diese Begriindunggilt
jedoch weniger dem sachlichen Gehalt der Behauptung,sonderneher der Art und Weise,
in der sie vorgetragen wird." The purpose of this, to say the least, unnecessary
qualification most clearly appears on 227: "Das mit yadpanschliessende Verspaarbe-
griindet,wie schon gesagt, nicht, weshalb der Mensch kein sicheres Wissen erlangt-eine
solche Begriindung lieferten jene Partien, aus denen die fr. 15 und 38 erhalten sind-
sondernerliiutert,wie Xenophanesdazu kommt, trotz allem Fortschritt(fr. 18), von dem
die Erfahrungja zeugte, auch ffir alle Zukunftzu behaupten,ein sicheres und endgiiltiges

134 AryehFinkelberg
Obviously, these lie in the impossibilityof ascertainingthe truthof our

inferences from the observed facts. One can only guess the meaning of
what one observes without being able to verify one's conclusions, so
that all such conclusions necessarily remainno more than guesses and
can never acquire certainty;for even if a guess is completely correct,
we have no means of ascertainingthis, just as we have none of ascer-
taining the contrary.82Nobody can say for certain whether the sun
indeed consists of small sparks as it seems reasonable to suggest;
nobody can prove whether the earth indeed was once mud as some
observationsseem to indicate;nobody can affirmwith entire assurance
that sea water is salt because of various mixturescarriedinto it and for
no other reason.83That is why there can be no certain knowledge
among humanbeings about all such things but only opinion, more plau-
sible and less plausible.84
Wissen ... werde es fiir den Menschennicht geben." Yet it is artificialto limit the expli-
cative force of the yd'pin fr. 34.3, and there is no justificationfor importingthe grounds
for the sceptical pronouncementof 34.1-2 from elsewhere.
82Cf. Plato, Meno 80D; Sext. Emp. Math. 7.46-52. Sextus' interpretationis therefore
generallycorrect,cf. U. von Wilamowitz,"Lesefriichte,"Hermes 61 (1926) 280; Kranzin
Diels-Kranz (above, n. 9) ad loc. pace Fr'inkel(above, n. 73) 124-125, followed by von
Fritz (above, n. 14) 1557-1558; Guthrie(above, n. 8) 1.395 n. 5; J. Lesher, "Xenophanes'
Scepticism," Phronesis 23 (1978) 2, and some others. Von Fritz (ibid.) objects that
Xenophanescould not say that he happenedto touch the truthin his criticism of anthro-
pomorphismbut did not know this himself; this is quite correct,but the objection is mis-
directed:Xenophanes'scepticism does not apply to the "theological"domain.
83AMt.2.20, 3; Ps.-Plut. Strom.4 and Hippol. Haer. 1.14.3; Hippol. Haer. 1.14.6; Hip-
pol. Haer. 1.14.4 (= Diels-Kranz [above, n. 9] 21 A 40, 32, 33). Nobody can know such
matters but God whose knowledge is free from the spatial and temporal limitations to
which humancondition is subject. It is here, if anywhere,that the traditionalopposition
between divine omniscience and human ignorance may be supposed to exercise an
influenceon Xenophanes. Deichgraiber,however, without any justification,specifies this
opposition as that between the Muse and the poet, for which he is correctly rebukedby
J. Mansfeld, Die Offenbarungdes Parmenides und die menschliche Welt (Assen 1964) 8
n. 1; Deichgr~iber'sapproachhas gained favor among Xenophanean scholars some of
whom go as far as to assume that this is all that Xenophanes wanted to stress (see esp.,
Steinmetz [above, n. 13] 40; Snell [above, n. 52] 139-141).
84This is, in fact, the well known demandby the Ionian ioroptrl (cf. Frfinkel[above,
n. 73] 131; Deichgraber [above, n. 6] 24, cf. 20; see also Heitsch [above, n. 73] esp.
194-205) of awto~ia as the prerequisite of reliable knowledge which Xenophanes
applied universally, i.e., also to the province of cosmological speculation where it is
impracticable. Frtinkel (ibid., cf. Jaeger [above, n. 8] 43 n. 2; Barnes [above, n. 17)
139-140) comparesthe HippocratictreatiseOn AncientMedicine, 1:

Having considered the reasons for Xenophanes' sceptical attitudeto

physical speculation we may now attempt to determine why the doc-
trine of the One does not fall within the scope of this scepticism.
Primafacie, if the impossibility of verification,that is, its purely specu-
lative character,is what in Xenophanes' view renders knowledge un-
certain, turning it into a mere opinion, this must apply to any kind of
speculative knowledge, both physical and "theological." Must we,
then, allow that Xenophanes did not trouble to be consistent, that he
simply declared that though human beings can never attain certain
knowledge about distant things and his accounts of heavenly and
undergroundthings were mere opinions, concerning the divine he pos-
sessed precise knowledge and therefore his account of One God was
If a man were to learn and declare the state of these [things in the sky or below the
earth], neitherto the speakerhimself nor to his audiencewould be clear whetherthe
statementswere true or not. For there is no test the applicationof which would give
certainty(o16yap oti, rpbqOTt Xpil &AvEviyKavTaE•i&vaitT •ocpq).
[Hippocrates,W. H. S. Jones, trans.(London-NewYork 1923) 1.15.]
Frankel (and Jaeger), misinterpretingXenophanes' position, fails to draw the true paral-
lels between it and the position of the Hippocraticwriter. Barnes (ibid.) points out the
continuity between Xenophanes' approachand that of the authorof the treatise. In fact
the quoted passage merely repeats what Xenophanes says in fr. 34 and about the same
things-it is precisely the knowledge of "heavenly and undergroundthings," the subject
of physical speculation, that is lacking certainty because it cannot be verified. To the
uncertaintyof the knowledge in the cosmological province the Hippocraticwriteropposes
medicine, where the empirical verification of assumptions is fairly possible and hence
precise knowledge is attainable. Barnes (ibid.), arguing from the treatiseto Xenophanes,
concludes that he "advocates a limited, not a general scepticism: it is theology and
naturalscience, not knowledge in general, that must elude our human grasp"(cf. Lumpe
[above, n. 8] 33-34). Barnes is correct,except for the "theology"which, as we shall soon
see, is quite a differentmatter.
It may be noted in this connection that the idea of the advance of mankind, which
seems to be the purportof fr. 18, is thereforesituated on a different plane, pace Heitsch
(above, n. 73) esp. 227. (Apropos: in 18.2 cannot mean more certain
ttetvov
knowledge, as notably Snell [above, n. 52] 139-140, claims, but betterconditions of life.)
If Xenophaneshere anticipateslatertheories,these are the Sophistic doctrinesof material,
social, and cultural advance rather than the idea of infinite scientific progress, as
K. Popper,"Back to the Presocratics,"in Studies in Presocratic Philosophy, D. J. Furley
and R. E. Allen, eds. (London 1970) 1.152, would have us believe. This advance in the
search for betterconditions of life involves, of course, the progress of knowledge, but of
practical knowledge whose attainabilityXenophanesnever denied. When we realize this,
the position of the Hippocraticauthorproves to be entirely identical with that of Xeno-
phanes.

136 AryehFinkelberg
the most certain truth? This seems highly implausible, if not to say
altogether paradoxical. At any rate, to saddle a thinker with incon-
sistencies and contradictions is not the best exegetical method, and
before resorting to it, it is always advisable to investigate other possi-
bilities. We should therefore explore the alternative, namely, that
Xenophanes drew an epistemic distinction between cosmological and
"theological"speculation,consideringthe latteras certainin itself. The
assumptionamountsto saying that in his monistic doctrineXenophanes
widely used apodeictic inference and, what is not less important,was
aware of its character. Theoretically speaking, this is not historically
impossible, for the distinction between apodeictic and non-apodeictic
knowledge is found in Parmenides.85
There is nothing inherently improbable in assuming Xenophanes'
use of logical proofs which, it seems, were alreadyemployed by Anax-
imander,86whose associate Xenophanes is reportedto have been. The
real question then is not whether Xenophanes could have used logical
argumentsbut ratherwhat extent and quality of argumentationcan be
admitted without transcendingthe limits of historical plausibility. It
may be stated at the outset that such an "upper limit" hardly exists:
Parmenides with his purely deductive doctrine of Being was, even if
much younger, a contemporaryof Xenophanes,87and nobody, I ima-
gine, would contend that had Parmenides been born some thirty or
forty years earlier, his doctrine could not have taken the inferential
form. Xenophanes thus could have resorted to logical proofs, and the
possibilities range between sporadic and rudimentaryuse and a con-
tinuous discourse like Parmenides'. The first question, therefore, is
what evidence we have to prove that Xenophanesresortedto argumen-
tation and what kind of argument,if any, did he use.
The allegedly Xenophaneanargumentationis amply presentedin the
relevant section of the pseudo-Aristotelian treatise De Melisso
85 As I argue in "Parmenides'Foundation of the Way of Truth,"Oxford Studies in

AncientPhilosophy 6 (1988) 39-67.
86 See C. Kahn,"Anaximanderand the ArgumentConcerningthe AFEIPONat Physics
203b4-15," in Festschrift Ernst Kapp (Hamburg 1958) 19-29; F. Solmsen,
"Anaximander'sInfinite: Traces and Influences,"AGPh 44 (1962) 109-131; E. Asmis,
"Whatis Anaximander'sApeiron?"Journal of the History of Philosophy 19 (1981) esp.
287-293.
87For Xenophanes' dates see: Steinmetz (above, n. 13) 13-34; cf. von Fritz (above,
n. 14) 1542.

Xenophane Gorgia (= MXG).88 Here Xenophanes is credited with

arguingfor God's being (1) ungenerated,(2) one, (3) homogeneous, (4)
spherical, (5) neither finite nor infinite, and (6) neither moved nor
unmoved.89Four of these, namely (1), (2), (5), and (6) are reportedas
Xenophaneanalso by Simplicius.9?Now of the six argumentsreported
in the MXG, four (3), (4), (5), and (6) should be dismissed at the very
outset as non-Xenophanean, for they are intended to prove theses
which, as we have seen in examining Aristotle's and Theophrastus'
accounts, Xenophanes never maintained. It may be added that (4) and
(5) are mutually incompatible, as is rightly observed by the author of
the Xenophanean section of the MXG himself. We are left thus with
two arguments(1) and (2) of which (1), as also (5) and (6), employ the
notions of being and not-being, thus betraying their post-Parrfienidean
date.91There remains then (2), about which we can say no more than
that it is not impossible that the kernel of this reasoning goes back to
Xenophanes.
Another source in which we find information about Xenophanes'
argumentationis Ps.-Plutarch-I mean his not too intelligible precis of
the proof of God's ungeneratedness. But before attemptingpenetration
into the rationalcore of this apparentlydistortedreasoningI would like
to discuss anotherpassage where Ps.-Plutarch'saccount also proves to
be patentlyconfused, the reporton Xenophanes' CtEpi 0EsV.
Its very location in the middle of the exposition of Xenophanes'
physical doctrine, between the report of the sun's origin and constitu-
88 For a review of current approaches to the MXG see: Untersteiner (above, n. 38)
xvii-cxviii.
891 would not subscribeto von Fritz' statement([above, n. 22] 228 n. 30) that "thereis
nothing in the argumentswhich he (the authorof the MXG) attributesto Xenophanes that
could not easily be retranslatedinto the comparatively simple form of the literal frag-
ments of Xenophanes' work."
90
Phys. 22.22 (= Diels-Kranz[above, n. 9] 21 A 31).
91Von Fritz (above, n.
14) 1550-1552, argues for the pre-Parmenideanorigin of (1)
suggesting that O5jotov,if taken as "gleich" ratherthan "ihnlich," makes the proof into
genuinely archaic reasoning. Yet von Fritz' argument is weak and requires additional
premises which are not in the MXG; incidentally 6a"otoooccurs twice in Xenophanes'
authenticlines (fr. 15.3, 23.2) and on both occasions not in the presumablyarchaicsense
"gleich" but as "ihnlich." Von Fritz' observation that the claim that "like cannot come
from like" would contradict common sense (which he takes as indicating the meaning
"equal"ratherthan "like") shows, I believe, that the argumentis a dialectical exercise of
relatively late origin.

138 AryehFinkelberg
tion and that of the earth's magnitude, leaves no doubt that there is
something wrong with our passage, for on no account can this be its
proper place. Moreover, not only does the passage prove to be mis-
placed, its content also betrays confusion. Indeed, &ico~etv•E icai
6pav KaO6ko ai ai•l icarX&gepo;, predicated here of the gods, is
associated in other sources with One God (see Table 1, line 12), and
this is undoubtedlycorrect, as is shown by the singularinfr. 24, olo;
6p0, oi~Xo8;&8voEt, 6~ ' &icoEt, to which this descriptioncan
olX)o;
be traced back.92But perhaps Xenophanes admitted the existence of
other, lesser gods who, he maintained,are also thoroughlyperceiving?
This suggestion does not save the situation: in Theophrastusthe phrase
was related to One God; whatever Xenophanes might have said about
the supposed "lesser gods," Theophrastuswas speaking of One God.
Two possibilities thereforeexist: eitherPs.-Plutarchor his source read-
dressed Theophrastus'reportabout God's perceiving as a whole to the
gods or he was mistaken about the true subject of Theophrastus'predi-
cation. The latter solution is clearly preferable:mistaking God for the
gods, in itself quite a possible doxographicerror, easily explains both
the surprisinglocation of the passage and the fact that in Ps.-Plutarch's
account of Xenophanes' monistic doctrinenot a word is said aboutGod
and his identity with "the whole," which in Theophrastuswas conse-
quent upon the statementof the Xenophaneanconcept of the "eternally
selfsame whole"; it is thus quite probablethat this Theophrasteanpas-
sage, not found in Ps.-Plutarch,was, as a result of doxographicconfu-
0tEv and then, being detached from the account
sion, altered into nrEpi
of the One, naturally relocated into a more appropriatecontext, the
cosmological part of Xenophanes' teaching. Now let us perform an
experiment. We shall take the nEpi O'Eov reportas the direct continua-
tion of the account of the "eternallyselfsame whole" and present the
entire passage as a column in synoptic table, while two other columns
will presentthe texts of the MXGand Simplicius (see below).
The parallels between the MXG and Simplicius, on the one hand,
and Ps.-Plutarch,on the other, are undeniable.93The argumentwhich
comes first in the Stromateis, though in itself differing from the first
92Cf. von Fritz (above, n. 14), 1555.
93This shows beyond any doubt that the nEpi Oestvreport in the Stromateis did not
come from the antianthropomorphicpolemics of the Silloi, as Reinhardt(above, n. 24)
94-95, suggests.

MXG3 Simplicius,Phys. 22.22 Ps.
[The 1st argument] [The 2nd argument]
(1) 'Al'6var6v 1(PlotvEdvat, ei' tI (4)ON F6Aieitiev ic toi
&aFyvyrov
&5 1 % ...
'aot, yevcoOat, Toiro XCywov Ecit 8E5V YVvd 1o
9%Evov 6•Lto ~ ei y
Cn yp irot 4i~
tol Oeo" &vdyI XX& bt' v
avogloiou ytveo0at. ~h yia
0
6cooio'i 4 1vogloio yevao0at tb 5v
•gotov dra0•; (ltv ln6 tolD
yev6pevovw uvatrbv&6 oi6v yhxpCRXhOv yevv&v
oi86&tepov" 6fto8ov" •R
oite yhxp
Ctgotov
69~' 6Lotol ii yevv&o0atIpooaiKetTob
'gTotov
v'
Rpooi5cEtvexcvwouivatgalt ov iK toT Ei F
•tk vx
6•olov" T Eavotiolo
teIrFvoat ... ovr' av , 5v
y(votto,~oaat t6toio i T
dAvogoiolo
OvTO;.
tdv6ototovyeva01at.
(2)ei Y•Xp t daOeveaot-
ytyvoto
poi tb iao p6tepov 7i E &c -
tovo; rIb.teiov , 'Ic XEipovo; tb
Kpettrov,1i tojvavTiov ta Xeipo
trv Ipettr6vov, t6ov E4 oi~c
K 1i
ifzvo % v yevio0at- 6enEp 6I-
vatov. di6tov gv oiv 8t&xra3tra
etvat trv Oedv.
[The 2nd argument] [The 1st argument]
(3) ei 86'a1tv 6 6S;&RdIdvtov (3) t6 yap v toiro cai irtv tv ano
icpdntatov, Eva air6bv 6
Oebyv XEyev0., yvi'va REv6et- o R
•vrlaiv
ipooliretv eTvat. ei tyXp 66i0 i FI T
1cV•otlV toi dvtO)VKpdtlatoxV
elvat* o~o
tva
RoEioci elev, o6Xav ai'6t pdat- •6to'vov yap, (raloiv,ovtov
ozov Kal 6logio; indpxetv av6ymcltnaoat
pPluXtoov abyv Evat

Rcvtov. yap civOEo;g
Ficaato;o ziv wpaeiv- b &6itvowv icpyTtoaov
Rot av otorog;etrl.
now0v 61o' Kaiaptarov0edg.
rouroy(p Oco'v
iat' EoAsdvalytv
Tvat, lcpa-rEiv, &x 1 Icpa-
reto• ami•a r cp
eivat. wa, wov prtoaxov
o ae ca0o il ipEir,
a-r&toooiVov oiwce vat Oe6v.
(4) nt'6vov oiv ivryov,ei Atv
id vah'AiAXWV
elev rai t
cpeiZttZroi;
Sitrotjo, o61&Cv eivat
Oeo6q-
nqxupc0vat yap b Eo yv il cpa-
retG0at. (5) i"ov & "i'vrow, oiK
av IV XV Go
exety U , 8v
eivat ow0t(PI)tV,
-EOf
Se
OV Slv
oti'T
pdatorov*
T6 ioov
pilttov o•'te XEipovevat rtol
Woot. that' Ecliep eI' xe Ical rotolu-
tov Eill 0e6;, Eva g6vov cvat r6v

8o6v. oi68kya'p o6& ndcvra -6-
vao0ati&v & po?hotto It et6vov
'va &padvat t6vov.
6vTmov"
[The3rdargument] [Absentin Simplicius]
(6) Eva 6' dvra tgotov dvcat ...
xc&vTr, 6p6tovra &t oiovra ra; tr
ica•
Ie aiAxa; ai'o0ioetS 6p
,Xov'ta
ei 'yapgir•, pa•rev&v Kai
xcvrl"
lt
KparGa a y th
)xc' Vi
&6Ln" agiep•p
Oeoi, tirep davatov.

argumentin the MXG (the second in Simplicius), is, as there, a proof of

God's ungeneratedness, whereas what in Ps.-Plutarch is mistakenly
reportedas Xenophanes' nEpi OEyvappearsto be originally a series of
arguments:the reportthat there can be no lordship among gods paral-
lels the argumentof the MXG for God's oneness, the argumentwhich
in both sources, the MXG and the Stromateis, comes second (first in
Simplicius),94while Ps.-Plutarch's statement that gods are free from
any want, conjoined with the paraphraseof fr. 24, is comparable,in its
position and reference to this Xenophanean line, with the third argu-
ment of the MXG. Before forming a general judgmenton the natureof
Ps.-Plutarch'sinformationwe must discuss it in more detail.
The only argumentpreserved as such in the Stromateis is that for
God's ungeneratedness. Yet as it stands the reasoning is neither
entirely intelligible nor possibly pre-Parmenidean. The premise is
pretty clear: "for had it ['the eternally selfsame whole,' God] come to
be, it is necessary for it not to be before this." But what follows is less
understandableand its wording bears the post-Parmenideanstamp:"but
not-being [must we spell 'not-Being'?] cannot come to be, nor can
not-being produce anything nor can something come to be by the
agency of not-being." Should we then translatead sensum "but being
not-Being it [changing the subject from not-Being to God] can never
come to be, nor can not-Being produce something, nor can something
come to be by the agency of not-Being"? But let us try another way.
Taking into consideration that at Xenophanes' time, and especially in
poetry, the article was used sporadically95and that supplying 'v with it
by later doxographerswould be only natural,we shall attemptto read
the argumentdroppingthe articles: Eiy•xp yiyvotro rouro, &vayicaiov
Rxp6ro5rou gi~lEvat- [-r] gili (E)6v 8 oicx &avy votro- (putting the
colon here) oi-r' (instead of o8)6' of MSS) &v [br] (E)6vnotioaot -t
til
o10E '3'o [Toi] gjrL TiV Tt,-"for had it ['the whole,' or
(')6vTo; 7YVOtZ'
God] come to be, it is necessary for it not to be before this; but not
being, it can never come to be: neither nought can produce anything
nor anything can come to be by the agency of nought." The sense is
94The whole report tEpi ~iOov being the misinterpretedaccount of One God, this argu-
ment must originally have been the proof of God's oneness. Cf. von Fritz (above, n. 14)
1554; Barnes (above, n. 17) 91-92.
95See R. Kiihner and B. Gerth, Ausfiihrliche Grammatikder griechischen Sprache4,
Satzlehre (Leverkusen 1955) 1.581-588; E. Schwyzer, Griechische Grammatik,A. De-
brunnered. (Munich 1950), 2.20-24.

142 AryehFinkelberg
perfect and is obtained precisely by eliminating what is post-

Parmenideanin the wording, the articles which suggest the conceptual
meaning of 6'v. Is this a mere coincidence? It should be noted that the
reasoning clearly implies God's being the whole of existence: had he
not existed, there would be nothing at all out of which or by the agency
of which he could come to be.96
Proceeding to the second argumentwe should note in advance that
the poor condition of the material allows no hope of an exact recon-
struction;all that can be done is to renderit intelligible while adding as
little as possible: "Among gods there can be no supremacy,for it does
not suit the divine holiness (ou0ytxp iotov) that god should be under
lordship;but were there many gods, there would be lords and subjects
among them (perhapsthe alternative'or all of them would be lords of
each other' was also posited, cf. the MXG and especially Simplicius);
hence there is only one God."97Whatever the detailed form of the
proof, what is obvious is that God's oneness was inferred from the
incompatibilityof the idea of the divine with that of being inferior to
one another. The way of arguing is characteristicallyXenophanean:
96 The
argumentas I reconstructit closely resembles Melissus' proof of the impossibil-
ity of coming-to-be (fr. 1). Barnes ignores the argument reported in the Stromateis,
preferringinstead to put his trust in (1) of the MXG which, he believes, though "contam-
inated by laterEleatic logic" nevertheless"containsa Xenophaneancore" ([above, n. 17]
87); in supportBarnes quotesfr. 1 of Epicharmus. Now if this fragmentparodiesXeno-
phanes' ideas, as it very probablydoes, the argumentalluded to seems to be that of Ps.-
Plutarchratherthan of the MXG; compareEpicharmus' to-j &B 'ica; g11i~Eovy' &rt6 tvo;
;"6tit Itp&aov with the restored argumentof the Stromateis. The "Xeno-
tLs8' i6ott
phaneancore" that Barnes extractsfrom the fragmentis that "a generatedgod must have
something to 'come from.'" This is again closer to Ps.-Plutarch'sversion than to that of
the MXG. Then Barnes abandons "the colourless" ylyvEaOat of Epicharmus(who is
presumablya live witness and at any rate, if he parodies Xenophanes, quite likely also
imitates his locutions) in favor of the mEicvovof the MXG (following in that Steinmetz
[above, n. 13] 52). Barnes finally concludes that the argument"statesthe necessary truth
that everythingthat is born has a parent"(ibid. 87). I wonderhow the authorof such rea-
soning gained Barnes' praise as "a brilliant,original,and sophisticatedtalent"(ibid. 84).
97Recognizing the identity of this argumentwith that in the MXG and Simplicius, von
Fritz (above, n. 14) 1554, points out that in itself the reasoning may serve for proving
both the oneness of God and G6itterdemokratie or rather,as he specifies, anarchy,prefer-
ring for this reason the setting of the MXG. Barnes (above, n. 17) 91, also says that the
argumentis "compatiblewith the pluralityof potent divinities each of which is at least as
great as anything else in existence" contending that the version of the Stromateis is the
misinterpretationof the reasoning found in the MXG. But Gotteranarchieis the reductio
ad absurdumratherthan a conceivable alternative.

the conclusion is obtained by explicating the idea, in this case

supremacy, necessarily present in the notion of the divine; the pro-
cedure is well attested in fr. 26 where the denial of God's locomotion
follows from its "unfittingness"for the divinity, and in Xenophanes'
sayings adduced by Aristotle where certain popular religious beliefs
and practices are dismissed as incompatible with the notion of the
divine.98The wording of the argumentbears, I believe, the unmistak-
able stamp of authenticity. I mean o y&p 6'atov, comparable with
n
o6 ~irtnp~nEt of fr. 26 and &~Epo~otvin the first of the two sayings
mentioned. Ea0cat also seems to be authentic. The word is
more specific8Eot6
than xpa-rEiv in the MXG and Simplicius and more shar-
ply conveys the contrastwith the divine dignity.
The last two phrases of Ps.-Plutarch'saccount of Xenophanes' iEpi
OEcovcontain what has remained of the inference in which, from the
premise that God is in want of nothing, the conclusion is drawnthat he
hears, sees and-we may safely add on the basis of fr. 24-thinks as a
whole. What therefore is lost-not simply remaining implicit as the
emphatic p il cKara ppo; added to Ka0oXou shows-is the statement
that possessing the perceiving ability only in one part of himself God
would lack it in anotherpart. We may thus outline the third argument
as follows; "God is altogether free from any want; but had he heard,
seen, and thought only in one part of himself he would be in want of
these in anotherpart;hence he hears, sees and thinks wholly and not in
one or another part of himself." The proof, like the previous one,
proceeds from the characteristicallyXenophanean analytical explica-
98Rh. 1399b5; 1400b5. See also Deichgriber (above, n. 6) 28-29; 0. Dreyer, Unter-
suchungen zum Begriff des Gottgeziemendenin der Antike (Hildesheim 1970) 21 n. 59;
Babut (above, n. 8) 431-434; Barnes (above, n. 17) 85-86. Babut is right as opposed to
Jaeger (above, n. 8) 49-51, who regards the prepon-category in Xenophanes as the
expression of a religious feeling; it also seems wrong to interpretthis category as purely
ethical, as in Gigon (above, n. 24) 191. For the meaning and development of the
prepon-category see M. Pohlenz, "TO HPEHON. Ein Beitrag zur Geschichte der
griechischen Geistes," in his Kleine Schriften (Hildesheim 1965) 1.100-139. Reiche
(above, n. 77) 93-95, objects that the understandingof the prepon-category in Xeno-
phanes as a purely aprioristic norm specifying the logically necessary connection of
God's essence with certain predicates, as Deichgriiber (above, n. 6) 29, interprets it,
causes the break between the "theology" and the cosmology in Xenophanes; yet this is
precisely what is alreadyproducedby Xenophanes' exclusion of the "theology"from the
scope of the sceptical attitude.

144 AryehFinkelberg
tion of the ideas necessarily implied in the notion of the divine, in the
presentcase, freedom from deficiency.
Having examined and assessed severally the argumentsretrievable
from the Stromateiswe should now appraisethe reliabilityof this infor-
mation in general. Our synopsis demonstratesthat all three accounts of
Xenophanes' arguments, those of the MXG, Simplicius, and Ps.-
Plutarch,can be traced back to one common source, and since the ulti-
mate source of the Stromateis is Theophrastus,he must be also the
source of the two other accounts. Theophrastus,it would then appear,
reported certain Xenophanean arguments, and this report has come
down to us in three versions. Two of them, those of the MXG and Sim-
plicius, preserved very little from their ultimate source while adding
much that is non-authentic;in consequence, these reports provide no
evidence for Theophrastus'account of Xenophanes' argumentsand a
fortiori of these argumentsthemselves." But have we reason to assume
that the informationconcerningXenophanes' argumentswhich we find
in the Stromateisfaithfully representsTheophrastus'account?
As we have seen, the Stromateis, certainly in their section on Par-
menides and in all probabilityalso in that on Xenophanes' God, closely
follow i~ of Theophrastus'Physical Opinions, which lends great relia-
bility to Ps.-Plutarch'saccount in general and the informationconcern-
ing the Xenophanean argumentationin particular. It is importantin
this connection that Ps.-Plutarchis the only doxographic writer (for
Simplicius, strictly speaking, is not such) who reportsthe Parmenidean
argumentand moreover, in its Theophrasteansetting;100 the authorof
the intermediarysource to whom the Stromateisgo back seems to have
had a taste for such things.
Now the orderin which the Xenophaneanargumentsare reportedin
the Stromateis (after restoring the original place and purport of the
99Reinhardt(above, n. 24) 91-93, was the first to assess the MXG's stemming from
Theophrastus'account (see also O. Regenbogen, "Theophrastusvon Eresos,"RE, Suppl.
7 [1940] 1544-1545), but was wrong in taking this as sufficient basis for using the MXG
as historicalevidence on Xenophanes,a mistake in which he is followed by Gigon, Stein-
metz, von Fritz, and Barnes. Those guided by the method recommendedby von Fritz
(above, n. 14) 1459, namely, to distinguish between the form and content of the MXG,
inevitably find themselves producing,as Barnesconfesses, accounts"of a somewhat arbi-
traryair."
'10 Two settings, Theophrastus'and Eudemus', are reportedby Simplicius, Phys. 115.
11 (Dieis-Kranz[above, n. 9] 28 A 28).

Table 4
Theophrastus Ps.-Plutarch Simplicius MXG
1 1. ungenerated: 1. ungenerated:
argument argument
la 2. ungenerated: 1. ungenerated:
argument(non- argument(non-
Theophrastean) Theophrastean)
2 2. one: argument 2. one: argument 1. one: argument 2. one: argument
3 3. it is not said
whether finite or
infinite
3a 3. homogeneous:
argument
3b 4. spherical,
argument
3c 3. neither finite 5. neither finite
nor infinite: nor infinite:
argument argument
4 4. it is not said
whethermoved
or unmoved
4a 4. neither moved 6. neither moved
nor unmoved: nor unmoved:
argument argument
5 5. actually
unmoved
6 6. thoroughly 3. thoroughly
perceiving: perceiving:
argument argument
7 7. governing all 5. governing all
things by his things by his
mind mind
alleged nEpiOE(ovpassage) strictlyparallels the order of Theophrastus'

descriptive statementsas reconstructedabove (see Table 3 above, page
126), thus complementing three of them with corresponding
arguments-see Table 4. Further,the comparison between our three
accounts of Xenophanes' arguments is instructive. The Stromateis
parallel the MXG and Simplicius only as far as they follow Theo-
phrastus,but differ from them, taking Theophrastus'side, where they
departfrom him, either in arrangementor in nomenclature;none of the
pseudo-Xenophaneanargumentscurrent among the doxographic writ-
ers is found in Ps.-Plutarch. As to the argumentsthemselves, none can
be dismissed as definitely non-Xenophanean because of a post-
Parmenidean character or over-sophisticated dialectics; they are not

146 AryehFinkelberg
only blameless in this respect but are apparentlyXenophaneanboth in

spirit and method. This being the case, we should assess the informa-
tion as authenticallyTheophrasteanand, as far as Theophrastus'sum-
maries are correct, which we also have no great reason to disbelieve,
the argumentsthemselves as genuinely Xenophanean.101
101 The assessment of the
authenticityof the argumentfor God's oneness, based on the
logical incompatibilitybetween the notion of the divine and the pluralityof gods (as well
as the recognitionthat Ps.-Plutarch'sreportIteptO 0Cv is but a misrepresentationof Theo-
o
phrastus' report Itept toi eooi) rules out the possibility that Xenophanes allowed the
existence of "lesser gods" besides One God, as some Xenophaneanscholars believe; see
Jaeger (above, n. 8) 43-46 and n. 34; Lumpe (above, n. 8) 26-28; Untersteiner(above,
n. 38) clxii; Cleve (above, n. 21) 8-9; Stokes (above, n. 8) 81; cf. G. Calogero,
"Senofane, Eschilo e la prima definizione dell'onnipotenzadi Dio," in Studi di filosofia
greca, eds. V. E. Alfieri and M. Untersteiner(Bari 1950), esp. 34. Von Fritz (above,
n. 14) seems to combine the "lesser gods" (1547. 48-68) with the argumentfor God's
oneness (1559). This being the case, the "polytheistic"interpretationof fr. 23.1 is pre-
cluded; those who, as, e.g., Guthrie(above, n. 8) 1.375, feel unhappywith Wilamowitz'
"polar expression" (see his Euripides Herakles [above, n. 74] ad 1106; cf. Uberweg-
Praechter[above, n. 8] 76; Diels-Kranz [above, n. 9] ad loc.; Burnet [above, n. 10] 129
and n. 1; Kirk, Raven, Schofield [above, n. 7] 170) should be referredto the modern for-
mulaic theory of the archaic hexameter. The explanations of the "lesser gods" as ele-
ments (see Kahn [above, n. 28] 165 n. 3, followed by Reiche [above, n. 77] 92; cf. Gigon
[above, n. 24] 178) is unacceptable: there can be only One God (so correctly Barnes
[above, n. 17] 83, 92) and, as the argumentshows, therecan be no other divine beings, be
they anthropomorphic,astral,or other. But let us set aside the argumentand examine the
problemof the "lesser gods" on its own merits with a view to explaining the 0eof in fr.
34.2 (see above, n. 74).
Supposing that Xenophanesallowed the existence of other gods beside One God, his
vigorous criticism of anthropomorphicbeliefs excludes the possibility that these gods
may be the anthropomorphicdivinities of popular religion. In that case they must be
"...the elements and sun, moon, and stars .. ." (Kahnloc. cit.). Reiche (loc. cit.), sub-
scribing to Kahn's explanation,draws the naturalimplication that "by physicizing them
[viz. the elements, fixed and errantstars] Xenophanesdefinitely does not mean that they
therefore cease to be theoi." This position is difficult to defend. First, Xenophanes
deprives these phenomena of what first and foremost makes the Greek god-
everlastingness;the sun, stars, and comets come from clouds (Ps.-Plut. Strom.4; Hippol.
Haer. 1.14.3; Aet. 2.4.11; 20.3; 3.2.2; 3.6 = Diels-Kranz [above, n. 9] 21 A 32, 33, 38,
40, 44, 45), the most ephemeralof phenomena,and because of this the sun and the moon
are constantlyborn and extinguished (AMt.2.24.9; 25.4 = Diels-Kranz [above, n. 9] 21 A
41a, 41); note that this is hardly compatible with Xenophanes' criticism of men for
believing that the gods are born and may "die"(it is not impossible, therefore,that it was
precisely the intentionto deprive the gods of popularreligion of their everlastingnessand
hence their divinity that promptedXenophanesto posit the somewhat extravaganttheory
of innumerablecoexistent and ephemeral suns and moons). Further,as Barnes (above,
n. 17) 96, points out, "... it is plain that by talking of 'what men call [Barnes'italics] Iris

Having seen that the accounts of the Xenophaneanargumentsof the

MXG, Simplicius, and Ps.-Plutarch ultimately derive from Theo-
phrastus we may attempt now to arrange all our sources, including
those on Xenophaneanreasoning,into a single stem.102
Discussing the views of the early philosophersin his commentaryon
Aristotle's Physics (in Phys. 20-28), Simplicius repeatedly refers to
Theophrastusas the source of his informationon the subject and con-
cludes his account of these philosophers by describing it as "an
abridged outline of what is reportedon the principles" (Phys. 28.30),
or the Dioscuri [fr. 32; A&t.2.18.1 = Diels-Kranz [above, n. 9] 21 A 39] Xenophanes

implies that there is, in reality, nothing divine about these phenomena ...." Lesher
(above, n. 82) 10 and n. 26, observes that Epicurus'explanationsof heavenly phenomena
"resembleboth in content and terminology those given by Xenophanes." (ThatEpicurus
was influenced by Xenophanes' doctrines is further supported by Sext. Emp. Math.
10.18.) Lesher concludes that Epicurus'rejectionof divine interferencein the movement
of the sky and the heavenly bodies is the explicit statementof what is implicit in Xeno-
phanes' natural account. At any rate, Epicurus used Milesian meteorology, of which
Xenophanes was a distinguished exponent. We can thus see that Xenophanes' accounts
of the "elements"and meteorological phenomenanot only do not suggest their divinity,
but ratherspeak against it (cf. Frinkel [above, n. 73] 130; Jaeger [above, n. 8] 42, 48-49;
Guthrie[above, n. 8] 1. 393).
Now what does Xenophanes say about gods? Two things. First, that they cannot be
such as men believe they are, which is germaneto his positing the alternativeconception
of One God. Secondly, that what men believe to be the gods are not such, which is per-
tinent to the physical realm: "She whom men call Iris, she too is by its naturea cloud,
purpleand red and green to see" (fr. 34); "... that which some call the Dioscuri are small
clouds glimmering because of their specific motion" (AMt.2. 18.1 = Diels-Kranz [above,
n. 9] 21 A 39). It follows that when Xenophanessays infr. 34 that nobody can have cer-
tain knowledge "aboutthe gods," he means that nobody can know what exactly is that
which men call Iris or the Dioscuri, or, putting it more generally, what men are in the
habit of calling gods. This is the domain which later came to be known as meteorology,
but for which Xenophaneshad no generalterm covering the scope of the naturalexplana-
tions as delineated by the Milesians. The "about the gods [that is, that class of things
which are commonly thoughtto be gods] and concerningeverythingI speak of [here, that
is, all other such things]" of fr. 34 is merely the archaicdescriptionof the domain later
known as meteorology.
102It should be stated immediatelythat in outlining the stem I do not intend to presenta
general picture of the Theophrasteantransmission;what I shall try to determineare the
main turning-pointsin the transmissionof the reporton Xenophanes' God which should
be postulatedto accountfor the state of the doxographicmaterialat our disposal.

148 AryehFinkelberg
that is, a short summaryof the "On the Principles,"the first book of
Theophrastus' Physical Opinions. Formally, Simplicius' account of
Xenophanes does not constitute an exception; here too he refers his
informationto Theophrastus,paraphrasingor even quoting his words
verbatim.103 It appears,however, thata considerablepartof the account
has nothing to do with Theophrastusand closely resembles the picture
presented in the MXG. This has made critics believe that in reporting
about Xenophanes Simplicius conflates Theophrastus' account with
that of the MXG,104 an assumption which leads to insuperable
difficulties.105Fortunately,it is quite unnecessary to assume that Sim-
plicius resorted to the MXG to explain the emergence in his reportof
the false propositions (3c) and (4a): these did not come into Theo-
phrastus' account from elsewhere but originated within it, being a
103See Diels (above, n. 13) 111-113, 480-481; cf. McDiarmid(above, n. 2) 116.
104This is the view of Diels (above, n. 13) 109-112, which has become standardin
Xenophaneanscholarship.
105It is quite inconceivable how Simplicius could have combined two such incompati-
ble reports(explanationslike McDiarmid's [above, n. 2] 118, are hardlyrealistic). Diels
himself (above, n. 13) 112, is at a loss to explain this-"nolo argutariqua ratione haec
cum priore sententia secundum metaphysica confirmataconciliaverit, utrum ambiguita-
tem illam allatis Xenophanisversibus demonstraveritan alia in capite de principioalia in
c. de deo attulerit";he then suggests (ibid.) that Simplicius had no access to Theo-
phrastus'work but drew his informationfrom Alexander(cf. Uberweg-Praechter[above,
n. 8] 74; Burnet [above, n. 10] 126; McDiarmid[above, n. 2] 116). This suggestion how-
ever is unwarranted-see Reinhardt(above, n. 24) 92 n. 1; Regenbogen (above, n. 99)
1536; Kahn (above, n. 28) 14 n. 1; but if abandoned,the hypothesis of the conflation
becomes altogether impracticable. Further,Simplicius nowhere indicates that he uses
some other account beside Theophrastus;we therefore have to allow that he takes the
MXG to be also by Theophrastus. But this suggestion makes Simplicius hold that Theo-
phrastusproducedtwo incompatible accounts of Xenophanes. Could he have believed
this? Could he, at the very least, not even have mentioned this extraordinaryfact?
Finally, we are told (Kahn,loc. cit) that Simplicius' presumablereliance on the MXGor a
similar source proves that "he was perfectly capable of ignoring the Phys. Opin." Let
alone that the isolated and doubtful example of such a practice on Simplicius' part can
prove hardlyanything,what Simplicius is admittedto be doing on this explanationcannot
properlybe called "ignoringthe Phys. Opin.";the only appropriatename is deliberatefor-
gery: Simplicius would then not merely have preferredanothersource to Theophrastus,
he would have combined Theophrastus'account with what he knew was not by Theo-
phrastusand moreover was entirely incompatiblewith him and proceededto ascribe this
amalgamto Theophrastus. Can one credit Simplicius with such conscious (and purpose-
less) falsification? (For the quality of Simplicius' scholarship see Diels [above, n. 13]
112).

product of garbling of (3) and (4)-see Table 3.106Evidently, this is

the same garbled version of Theophrastus' account from which the
MXG also ultimately derives its information. And, indeed, the differ-
ences in wording and even, to an extent, in the setting of the arguments,
between Simplicius and the MXG suggest their independentderivation
from a common source ratherthan the former's direct dependence on
the latter.107
To judge from Simplicius, the distortedaccount at his disposal con-
tained (la), (2), (3c), (4a), and (7)-see Table 4. It follows that it
could not have included (3), (4), and (5). Nor did it contain (3b), which
Simplicius, having found it in Alexander, dismisses as incompatible
with the "Theophrastean"(3c); nor, in all probability(3a), for Simpli-
cius does not reportGod's homogeneity, mentioning it only in connec-
tion with Alexander's (3b). Now since argument(3a) originates from
the reframingof (6), the lack of the former suggests the presence of the
latter;yet though reporting(7) Simplicius does not mention (6) which
is closely related to it, and this seems to suggest that in his source (6)
was lacking.
106 That the version of Theophrastus'work Simplicius used was mutilatedin more than
one respect seems to me obvious. First of all, this was very probablynot the whole of
Theophrastus'work but only the first book, "On the Principles": Simplicius' commen-
tary offers no evidence that he possessed something more. Further,it has been observed
that his reliance on Theophrastusis sometimes mediated by Alexander. This suggests
that he had no access to these pieces of the Theophrasteaninformation,i.e., they were
missing in his version of the "On the Principles." Diels (above, n. 13) 113, lists three
cases where Simplicius, as he maintains, clearly resorts to Alexander's reproductionof
Theophrastus: Phys. 38.20 (= Theophr. fr. 6), 115.11 (=fr. 7), and 700.18 (=fr. 15).
As to fr. 15, the mere mention of Alexander's name after Theophrastus'is not sufficient
testimony to reliance on Alexander;butfr. 6 and 7 are undeniable. In both cases the sub-
ject is Parmenides'doctrine,and Simplicius' knowledge of Theophrastus'account of Par-
menides seems not to go beyond this information. The conclusion that the section on
Parmenideswas missing in Simplicius' manuscriptof the "Onthe Principles"seems una-
voidable. Now in Theophrastus,the accountof Parmenidesfollowed that of Xenophanes
which, in turn,was precededby the account of the Milesians; in Simplicius, the confused
account of Xenophanesappearsat the very beginning, before the Milesians. It would fol-
low that not only did Simplicius possess only one detachedbook of Theophrastus'work,
but that the text of this book was damagedin the partdealing with Xenophanes and Par-
menides: the account of Parmenides dropped out entirely, while the report on Xeno-
phanes, fundamentallygarbled,was relocatedat the beginning of the book.
107Cf. von Fritz
(above, n. 14) 1552.

150 AryehFinkelberg
Simplicius' report shows that (3c) and (4a), having originated as a

result of the garbling of Theophrastus'text, were then provided with
proofs, presumably to bring them in line with other Xenophanean
theses accompanied by corresponding arguments; perhaps the cor-
rupted text made it obvious that some argumentshad been originally
reported, thus prompting attempts to reconstitute them. This also
seems to be the occasion on which (1) was provided with a new argu-
ment, thus being turnedinto (la): all three arguments,(la), (3c), and
(4a), have one essential distinctive feature in common-all resort to
the Parmenideannotions of Being and not-Being. Since it is hard to
see why anyone might be interested in replacing one proof of God's
ungeneratednesswith another, the simplest explanation seems to be
that the original argumentwas missing or mutilated. We should there-
fore conclude that Simplicius' source was a product not only of a
degradationof Theophrastus'text, but also, which is worse, of a "res-
torative" work of some doxographic writer. We must then postulate
that between Theophrastusand the "restored"account used by Simpli-
cius there was a version of Theophrastus'work in which the account of
Xenophanes was garbled (perhapsas a result of some abridgmentand
condensation, but perhaps merely because of careless copying and
similar technical corruptions): theses (5) and (6) as well as argument
(1) were lacking or unrecognizably mutilated, while (3) and (4)
appeared,because of rephrasingor textual corruption,in a form open to
just such a misunderstandingas that attestedin Simplicius' source.
Beside the view found in his manuscriptof the "On the Principles,"
Simplicius reports the views he found in Alexander and Nicolaus,
which diverge from what he takes to be Theophrastus. Thus he
informs us that, according to Alexander, being homogeneous, God is
finite and spherical,which amountsto argument(3b). Since Simplicius
takes this view to be Alexander's own and dismisses it as non-
Theophrastean,we must conclude that Alexander neither presented
(3b) as Xenophanes' argumentnor claimed it to be Theophrastus'rea-
soning. Further,Simplicius' &&icai oqptIpoEt0E; acxr
tnrexpaoaokvov
8ta
[sc. t6 na&v 'plaot] 6brEnavtaX60ev igotov [EEvopd6-
'AXcavSp6r
vrlv] XyEtv shows that in Alexander homogeneity was presentedas the
authentic Xenophanean statement. Does this mean that argument(6)
was reframedin the way known to us from the MXG, that is, as argu-
ment (3a)? Simplicius' wording seems to suggest that it still was not.
In our passage Simplicius consistently uses SeticvUco in references to

arguments,108while using Xyco for statements.109If so

iavtaXov60ev
Sgotov[EEvo<ptivrlV] Vyetv must indicate that in Alexander the homo-
geneity was reportedas Xenophanes' statementratherthan as an argu-
ment. To sum up, Alexander ascribed to Xenophanesthe assertionthat
God is homogeneous and, on applying Parmenideanlogic, claimed that
God is finite and spherical. We can therefore conclude that
Alexander's report stems from what has been termed above (Table 3)
the "Parmenidized"version of Theophrastus'account.
It is not hard to see that the "Parmenidized"version is also the ulti-
mate source of the overwhelming majority of our authorities-
Hippolytus, Cicero (in Acad.), Theodoretus, Diogenes, Sextus, and
Ps.-Galen, for all of them report(3a) or (3b) or both (see Table 1). At
the same time none of them gives (4) and (7), which indicates that
these were not in the "Parmenidized"version. Now the "Parmenidiza-
tion" of Theophrastus'account, i.e., the insertion into it of the theses
(3a) and (3b), which are closely related, is possible only if (3) was pre-
viously excised (provided, of course, that one does not assume a deli-
beratefalsification). It follows that between Theophrastus'account and
its "Parmenidized"version there must have been a report allowing the
insertion of (3b), i.e., lacking (3). Now the droppingof (3) must have
followed from epitomizations in the course of which Theophrastus'
account was abridgedby excising what seemed dispensable;obviously,
the Theophrasteanstatements of "what Xenophanes did not say," (3)
and (4), must have been the first to be dropped. It is therefore most
probable that (4), as well as (7) which may have seemed too self-
evident to require a special mention, were dropped together with (3),
that is, in the source between Theophrastus'account and its "Parmeni-
dized" version, a source which I shall therefore term the Abridged
Summary. This Summarymust thus have contained (1), (2), (5), and
(6), while dropping (3), (4), and (7); the "Parmenidized"version then
was the expansion of this Summaryby the addition of theses (3a) and
(3b) which, as far as we can judge from Alexander, were connected by
means of the inference after Parmenides' pattern, which amounts to
argument (3b), without, however, presenting it as Xenophanes' own
reasoning.
Of six accounts, those of Hippolytus, Cicero (in Acad.), Theo-
108 Diels-Kranz (above, n. 9)

1.121.29 and 31; 122.1 and 12.
109Diels-Kranz(above, n. 9) 1.122.7; cf. 121.28.

152 AryehFinkelberg
doretus, Diogenes, Sextus, and Ps.-Galen, which stem from the "Par-
menidized" version of the Abridged Summary, only Hippolytus des-
cends directly from it, for, while droppingthe arguments,as in all other
sources of this group, he reportsall the theses presumablycontainedin
it and in the same order (only (5) is relocated, being given earlier, see
Table 1). The rest of the accounts stemming from the "Parmenidized"
version report only some of the theses, in various combinations and
orders;what, however, is more significantis that all of them lack thesis
(3a). Such consistency can hardlybe incidental;ratherit suggests their
derivation from a common source differing from the "Parmenidized"
version in that it lacks (3a). In determining the mutual relationship
between these two sources we should take into account that the doxo-
graphic idea that Xenophanes conceived of God as finite and spherical
had to be, and actually was, as Alexander's reportshows, the result of
inference by the analogy with Parmenides from God's homogeneity
which, in turn,was easily suggested by God's perceiving as a whole.Ito
It is thereforenot very plausible that homogeneity and finitudeentered
the transmission separately, and at any rate homogeneity must have
come before finitude, not vice versa. It follows that the account in
which finitude and/or sphericity is reportedwhile the homogeneity is
lacking should be assumed to derive from a text in which both attri-
butes are present, being, in all probability, its abridged form. We
should therefore postulate the existence of an abridged form of the
"Parmenidized"version as a common source to which the accounts of
Cicero (Acad.), Theodoretus,Diogenes, Sextus, and Ps.-Galen can be
traced.
Beside the "Parmenidized"version, yet another expansion of the
Abridged Summary seems to have existed where Xenophanes' God
was described as infinitepresumablyby analogy with Melissus' Being.
The existence of this version, termed "Melissized" (see Table 3), is
suggested by the coincidence of the reportsof Cicero (in Nat. D.) and
Nicolaus Damascenus (ap. Simpl.). Finally, Ps.-Plutarch's account
most probablyis in direct descent from the Abridged Summary,being
mediatedneitherby the "Parmenidized"nor "Melissized"versions.
It remainsto consider the origins of the Xenophaneansection of the
MXG. This is obviously the productof the combinationof two sources,
the "Parmenidized" version of the Abridged Summary and the
110 Cf. Stokes (above, n. 8) 75.

"restored"version of the "On the Principles"which was also the source

of Simplicius. In the MXG two diverging transmissions of Theo-
phrasteanorigin met again to producethe "full"pictureof Xenophanes.
The "restored"account of Xenophanes from the "On the Principles"
was taken as the basis of the compilation, as the fact that the MXG
gives (la) ratherthan (1) of the "Parmenidized"version shows. This
must be put down to the compiler's respect for Theophrastus'name,
under which the "On the Principles"came to him, as distinct from the
"Parmenidized" version: we remember that Simplicius takes
Alexander's report as the latter's own view, evidence that it was
presented by Alexander anonymously, which would hardly have been
the case if it had circulated under Theophrastus' name. From the
"restored"version came (la), (3c), and (4a); (2) was in both sources;
(3b) which in the "Parmenidized"version was intended as explicative
was taken as Xenophanes' own argument;(6) of the "Parmenidized"
version was reformulatedin the Parmenideanspirit, thus being turned
into argument(3a), which made the thesis (3a), already present in the
"Parmenidized"version, inferential and thus provided (3b) with the
inferential premise and the whole with the characterof a continuous
deduction. (Of course, these changes, the turning of (3b) into Xeno-
phaneanargumentand the reframingof (6) as (3a) could have occurred
earlier, in intermediatesources.) To reconcile the apparentincompati-
bilities, (5) of the "Parmenidized"version was excised in favor of the
"Theophrastean"(4a) of the "restored"version of the "On the Princi-
ples"; in (3b) the "finite"was droppedto avoid the clash with "neither
finite nor infinite,"which, of course, did not save the situation,and the
compiler himself points out the incompatibility of "spherical,"which
implies finitude, with (3c). The compiler's lavish elaboration of the
"Xenophanean"arguments111was not, I believe, a conscious falsi-
fication; it was ratherthe "reconstruction,"on the basis of doxographic
summaries, of the argumentsas the compiler imagined they must ini-
tially have looked.
The entire stem is presentedin Table 5.
I1 On the non-Theophrasteanlocutions in the MXG see Diels (above, n. 13) 113.

Table 5 THEOPHRASTUS:
(1),(2),(3),(4),(5),(6),(7)
THE CORRUPTEDVERSIONOF T
THE 'ON THE PRINCIPLES': O
(1) withoutits argument;(2); (3) and (1
(4) corrupted;(5) and (6) missed; (7)
'RESTORED'VERSIONOF THE 'PARMENIDIZED'VERSION: T

THE 'ON THE PRINCIPLES': (1),(2); theses (3a) and (3b) added; (1
(1) suppliedwith the new argu- argument(3b) admittedas (5
ment = (la); (2); (3) and (4) explicative; (5),(6)
misinterpretedand suppliedwith
the arguments= (3c) and (4a); (7)
THE ABRIDG
THE 'PARME
VERSION:
(1),(2); (3a) dr
(3b),(5),(6)
Simplicius: The Xenophanean Alexander: Hippolytus: Cicero (Acad.):

the orderof (la) section of the theses (3a) (5),(1),(2) (2),(5),(1),(3b)
and (2) reversed; MXG: and (3b); (3a),(3b),(6) Theodoretus:
(3c),(4a),(7) (la),(2); (6) argument(3b) (2),(3b),(1),(5)
restatedas as explicative Sextus:
(3a); (3b) as (2),(3b),(5),(6)
Xenophanean; Ps.-Galen:
(5) and (7) dropped (2),(3b),(6),(5)
Diogenes:
(3b),(6)

Some of Xenophanes' argumentshaving been explicated and his use

of inferential proofs thus firmly established, we may proceed to ask to
what extent the Xenophanean monistic doctrine was inferential. For
indeed, the fact that Xenophanes used logical proofs does not neces-
sarily mean that his conception of One God was deductive or predom-
inantly deductive, for his argumentsmight well have been sporadicand
of secondaryimportancewithin the general framework. Yet we have, I
believe, good reason to conclude that precisely the contrary was the
case. Let us compare Xenophanes' argumentationaccording to its
order in Theophrastus' account (Table 4) with fr. 8 of Parmenides'
Aletheia:112
Parmenides Xenophanes
(1) Ungenerated andeternal (1) Ungeneratedand eternal
[8.6]. ...rva yxp y7vvav8tlioCae [Ps.-Plut.Strom.4]: ei y&pytyvotro
o
TO)Co,avaylccaovRp TO)toR) gi1
[7] m' n"60eva6I\O'v; o06'
8K rl evalt [To] (E)6v &8 o3DC&v
R•7
dovzogEaooom yvotTo- ouir' [insteadof o68' of
[8] o' voesvicrX. MSS] av ['r] Ril (A)6V Ro017iuatlit
(po•oOat r9o6Di
[12]o~'6 inor' p
crl d6vro;~1prioet oIte %)r[Tol] A(A)6vro;g
yvoti'
Rioeto; ioXClg av Tt.
[13]yyvsaeai 'rt nap'C9a6 'ricX.
(2) Oneandcontinuous(8.22-25) (2) One
(3) Unchangeable (8.26-31) (3) Unmoved
[8.29] rau6r6v t' v te gvov [fr. 26.1] aleFl8' v zaj'fP gigve
1
Ica0'haDTy6 teTaUiT,
cei'rat vo 9evoo86'v IdX.
[30] xo ;FoS9'gReS6ov aiOt giCve
(4) Perfect(8.32-49) (4) Thoroughlyperceiving
[8.32]otveFievoix d&reh [Ps.-Plut.Strom. 4]
ErrlXov Tb itti6eo0ai Te
F6vOIt; dvat- grnSvb; ... gr8',k;-
[33] Foat
yTp o6•ICIt6ieviq.rl- bv
8' &v Rav'bS9Se&iTo.
[34] TaDb9yv 8' oati voeiv e Kcai 8%iali6pav(IcavoEsv)
adCoItyv
oiveFievi~For V6nTIa. cai rl caraTg po; =
Ica0dkoxoD
112In the segmentation
of the fragmentI follow Schofieldin Kirk,Raven,Schofield
(above,n. 7) 249-253.

156 AryehFinkelberg
[35] 0o ya'pa&veU
Toi 6vTo;,9v fr. 24:
oaytv, voe' , o~o; &
oieo 6p , oi~Xo&5
n•~paqctoievov
b voEIv...
[36]eFiprijaet; I' a&couet.
[42] au6,rxp~ineinieipa;q n~iatov,
C0aTi
teTeXehRTLevov
[43] ni&vroOev, etGmrickou aq(paiprI
EvalyKtov icyK(p,
[44]geCaa60evivaoinag idv'rn'T6 "
ytp orTe Rt Ciei ov
[45] oTrEe It 1pai6repoviteXvat
1i
pe6v a'tt'tf7
[46] o rzeyazpoKEcov'tf7.gaoUt,x6 cev
aotaIoi1viiceio0at
[47] ei; 6Og6v,oj ' byvagrtyviuno;
ei'rcev d6vro;
[48] 'i &Xhovti 8' 8 faooov, nEi%
Rnavanttv~atoukov
[49]of yapiauvroOev•Tov, v
RneipautKicpet. 6.ilo;
The order of Parmenides' argumentsfollows the reportedorder of

Xenophanes', and this becomes even more evident if we take into con-
siderationthat God's thoroughperception is undoubtedlya manifesta-
tion of his perfection.113The parallelism is the more notable in that
Parmenides'argumentsare only loosely connected with each other and
could actually have been arrangedotherwise. The parallels are found
not only in the arrangementof the reasoning but also on the linguistic
plane, and though scanty, they are especially significant, for Xeno-
phanes' argumentshave come down to us in summariesand corrupted
ones at that; where we have the authenticwording, namely in the case
of "unmoved,"the resemblance is striking indeed.114Finally, in the
Parmenideanarguments(1) and (4) we can easily discern variationson
the correspondingXenophaneanideas,115(4) being especially interest-
113Cf. fr. 23.2: "[God]is not like mortalseither in body or in thought." Barnes(above,
n. 17) 94 and n. 20, rightly equates God's being free from any want with his perfection,
adducing as the parallels Eur. HF 1345 (cf. Guthrie [above, n. 8] 1.373), Antiphon'sfr.
10, Xen. Mem. 1.6.10, and Diogenes the Cynic ap. Diog. 6.105.
114Cf. Reinhardt(above, n. 24) 112-114; Stokes (above, n. 8) 83 n. 53.
115Cf. Diels (above, n. 13) 111 n. 2: "Xenophanisrationemrepresentarilicet ex Par-
menid. v. 66 St." (Diels quotesfr. 8.5-8 and 12-13).

ing in this respect. Parmenides begins by applying to Being Xeno-

phanes' denial of any want to God; the meaning of the following lines,
8.34-36, is obscure and controversial,but the undisputablefact that a
certain connection is establishedbetween the One and thinkingsuffices
for our purpose: the sequence of Parmenides' ideas is the same as in
Xenophanes, from oiiccitt6ei;g to voeiv. After a short digression in
which he dismisses mortal errors (8.38-41), Parmenides returns to
Being and proceeds to describe its perfection-the idea manifestly
present in the Xenophanean God's thorough perception-and then
especially dwells on the homogeneity of Being, again the very charac-
teristic implied in God's being a single common sensorium. Par-
menides thus moves in the circle of Xenophanes' ideas, developing and
explicating them, a fact which points to an influence much deeper than
any formal dependence.
The comparisonis indeed illuminating: in his deductionof the attri-
butes of Being, Parmenidesmutatis mutandisfollows fairly closely, as
our comparative material, scarce as it is, shows, the patternof Xeno-
phanes' "theological" poetry. This proves the essentially, though not
necessarily wholly and thoroughly, inferential characterof the Xeno-
phanean monistic doctrine.116I am preparedfor the objection that, if
Xenophanes was indeed the pioneer of inferential discourse, why do
we hear nothing about this from our ancient authorities? Well, suppos-
ing that Parmenides was this pioneer, do we hear anything about
this?117
116 In reconstructingTheophrastus'account I did not have Parmenides'fr. 8 in mind;

the idea of tentativelycomparingthese occurredto me later,and the result was surprising.
The close correspondencebetween Parmenides'fr. 8 and my reconstructionof Theo-
phrastus'accountboth verifies the latterand attests the basing of Parmenides'thought on
Xenophanes. The extent and specificationsof these influences are naturallyopen to argu-
ment, but not, I believe, the fact itself.
117 We are in a position now to decide whetherXenophaneswrote a formal composition
on the philosophic subjects. The ancients say that he did; Burnet (above, n. 10) 115, fol-
lowed by Jaeger (above, n. 8) 39 and nn. 6, 7, 8; 40 and nn. 9, 10, 11; Steinmetz (above,
n. 13) 54-68; and von Fritz (above, n. 14) 1545-1546, disagree; see in detail in: Unter-
steiner (above, n. 38) ccxxxiii ff. Now it is obvious that the idea that Xenophanes
presented his "theological" conception in sporadic utterances scattered throughout his
poetry is not really compatible with the inferential,or close to such, form of his doctrine;
and it is altogether incompatible with the fixed order of the arguments,the existence of
which is proved by the parallelism between Theophrastus'account of Xenophanes and
Parmenides'fr. 8. It follows then that Xenophanes' monistic doctrine could have been
expounded only in a continuous poetic passage, either a separatepoem or a part of a

158 AryehFinkelberg
In the light of our conclusions we can understandthe purportof a

curious testimony by a man who most likely was personally acquainted
with Xenophanes at Hieron's court in Syracuse, had occasion to listen
to his performances118 and parodiedhis doctrines in comedies. I mean
Epicharmus.'19
In Metaph. 1010a6 Aristotle, criticizing his predecessors, remarks:
"... therefore,while they speak plausibly, they do not say what is true
(Eisc6oKggvAvXYO1otv,
oiU anrl0fi
& •• • (for it is fittingto put
the matter so ratherthan as Epicharmus,youatv)
put it against Xenophanes)."
What then did Epicharmussay? There are three theoretical possibili-
ties: (a) "he speaks plausibly and what is true,"(b) "he speaks implau-
major inclusive piece. At the same time,fr. 34 presupposesas its context a more or less
extensive passage concerned with philosophic topics (cf. Deichgraiber[above, n. 6] 19;
Gigon, Untersuchungenzu Heraklit [Leipzig 1935] 151; Untersteiner [above, n. 38]
c-cxiv; Barnes [above, n. 17] 83-84), specifically, as we have seen, with naturalexplana-
tions. There are therefore two possibilities: either the monistic deductions and the
naturalexplanationsconstituted two parts of a unified philosophic composition, or they
were two separatepoems, on God and on the world (whetheror not combined with other
topics such as criticism of popularbeliefs, etc.). Now there are three reasons for prefer-
ring the former option. First, the ancients knew only one Xenophanean philosophic
piece, which is thrice referredto in our sources as lepi P•0ioe;,-by the grammarians
Crates of Mallos and Pollux (Diels-Kranz [above, n. 9] 21 B 30, 39) and presumablyby
AMtius(Diels-Kranz [above, n. 9] A 36, the second text, wrongly assessed by Diels as
"aus d. homerischenAllegorien,"-see J. Mansfeld, "Aristotleand Others on Thales, or
The Beginning of NaturalPhilosophy,"Mnemosyne38 [1985] 127 n. 64); cf. also Diog.
8.56 (= Diels-Kranz [above, n. 9] 21 A 5). The title by no means indicates that the con-
tent of the piece was restrictedto cosmological speculationalone; HEptipoEoS is usual
in ancient references to the early philosophic compositions, and even the Parmenidean
poem was referredto as HEpi ple o; (Suda s.v.; Diog. 8.55; Simpl. De Caelo 556, 25 =
Diels-Kranz [above, n. 9] 28 A 2, 9, 14). Secondly, the combinationof the monistic and
cosmological doctrinesas two formallydistinct partsof the same poem is attestedfor Par-
menides, which, in view of the overt influence, may be taken as suggesting the unified
composition in Xenophanes as well. Thirdly, in the absence of contraryevidence, we
should admit the unified poem in orderto meet the requirementof the most economical
solution. We may thus assume that Xenophaneswrote a poem known in antiquityunder
the conventional title Hepi <piamo;, which formally fell into two parts, on God and on
"the gods and all other such things," that is, cosmology (cf. above, n. 101). It is quite a
different question whether the piece was a didactic epic or a Sillos, i.e., whether it also
included criticism and polemics, and whether it was in hexametersor in mixed meters,
questions which, it seems, can hardlybe solved.
118See Diog. 9.18 (= Diels-Kranz [above, n. 9] 21 A 1): A&k& icKaacrgT6 et
ppa•4
119Cf. above, n. 96; see also Reinhardt(above, n. 24) 122-125.

sibly and what is not true," (c) "he speaks implausibly but what is
true." Of these only (c) is a true witticism capable of sticking in the
memory and worth recalling and at the same time providing Aristotle's
"plausibly but what is not true" with the best contrast.120The next
question is, what was Epicharmus' point of reference? As we have
seen, Xenophanes declared that his cosmology was not true but only
plausible, using, it must be noted, a similar expression,,otico6Ta toi
T6CIotat(fr. 35). Now if it was the Xenophaneancosmology that was
Epicharmus' intended target, i.e., if he simply derided Xenophanes'
sceptical qualification of natural accounts, the joke appears to be an
idle mockery which could hardly appeal to Aristotle and be described
by him as what "Epicharmusput against (or 'in regardto'-eiq) Xeno-
phanes." It seems therefore much more probable that the target was
Xenophanes' "theology." Now it is perfectly understandablethat one
can call the concept of a thoroughlyperceiving and totally immovable
cosmic god implausible, but it is not immediately obvious why one
should at the same time assent, though ironically, to the truth of this
idea. This, however, is easily explained if we bear in mind that Xeno-
phanes' implausible doctrine was supportedby "irrefutable"proofs. It
seems, therefore, that Epicharmus ironically alludes to Xenophanes'
own qualification of his natural explanations: they are, as he himself
confesses, plausible yet hardly true; his account of the divine is, of
course, true, yet, regrettably,implausible. The situation, amply illus-
trated in Plato, is quite puzzling for anyone inexperienced in logical
discourse as all Xenophanes' hearerscertainly were: the starting-point
seems undeniably true but the conclusions are surprising and hardly
acceptable. It was precisely this paradox, I believe, that made Epi-
charmus call Xenophanes' "theology" "implausible truth." Such was
its effect on contemporaryhearers and such witticisms were current
among them.121
120 See Th. Gomperz in Diels-Kranz (above, n. 9) ad loc.; Ross' note ad loc. in the
Oxfordtranslation,and Ross (above, n. 2) 1. 276, ad 1010a6.
121We can now appraise the extent of
Jaeger's misinterpretationof Xenophanes'
thought when he writes: "It is nothing that rests on logical proof, nor is it really philo-
sophical at all, but springs from an immediate sense of awe at the sublime of the Divine"
([above, n. 8] 49). Von Fritz (above, n. 14) 1547, correctly warnsthat "das logische Ele-
ment in X.s theologischem Denken ... sollte ... nicht Uibersehenwerden"(cf. Reinhardt
[above, n. 24] 100; Deichgriiber[above, n. 6] 28-30; Barnes [above, n. 17] 94). Yet the
very controversy as to whether Xenophanes was a religious mystic, as Nietzche first
claimed (Die Philosophie im tragischenZeitalterder Griechen in Werke,ed. K. Schlechta

160 AryehFinkelberg
After this long discussion of Xenophanes' argumentation and

related subjects we may return to the question which necessitated it,
whether the contrast he drew between the two facets of his teaching,
"theological"and physical, was of epistemic character. We know now
that Xenophanes used inferential proofs and these proofs were not
accidental, but ratherconstitutedthe very texture of his monistic con-
ception. Now this inferential or almost inferential doctrine was
intended as true, as its exemption from the scope of the sceptical atti-
tude shows and as Epicharmus' testimony seems to attest, while the
cosmological account was professedly declared to be not true but only,
at best, a plausible opinion, since the conclusions arrived at in this
quarterlack certainty-they are possible but never the only ones possi-
ble, that is, never necessary; it follows, then, that the conclusions
obtained in the "theological"sphere were regardedby Xenophanes as
certain,122that is, the only possible, or necessary ones. This means that
he was aware of the apodeictic characterof his inferences,123and we
must therefore recognize that he in effect distinguishedbetween apo-
deictic and non-apodeictic knowledge.124The truth of Xenophanes'
monistic doctrine is thus guaranteedby the very fact that it is attained
throughnecessary inferences; physical speculation, on the other hand,
is not apodeictic and because of this requires empirical verification
which, however, is impracticable,reducing this kind of knowledge to
no more than opinion.125
(Munich 1956) 2. 385; cf. Jaeger, loc. cit.) or a rational"theologist"(see Reinhardt,loc.

cit.; Barnes, loc. cit.) is somewhat artificial:the mystical intuition may well be followed
by conceptualelaborationand argument,-pace Kirk,Raven, Schofield (above, n. 7) 165.
122Cf. Deichgriber (above, n. 6) 30; Gigon (above, n. 24) 19; von Fritz (above, n. 14)
1557-1558.
123Cf. Deichgriber (above, n. 6) 29: "Die Pridikate, die der Gott erhilt, ergeben sich
mit der im nrpatxov gelegenen Notwendigkeit." Yet as the first argumentin Ps.-Plutarch's
account shows, it would be wrong to restrictXenophanes' argumentationto the "logisch-
explicative Methode"(ibid. 30).
124My conclusions concerning the logical standing of Xenophanes' monistic doctrine
and the epistemological opposition between it and the cosmology are close to those of
Deichgriber (above, n. 6) 29-30.
125Lumpe (above, n. 8) 35, is therefore wrong in concluding that Xenophanes'
"Methodeist ... weder rein empiristisch noch rein rationalistisch";as a matter of fact,
Xenophanes uses two different methods in his monistic and his cosmological doctrines.

The next question is how Theophrastus defined, if at all, Xeno-

phanes' epistemic approach. The relevant sentence in Ps.-Plutarchruns
as follows: &artopaiveTateictaitag aio0rojet; esV~iE Icaci
icaO•6ko
aov aiarSi rai a3byvb6v 6yov tacld4Xt,-"Xenophanes says that
senses lie and together with them he discredits on the whole the logos
itself." The reportas it stands is undoubtedlywrong and seems to have
come from inferior Sceptic sources. Yet this can hardly be the case.
First, the section of the Stromateis concerned with Xenophanes has
thus far proved itself a source not contaminated with alien, non-
Theophrastean, doxographic traditions, and we need not make an
exception for a single phrase, the more so that it does not display any
specifically Sceptic feature.126Secondly, the location of the sentence in
the account, after the exposition of the monistic doctrineand before the
reporton Xenophanes' cosmology, is undoubtedlycorrect and points to
the original purpose of the phrase, namely the determinationof the
epistemological standing of the "theological"doctrine in Xenophanes'
teaching. We thereforehave every reason to believe that the statement
did not come from any other source but is yet anotherexample of the
garblingof Theophrastus'wording in Ps.-Plutarch.
Now supposing that Theophrastusreported Xenophanes' epistemo-
logical approach, how did this report read? In the passage from
Metaph. 5 discussed at the beginning of this paper Aristotle puts the
Parmenidean opposition between Aletheia and Doxa as follows:
"...but being forced to follow the observed facts, and supposing the
existence of that which is one in definition (iarax b6v but more
,6yov), he now
than one according to our sensations (icarx ri~vai'O0rltiv),
posits two causes ... fire and earth ..." This is to say that in Aletheia
Parmenides follows logos while disregarding aisthesis, whereas in
Doxa he is guided by aisthesis while abandoninglogos. This was also
Theophrastus' view, as appears from Ps.-Plutarch's report that Par-
menides tz; aio0GioEtS3KEP *K tig ag,-"throws out
,lXEt &,rl9ei
Nevertheless Lumpe is much closer to the truth than those critics who, as Frinkel and
Reiche, try to reduce Xenophanes' approachto a "rigorousempiricism"(Frinkel [above,
n. 73] 130).
126Pace Barnes (above, n. 17) 137, who ranges Ps.-Plutarch'sreportwith Sotion's (cf.
also Lumpe [above, n. 8] 32),-compare Sotion's wording (ap. Diog. 9.20 = Diels-Kranz
[above, n. 9] 21 A 1, above n. 70) echoed in Hippol. Haer. 1.14.1 (= Diels-Kranz [above,
n. 9] 21 A 33).

162 AryehFinkelberg
sense-perceptionfrom the truth."127 Now what applies to Parmenides'

Aletheia must also apply to Xenophanes' "theology": Xenophanes, in
Theophrastus' view, was the inventor of the "way of the eternally
selfsame whole" followed by Parmenidesin Aletheia, and in his "theol-
ogy" Xenophanesalso largely resortedto logos, that is, a priori deduc-
tions. It follows that Theophrastusmust have said that Xenophanes
followed logos while disregardingaistheseis. Now this is the formula-
tion found in Aristocles,128who classes Xenophanes with the Eleatics
and the Megarians, describing all of them as thinkers who disregard
aistheseis and relied on the logos alone. But why conclude that Aristo-
cles' characterizationof the epistemological position of a number of
thinkersgoes back to Theophrastus'descriptionof Xenophanes' indivi-
dual position? Let us examine the wordingof Aristocles' report:
oi'ovrat yap 6Eiyv r&;aRv Ps.-Plut.Strom.4: atoqpaiverat

a
aiaoroetl; cai Taxgpavaoia(; 68'Kaici
rz;gaia,0Aa& \jIg tSIg
al'T 68 ~E6vov Kai ia06xo'uOav alzaigi Kai
ca-ra•3&rxxtv, V
To) X6yTp_nttoyEIitv" otalza aplrv Thv k6yov8tapac"Lxt.
ydp tva lrp rEpov Kai
•Riv E. ial
al ZIivyov
"
HapptEvi86rli
-
MXtioao eyov, i9oaEpov 8' oi
nEpi XzriXnova icai tot; ME-
yaptco't;. 08Ev oLTol Hippol. Haer. 1.14.2: k*ytt 6~
iotvv
YE O OVEVElvatZal 'TO 0rt o0i6v
y•veFat o0- cE0eipe-
b EpovV
EdVvat
rat o riatKaizt&riv TO
Rtil tnrl•p 7yEvv&a( tI ltvE
Rn8& E'(pOIpEacxiatRi6lE tIVEi- nav FOatvKrx.
6
aoat T'napdanIav.
If the resemblance between Aristocles' and Ps.-Plutarch'swording is

not enough to ascertain their common descent, then the additional
127Ps.-Plut. Strom. 5 (= Diels-Kranz [above, n. 9] 28 A 22). Kahn (above, n. 28) 24,
notes that "the two sections of Parmenides' poem are distinguished [by Theophrastus]
Ica' jRiv and jara
86t av 6& Oi)vnoh according to the very words of
Parm.&dXAO•tav
B 1.29-30 ... Aristotle's own distinction in yov,
terms of X6yoq and ai'atOriot or
qatv6gieva (Vors. 28 A 24-25) is less accurate,for the terms are his own ratherthan
those of the writerin question." Yet having describedthe distinctionin Parmenides'own
termsTheophrastusinterpretsit in termsof X6yoqand ai'a0loOnq.
128 Flepiptlooopiaq if ap. Eus. Praep. Evang. 14.17.1 (= Diels-Kranz [above, n. 9] 21
A 49).

parallel with Hippolytus finally proves this. Aristocles, that is, here
draws on the Theophrasteandoxographyof Xenophanes (in all proba-
bility, on the "Parmenidized"version, as the parallel with Hippolytus
suggests-see Table 5) applying what is said with reference to Xeno-
phanes to a whole group of thinkers.
We can thus conclude that Theophrastusdid define Xenophanes'
epistemological approach and that his definition was exactly as we
would expect it to be-"he throws out aistheseis, while trusting logos
alone." The definition is loose and misleading, for it implies that, in
contrast to the "theology," the "physics" is "accordingto aistheseis";
yet the actual contrast is not that between two kinds of cognition,
purely intellectual and that based on the senses but that between two
kinds of inferences, a priori demonstrationlabelled certain knowledge
and a posteriori inferences which, when unverifiable, are but fallible
guesses. 129
Hereour reconstruction
of Theophrastus'
accountcan be regarded
as completed: we have determinedhis generalapproachto Xeno-
129 We can now see whether, as is usually believed, Xenophanes held that the sense-
perceptionsare unreliable. Comparingthe reportin the Stromateiswith thatof Aristocles
we observe that Ps.-Plutarchnot only completely misrepresentsthe second partof Theo-
phrastus' statement,but also glosses its first part as "senses lie," and thereforewe must
not take these words as suggesting that TheophrastusreportedXenophanes'distrustof the
reliability of sense-perceptions. Theophrastus'"throwingout aistheseis, while trusting
logos alone" must also be Aetius' reason for listing Xenophanes together with ten other
philosophers under the title "the senses lie"; and indeed, the names listed (AMt.4.9.1 =
Diels [above, n. 13] 396; cf. Diels-Kranz[above, n. 9] 21 A 49) suggest this interpretation
ratherthan the unreliabilityof sense-perceptions. The doxographicmaterialat our dispo-
sal thus provides no evidence that Theophrastusever said that Xenophanes claimed the
unreliability of our sense-perceptions, and the fact that the Sceptics did not explicitly
ascribe this view to Xenophanes furtherproves our conclusions. We are left with Xeno-
phanes' fr. 38 which is usually interpretedas stressing the relativity of our perceptions
and our empiricaljudgments (e.g., Frinkel [above, n. 19] 430; G. Rudberg,"Xenophanes,
Satirikerund Polemiker,"SO 26 [1948] 133; Guthrie [above, n. 8] 1.401). This is the
possible purportof the lines, but in the absence of the broader context it is not clear
whetherthis was the intendedone. At any rate, it is evident that the idea of the unreliabil-
ity of the senses, if articulatedby Xenophanes at all, was casual and peripheric in his
thought.

164 AryehFinkelberg
phanes' twofold teaching and dwelt on the main points of his reporton
Xenophanes' monistic doctrine; the examination of Xenophanes'
cosmological conception, however interesting and desirable, is a
separatetask which should be left for anotheropportunity. After such
a lengthy discussion one should perhapsbriefly recapitulatethe results
arrivedat, and the best way to do this seems to be to presentthe Theo-
phrasteanaccount in the form of the orderedseries of statementsrecon-
structedabove. In this list the sources from which a given statementis
excerpted or on the basis of which it is formulatedare referredto by
the name of the authorand the page and line(s) of the Diels-Kranzedi-
tion; statements and parts of statements which are purely conjectural
are italicized.
i. [Ps.-Plut.;122.15-18] But Xenophanesof Colophonwho pursued

a certain way of his own different from [thatof] all those spoken
of beforehand[sc. the Milesians] allows neithercoming-to-be nor
destructionbut says thatthe whole is eternallyselfsame.
ii. [Simpl.; 121.28] He says that this One and Whole is God, saying
thus-fr. 23.
iii. [Simpl.; 121.27-28] The mention of this Xenophanean opinion
ratherbelongs to a study other than that concerned with natural
philosophy [thatis, in thatconcernedwith firstphilosophy].
iv. He says that God is ungeneratedand eternal which he proves as
follows: [on the basis of Ps.-Plut.; 122.18-20] had it [the Whole
or God] come to be, it is necessary for it not to be before this; but
not being, it can never come to be: neither nought can produce
anythingnor anythingcan come to be by the agency of nought.
v. That God is one he proves so: [on the basis of Ps.-Plut.; 122.
23-24] among gods there can be no supremacy,for it does not
suit the divine holiness that God should be under lordship; but
were there many gods, there would be lords and subjects among
them (perhapsalso: or all of them would be lords of each other).
vi. [on the basis of Simpl.; 121.24-25; 122.3-6] He does not say
whetherGod is finite or infinite,
vii. nor does he say [on the basis of Simpl.; 121.25; 122.3-6] whether
he is moved or unmoved,
viii. but [on the basis of Simpl.; 122.6-9] actually he conceives of
God as unmoved,for he calls him eternally selfsame and says-
fr. 26.
ix. He says that God is thoroughly seeing, hearing, and thinking-

fr. 24; he demonstratesthis in the following way: [on the basis

of Ps.-Plut.; 122.25-26] God is altogether free from any want;
but had he seen, heard and thought only in one part of him he
would be in want of these in anotherpart; hence he sees, hears,
and thinks wholly and not in one or anotherpartof himself.
x. [on the basis of Simpl.; 122.13-14] And he says that God
governs all things by his mind, saying-fr. 25.
xi. [on the basis of Ps.-Plut.; 122.20-21; Aristocles; 126.6-8] Thus
he throws out sense-perceptions,while trustinglogos alone.
xii. [on the basis of Ps.-Plut.; 122.15-18; Theophr. ap. Alex.;
219.31-33] The other way, that of accounting for coming-to-be
of existing things he dismisses declaring such accounts to be no
more than opinion deprived of any certainty saying this in such
words-fr. 34.
xiii. Nevertheless he proposes some such opinion which he himself
seems to adjudge looking plausible, as his own words show-
fr. 35.
xiv. [Theophr.ap. Alex; 219. 31-33] But Parmenideswho came after
him took both ways [sc. that of Xenophanes and that of the Mile-
sians, cf. (i)]. For indeed, he both says that the whole is eternal
and tries to account for coming-to-be of existing things not how-
ever thinking about both [ways] alike, but according to truth
assuming the whole to be one, ungenerated,and spherical,while
according to the opinion of the many as to accounting for the
coming-to-be of perceptible things, positing two principles, fire
and earth,etc.
The reconstructedaccount represents that of the first book of the

Physical Opinions. Indeed, (i) is the counterpartof (xiv) which is
explicitly related by Alexander to the first book; (ii)-(x) also belong
there, for they either come from Simplicius' reportor correct and com-
plement it where it is wrong or incomplete, while this report itself
comes from the first book of the Physical Opinions. That the account
of Xenophanes' argumentswas given by Theophrastusthere is further
supportedby the fact that the Theophrasteanpr6cis of the Parmenidean
argumentwas found by Alexander in the first book.130Finally, the fact
that in (xiv), which is a quotation from the first book, Theophrastus
130 Simpl. Phys. 115.11 (= Diels-Kranz [above, n. 9] 28 A 28).

166 AryehFinkelberg
outlines the respective standing of Parmenides' Aletheia and Doxa,

suggests that the statements (xi)-(xiii), where Xenophanes' epistemo-
logical approachis reported,also belong here.
If Theophrastus'account was as I suggest, it seems to have been of
great accuracy. True, it misrepresentsXenophanes' position in that his
epistemic approach is interpreted in terms of the contrast logos:
aistheseis, but this is the only majormisinterpretationI can find in the
account. On the whole, this is a precise report which moreover does
not show any tendency to assimilate Xenophanes' teaching to that of
Parmenides,a tendency which, as we have seen, is easily discernablein
later doxographic writers. Theophrastustakes Xenophanes' text sim-
ply and literally and reports it as it stands, accurately and precisely,
stressing the specific and indicatingonly the most general parallelsand
most apparentlines of historical continuity-as a matter of fact, he
mentions only two: thatParmenideswas Xenophanes' "pupil"and that
he adoptedin his Aletheia the "way" initiatedby his "teacher"combin-
ing it, in Doxa, with the "way"of Xenophanes' Milesian predecessors.
Yet it would be hard to point out even one importantParmenidean
doctrine which is not, in one or another way, rooted in Xenophanes'
teaching. Such is, firstand foremost,the Parmenideanidea of the intel-
ligible unity of the sensible manifold which in Xenophanes himself
was, as we have suggested, the development of one of the facets of
Anaximander'sApeiron; more specifically, this is the view of unity as
one of two aspects, true, the most essential, significant, and sublime,
but nevertheless one aspect only, of reality, complementaryto its other
aspect, that of the manifold. The inferential form of Parmenides'
Aletheia was obviously inspiredby Xenophanes' attemptto present his
monistic doctrine as deductive, and we have seen that the order of the
argumentsin that part of Aletheia where the attributesof the One are
deduced (fr. 8) follows that of Xenophanes' "theology." Again, Par-
menides' contrast of Aletheia and Doxa as necessary and possible
accounts of reality has roots in Xenophanes' distinction between apo-
deictic and non-apodeictic kinds of knowledge.131We may go on to
point out that the main attributesof Parmenides' One-its eternity,
oneness, unchangeability,and, to an extent, perfection-were already
131 The importanceof Xenophanes' distinction between knowledge and opinion for the
subsequentdevelopment of Greek thought is rightly stressed by Guthrie(above, n. 8) 1.
399.

established, and most of them demonstrated,for his One by Xeno-

phanes; that Parmenides' terminologicaluse of doxa, and perhaps also
of logos,132 can be traced back to Xenophanes; and that it is not
improbablethat the very idea of the unity of all things as that of Being
was suggested to him by Xenophanes' argumentfor God's ungenerat-
edness. To exhaust the list, we should also mention Parmenides' con-
cept of the universe as the intelligent sensorium,133and his dualistic
physics, two ideas of undoubtedly Xenophanean origin.134And if we
add to all this the Xenophanean thought-patternsoperative under the
surface of Parmenides' poem, the depths of which we have had occa-
sion to catch a glimpse, we shall be in a position to appraise what is
implicit in the lapidaryAristotelianremark"Parmenidesis said to have
been his pupil."
132 See, however, von Fritz

(above, n. 14) 1559, but also H. Fournier,Les verbes 'DIRE'
en grec ancien (Paris, 1946) 55-57, 217-221.
133This doctrine, as I argue in "'Like Like' and Two Reflections of
by Reality in Par-
menides," Hermes 114 (1986) 405-412, accounts for two alternativeways of cognizing
reality, as Being and as the mixtureof fire and night.
134The dualistic pattern,at least in rudimentaryform, is found in the Milesian cosmo-
gonists (Ps.-Plut. Strom. 2 = Diels-Kranz [above, n. 9] 12 A 10; Hippol. Haer. 1.7 =
Diels-Kranz [above, n. 9] 13 A 7). Yet while adopting and furtherrefining the Milesian
view of cosmic process as propelled by the interactionof two main cosmic opposites,
Xenophanes introducesa significantnovelty: he was the first to see the interactionof the
pair of the opposites, at least in part, as their mingling (Hippol. Haer. 1.14.5 = Diels-
Kranz [above, n. 9] 21 A 33; cf. fr. 29, 33). It is precisely this idea, this time in more
developed and universal form, that operates in Parmenides. According to him, the world
is constitutedby the pair of opposites, fire and night, which partly are in the mixed state
(fr. 9), while all things-the animal body (fr. 16), water and air (Arist. Gen. Corr.
330b13 = Diels-Kranz [above, n. 9] 28 A 35), the heavenly bodies (AMt.2.7.1; 20, 8a;
3.1.4 = Diels-Kranz [above, n. 9] 28 A 37, 43, 43a) are but various mixtures of fire and
night; the mingling of these was the way in which the developed world came into being
(fr. 13) and is the way in which it is maintained(fr. 12. 3-6).


Department of The Classics, Harvard University

Uploaded by

Copyright:

Available Formats

Department of The Classics, Harvard University

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Department of The Classics, Harvard University

Uploaded by

Copyright:

Available Formats

Department of the Classics, Harvard University

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

T seemsconvenientto begin our discussion of Xenophanes' teach-

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

notice of many critics, though it is explicit in Aristotle. I mean the

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

Turning now to Aristotle's account of Xenophanes, we should note

7 Notably by Kirk in G. S. Kirk, J. E. Raven, M. Schofield, The Presocratic Philoso-

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

ready to allow Xenophanes as one of the "Eleatic tribe,"9Aristotle

9 Sophist 242D (= H. Diels, Die Fragmente der Vorsokratiker,5W. Kranzed. [Berlin

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

phanes' lack of clarity as referringto the lack of specificationof God's spatialcharacteris-

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

Thus, not finding in the "theological"context evidence for whether

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

Aristotle's attributionof unchangeability to God can be checked

Always he remainsin the same [place], not moving at all,

It has been suggested that dvqonagis to be taken here in its "philo-

The fact that God's accomplishmentsare effortless entails his absolute

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

view of God. The consensus is indeed impressive: including all our

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

32"The Milesian Monistic Doctrine" (above, n. 17). The

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

Turningnow to Theophrastus,let us begin with his general approach

Theophrastussays that Xenophanesof Colophon, the teacherof

Some critics take these words to mean that Theophrastus,presum-

Some removed generationand destructionfrom the world alto-

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

they belong to anotherand higher inquirythan the study of nature.

In criticizing the Eleatics Aristotle proceeds here from his principaldis-

39Cf. Ph. 184b25; 185a12.

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

Being,42but so far it is still unclear with which kind of unity he counts

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

oi03 icvEi^zatseems to be superfluousin view of 50 tRetapooXfig in the

45 Cf. fr. 14.

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

5 datov [dIyvriltov: 8vFva CvV neque natum

12 Kati taot adeta-

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

... ei 6 kov o-pavbv

49 & )o;opav6; doesnotheremean'sky' (in which W. P6tscher,"ZuXenophanes,Frgm. 23,"

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

doxographicadditionafter the patternof Parmenides'Being.5o

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

mind is that Theophrastus,who, as distinct from Aristotle, took into

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

predominanttendency.54When homogeneity and sphericity penetrated

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

applicablealso in the lattercase.

This content downloaded from 128.235.251.160 on Fri, 26 Dec 2014 05:58:06 AM

Theophrastus Simplicius 'Parmenid