The Unmoved Mover in Early Aristotle

Author(s): H. J. Easterling
Source: Phronesis, Vol. 21, No. 3 (1976), pp. 252-265
Published by: BRILL
Stable URL: http://www.jstor.org/stable/4181995 .
Accessed: 23/02/2015 14:13
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 support@jstor.org.

BRILL is collaborating with JSTOR to digitize, preserve and extend access to Phronesis.

http://www.jstor.org

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions

Moverin earlyAristotle
TheUnmoved
H. J. EASTERLING

It

is still worth discussing whether the Unmoved Moveris present in

the early stages of the development of Aristotle's thought. Jaeger

thought that the doctrine of the Unmoved Mover was formulated
by Aristotle at the beginning of his independent career and that it was
to be seen as the successor to the Theory of Forms'. Perhaps few
scholars would subscribe to this extreme version of the thesis today,
but there is currently considerable support for the belief that the
doctrine was an early one and that it appeared in de Philosophia2.
There has been a good deal of interest in the subject recently3, but no
account has been taken in recent discussions of the references to the
Unmoved Mover (or to an unmoved mover) in the earlier books of the
Physics, which I wish to consider here. To anticipate my conclusions, I
intend to argue that the passages containing these references, which
occur in Physics B and F, are later additions and that the doctrine
of the Unmoved Mover did not figure in the original version of these
books.
As there can be no doubt about the presence of the Unmoved
Mover as a central topic in Physics (, the intention to treat the early
books separately may need some justification. Though it is hard to be
dogmatic about the structure of the Physics, it is perhaps safe to say
that the proof of the Unmoved Mover's existence, which occupies 0),
forms the culmination of the discussion of change that begins in E and
W. Jaeger, Aristotle' (Oxford, 1948), trans. R. RZobinson,p. 219.
So, for example, E. Berti, La Filosolia del primo Aristotele (Padova, 1962), pp.
344-357; M. Untersteiner, Aristotele, Della Filosolia (Roma, 1963); I. During,
Aristoteles (Heidelberg, 1966), pp. 185-189.
3 See especially J. P6pin, 'L'Interpr6tation du de Philosophia d'Aristote',
(R.E.G. 77 (1964) 445-488); K. Gaiser, 'Das zweifache Telos bei Aristoteles' (in
Naturphilosophie bei Aristoteles und Theophrast, hg. von 1. During, Heidelberg,
1969, pp. 97-113); W. P6tscher, Strukturprobleme der Aristotelischen und Theophrastischen Gottesvorstellunzg (Leiden, 1970); B. Effe, Studien zur Kosmologie und
Theologie der A ristotelischen Schrift 'Uber die Philosophie' (Zetemata 50 (1970));

A. Graeser, 'Zu Aristoteles peri philosophias (Mus. Helv. 27 (1970) 16-27);

A. Graeser, 'Aristoteles' Schrift Uber die Philosoplie und die zweifache Bedeutung der causa finalis'

(Mus. Helv. 29 (1972) 44-61).

252

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions

continues in Z4. Leaving out of account the puzzling book H, one can
say that EZO, whether or not they were originally written as a single
unit, certainly form one now, with H awkwardly intervening, and
that they are so regarded by Aristotle, who refers to them as 'a' 7tepl
XLIvaew)5. Similarly ABUrA
form a loose unit, also recognisedas such by
Aristotle under the title

7cep'typuaec

or

&

(puaLxa6;this unit is

clearly earlier in date than EZ?7. Thus it seems a justifiable procedure

to examine ABFLAas an independent unit early in date.
The proof of the Unmoved Mover's existence appears in E),and
it is natural to ask whether this was the place where Aristotle first
announced it as a new doctrine, or whether he is here setting out a
proof to support a doctrine already formulated and enunciated elsewhere. At first sight the latter explanation might appear correct, for
there are several referencesto an unmoved mover in the earlier part of
the Physics. I now wish to examine the passages in which these occur.
There are five in all:
For convenience I will take first the three passages from the discussion of xLviwcm
in F 1-3, which hang together.
1.

aOX?L pd.V OUNV'TLCLVOC&V XLVCZa4CL TO XLVQOV,Qui [L7)V XXX& 7CpCLTOuTOU

E'V
'CiXXWV

eiU

8n,O'V 67r4

'/eL

(gaTt yOp TL XLVOUV XOcOxLvoV).

201 a 25-27
After some preliminaries, r 1 begins by defining change as the
actualisation of the potential (a 11), giving examples (a 16). Aristotle
then makes the point that some things can be both actually and
potentially, and thus can stand both as agent and as patient. This
applies to To xLvo5v qpuaLXW,
which is %Lv7qTov
as well as xvv-rtsx6v(a 24),

because it produces change while being changed itself. A digression

(a 25-27) then remarks that there is another kind of mover (viz. unmoved), though some thinkers fail to realise this.
2. xweVcoctL
XOct

Oi

xatL so

axwVatL' npqao

xLvo5v
ea-.Lv(V

'Oaitsp

XepraxL 7stWv, TO\ aUvo4ta

y&p I XVaCL

u7rtpxel,

OV XLVT6V,

OUTOou XxLvacta

202 a 3-5
This is merely the same point as that made at 201 a 25-27, appearing
again in the summary at the end of the chapter. Every mover, says
4 This view of 0) is perhaps an over-simplification; cf. G. A. Seeck, Nachtrage im
achten Buch der Phtysik des Aristoteles (Abh. Mainz, 1965. 3).
5 e.g. deCaelo 272 a 30, 275 b 22.
6 e.g. Physics 251 a 8.
7 For a fuller discussion of these references and of the relative dating, see W. D.
Ross, Aristotle's Physics (Oxford, 1936), pp. 1-19.

253

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions

Aristotle, will itself be in motion (i) if it is potentially xLvo6v and (ii) if

its previous lack of movement was 'psurax,rest (i.e. temporary absence
of movement), rather than absolute negation of movement. This
implies that there are movers that are not potentially xtvryov, and
whose nature is incompatible with movement.
3. el)E 8' &atopLocvXoyLxrJv avyxoF.ov yxp 'ac
tou
e'VepyeLoev
VOu tLVOc
7COL7=LXOU XOCLrou

E'X/o 705 V?'

rcVL;

7rOcLM[a, to)

yap

Lty)

r6
7

7T-7rLXOl

eV

8C 7

vo

tCXOVaoT

Y; 87?

C.

&-

t6
ro' 7E

XOL XLVGUpLev(d,

E`th

, :pYOV 8? xat

, ECRU?V I t
e V notLGq

V
eV t&

xav,
O Yap
~V"-tL at'ou( y&f.p
7' X LvYaL
V
e V TcepXLVOU
t:5 7VE no XtV0ovXLV CErcu,
xLvourvoou),

ev TCO 7tCa(OVTt
7COLo5vVt,
1 ovSro,
. att pnv
ocucovvc,av ej a'f7ec,
oX
7 V.e'tOt)tO' ,
OOCV
c E7)
k 'a ~.L
OULC)VUULO4

uro6 Xo6yoq &S7LXLVouvroq XaL

LoIlo,

StSEL 01V c/4Lp(G)XLV1fjG

7tCOaLv
es 8tL XC TOCUTar-V
XcrL
xoCl XIVVV
?Vz
Xt8CL
L

202 a 21-31
F 3 deals with the relation between mover and moved. It motion is the
actualisation of the potential, where does this actualisation take place?
Is it in the agent or in the patient? Aristotle's answer is that it is in the
patient (a 13-14). But this raises a difficulty (a 21). There are really
two actualisations, that of the potential agent and that of the potential
patient. Do they both take place in the patient (alternative A) or is one
in the agent and the other in the patient (alternative B) (a 25-27)?
Aristotle examines these two possibilities in turn.
I

oU ztvyatroa.
zZOV xLVYnaLv

1. Take alternative

't)
B, that n teV noLLag eV

-l
noto5vwn,

8? iaSCt,O

EVJTC) wa7rocovwn.

This entails that there will be movement in the mover (a 29),

which in turn leads to one or the other of two alternatives,
either 1.1 every mover will be moved;
or 1.2 something that has movement in it will not be
moved.
These are both impossible consequences and so by this reductio
ad impossibile alternative B can be rejected.
2. Take alternative A, that both are in the patient (a 31). Various
difficulties result from this (a 33-b 5), but these can be resolved
(b 5-22).
In this argument, the section relevant to my present discussion is 1.1,
which shows that Aristotle implicitly rejects as impossible the suggestion that every mover is in motion. I.e. he believes that there is such a
thing as an unmoved mover.
In these passages, then, Aristotle clearly mentions or presupposes the
-lile passages 2 and 3
existence of some kind of unmoved mover. And wh
might be dismissed as referring merely to some unspecified kind of
254

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions

unmoved final cause, it is nevertheless possible to say that passage 1

must refer to the Unmoved Mover, i.e. the prime mover of the universe;
Aristotle promises a future discussion of the subject, which can only
mean Physics E) or Met. A, so that he is clearly thinking of the Unmoved Mover here.
I now wish to ask whether these references to the Unmoved Mover
are firmly anchored in their contexts, or whether they show any signs
of being later additions. I assume for the purpose of this discussion
that the Physics can be treated as an original basic structure to which
notes were (or may have been) added later; it is possible to do this in the
case of Physics r because we have a control in Met. K.
The discussion of change in F 1-3 is reproduced in Met. K 9. The
version of Met. K is clearly later, because it reproduces the other version
in a briefer and more concise form; e.g. it omits unnecessary examples,
adopts simpler forms of words, and in general simplifies any elaboration. The version of Physics F is reproduced completely in Met. K,
with the exception of four passages:
(i) 200 b 28-32. A note, not very relevant to the context, on relative
terms. Aristotle has just pointed out that the distinction Kuvoc?rq/
evre)S?iXCL, which makes change possible, occurs in all categories. He
goes on to remark in b 28-32 that while XLVYCatq
thus takes place in all
categories the elements involved in change, viz. the xmvyov and the
themselves fall in the category of relation.
xLvvLqx0v,
(ii) 201 a 19-27. A parenthetic note to point out that some things

are capable of interaction and so can be both

XLVOVVcMand

XLwou[Lva.
(iii) 202 a 3-12. This paragraph looks at first sight like a summary
re-stating the point made earlier (in 201 a 23-25) that a natural mover
will always itself be in motion, but in fact it goes further than this,
introducing the new point that such a mover always has an eloq
which is its apxZxxvrCam (in the sense that e.g. a builder's possession
of the form of house enables him to build houses).
(iv) 202 a 21-29. This is the oinop'oa
concerning the localisation of the
two ZVepyetLa; it arises out of the short paragraph on the relation
between mover and moved, and was discussed above in relation to
passage 3.
These are the only passages of F 1-3 that are not represented, even in
summary form, in the version of Met. K. Their omission may of course
be pure chance, or it may be because the excerptor (whether Aristotle or
another) thought that they were not essential to the summary that he
was making - and it is perfectly true that they are not. On the other
255

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions

hand, this does not seem to fit the excerptor's method; he does not
systematically omit inessentials, but reproduces almost word for
word a number of passages that are no more essential to the general
scheme than the four that are omitted, e.g.
201 b 5-15. A summary of points made in the preceding chapter,
with a detailed example to illustrate them.
201 b 16-202 a 2. Some rather disjointed and rambling remarks,
mainly about other philosophers' views of change and their shortcomings.
In view of this it seems unlikely that the four passages were omitted
merely because the excerptor was systematic and thought them inessential; he does not seem to work in this way. A more probable
explanation is that these four passages did not appear in the original
version of Physics F from which he was working.
This suggestion is supported by a closer examination of passage (ii).
This passage, unlike the other three, is not entirely omitted from Met.
K, but is represented by a single sentence:
, xcL oure npo6rspovouo
S'orv I ev)Xe zerm
3U[PML"V or xcvLXtVeLa43acL
va'rpov. (1065 b 20-21) As was first pointed out by Diels8, this sentence
does not at all convey the sense of the passage that it represents (201 a
19-27). This can be explained on the assumption that 201 a 19-27 as it
now stands is a later amplified version of something that originally
answered more closely to what we have in Met. K. There are two
points that support this explanation:
1. In the recapitulation at the end of F 1 (201 b 5-15), Aristotle
claims to have demonstrated, and gives a summary of, not what now
stands in the text of the chapter, but what we have in Met. K. At 201
b 6-7 he uses the exact words of Met. K, although in the present
version of Physics F this point has not been mentioned.
2. As Diels observed, the words that follow the suspect passage at a
27, a ' o058uv04LEOvtoq,do not follow on what now precedes them.
Moreover the word ivr?Xk(La, which is supplied by all editors and is
necessary to make sense of the sentence, does not appear in any MS of
the Physics. It is worth noting that it does not appear in the version of
Met. K, and does not need to appear there since it is easily supplied
from the previous sentence.
I would argue, therefore, that these four passages as they now
8 H. Diels, 'Zur Textgeschichte der Aristotelischen Physik' Abh. d. k. Akad. d.
Wiss. zu Berlin, 1882, Philos.-hist.

classe, Abh. I, pp. 34-35 = Kleine Schriften,

Hildesheim 1969, pp. 230-1.

256

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions

stand were absent from Aristotle's original version of Physics F. This is

striking because, as will be observed, the references to an unmoved
mover considered above are all contained in these suspect passages.
There are two passages to be consideredin Physics B. I wish to deal
first with B 7.
This chapter falls into two sections: 7.1 (198 a 14-21), which summarises the doctrine of the four causes, and 7.2 (198 a 21-b 9), which
contains a series of rather disjointed remarksabout the four causes and
the physicist's interest in them. There are two odd inconsistencies
between these two sections.
1. The main subject of 7.2, which constitutes the bulk of the
chapter, is the relevance of the doctrine of the four causes to the
physicist; Aristotle argues that the physicist must take account of all
four causes, i.e. that all four must be invoked in the explanation of
Ta c(SaL. As the text stands at present, 7.1 forms a link leading on from
the earlier discussion of the causes to 7.2; if Aristotle intended it so, we
should expect the examples of causation given in 7.1 to look forwardto
7.2 and to be relevant to the physicist and his subject. But it is odd
that in 7.1 Aristotle chooses mathematics, an example drawn from TOC
to illustrate the concept of formal cause, and in fact appears
OxMvy17m,
to be saying that formal causes are restricted to the sphere of &xLvvqtx9.
The use of a mathematical example is not odd in itself (a similar
example was used to illustrate the formal cause when it first appeared
in B 3, at 194 b 28), but it is odd that Aristotle appears to restrict
formal causes to this type; more odd still is the fact that this nonphysical example is used in a summary which, as the text stands,
leads up to Aristotle's discussion of the physicist and his interest in the
causes. This suggests that when 7.1 was written it was not designed to
lead on to 7.2, but that the latter was added at a later stage.
2. The parenthesis at a 27-31 begins by mentioning unmoved movers
and digresses from this to remark that they are outside the scope of
physics and form the subject matter of a different science. Three
sciences are then listed, which are differentiated by their subject
matter, viz.
l Of course, as Ross remarks ad loc., Aristotle cannot mean to say this, because
it is obvious that the formal cause is relevant to natural objects; Aristotle must
be taken to mean that MxivrrTa provide a particularly clear illustration of formal
causation (because the other three causes do not, in the nature of the case,

operate

'v -oZq
t'CxMLVoL,

so that there is no danger of confusing

the formal

cause with any of the others).

257

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions

(1)

(2)

7tCpt &XLVT6V,
h

(3)

7reptxwovu4Lv)v ,L&vdIvpprwv

&U,

7
NeplTX
X,?&pr4.

Of these three, (1) is clearly 7prIn (pL?oaocpqx

and the other two are subdivisions of physics, (2) being astronomy and (3) the study of the
sublunary world.
It is worth noting that this trichotomy does not occur elsewhere in
the Physics. In B 2 Aristotle discusses the sk--pe of physics. His main
point is the distinction between physics and mathematics, while a note
at the end of the chapter adds first philosophy as a third science
(194 b 14-15). Thus B 2 may be said to recognise a trichotomy, but it is
a different one from that given in B 7, viz.
(1) physics
(2) mathematics
(3) first philosophy.

This trichotomy appears again in Met. E (Chapter 1, esp. 1026a

18-19) and K 7 (1064 b 1-3); it also underlies the discussion of Met. F 3
(though it appears more clearly in the summarised version of F 3 that
is given in K 4, (1061 b 17-33)), and it is clearly a common Aristotelian
schematisation (cf. also E.N. 1142 a 17-20, de Anima 403 b 7-16)1O.
The trichotomy of Physics B 7 is not found in any of these passages;
astronomy is usually included as a part of physics (cf. Physics B 194 a
7-12, Met. M 1078 a 14-20). It corresponds, however, exactly to the
trichotomy given at the beginning of Met. A, where the three sciences
are again differentiated by their subject matter. At 1069 a 30, Aristotle
lists three sciences that study three different kinds of o'uaLx,viz.
(1) &X1V-nTO
oCatX,
(2) o6alocxda5nrr xxlt OU&oq,
(3) oija:c tatxxhjrh xat cx&psi.

(2) and (3) are here said to be sub-divisions of physics, but mathema-

tics does not appear as a third science; the only mention of

hcjrex'nxa

ot6aUo. Thus
here is as possible candidates for classification as MxLvThro4
the presuppositions that lie behind the trichotomy of sciences in
Physics B 7.2 seem to be in sympathy with Met. A, and not with the
earlier part of Physics B. This again suggests that B 7.2 is not part
of the original version but is later in date than the rest of B"1.
10

For a fuller discussion of this subject see P. Merlan From Platonism to Neo-

platonism'

(The Hague, 1960), pp. 59ff.

258

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions

Some scholars have suggested that particular sections of B 7.2 are

later insertions. Thus von Arnim argued12that the parenthesis a 27-31
is a later addition, on the grounds that (1) it intrudes unsyntactically
on the preceding sentence, (2) the mention of three sciences is irrelevant, and (3) the statement here that unmoved movers are outside the
province of physics is contradicted by the dictum a few lines later
(a 35 - b 5) that an unmoved mover x%veZucptxia.These do not seem
compelling arguments to support von Arnim's case'3, though of course

in the text of Aristotle any irrelevant parenthesis of this kind is likely

to arouse suspicion, and von Arnim may quite possibly be right about
this passage. More recently Solmsen has tentatively suggested14that
in its original version 7.2 ended at the words xx' e'L 6 r[ CLM at a 32.
The succeeding words xot kETO 7rpiO'rOvxLVY)cvXV,together with what
follows, "change the direction of the thought and may well have been
added at a later time". Again this is a possible hypothesis, but (as
Solmsen himself goes on to say) one that it is difficult to confirm.
It may be right to suspect further fragmentation of 7.2 along these
lines, or it may be that the whole section (i.e. 198 a 21-b 9) should be
"1 I am assuming that Met. A is later in date than the Physics. Although many
scholars take Met. A to be an early work (so most recently I. During, Aristoteles,
pp. 189-190), I find it difficult to believe that it could have been written before
Physics 0. It must surely be later than the Physics, if only because its theory of
the Unmoved Mover represents a clear advance on that of Physics 0, and
because it presupposes and refers to a number of doctrines expounded in
Physics 0 (e.g. 1071 b 6-10, b 10-11, 1073 a 7, a 10; these passages are integral to
the argument and can hardly be explained away as later additions).
12 H. von Arnim, 'Zu W. Jaegers Grundlegung der Entwicklungsgeschichte des
Aristoteles' (Wiener Studien 46 (1928) lOff).
"I Of von Arnim's points, (1) and (2) are hardly conclusive, while (3) seems to be
an unreal difficulty. The statements about unmoved movers seem to be paradoxical rather than contradictory. An unmoved mover is not a TUMxh &pyn(a 36),

in the sense that it is not a natural object and does not have an &pyi xIacv(

kV

murCo
(a 37). The study of such a mover in itself is therefore not a department of
CpuaLxY (a 28, oux&rL
9uputx<), but comes under the heading of 7rpcu11 PL),OaOqta.
This, however, does not prevent it from having an effect on natural objects
which is 9uatx6v (i.e. it XLVeL
9ulamXi, a 36), in the sense that when it operates on
them its effect is to make them move in a natural way (i.e. it actualises their
&pX?xLV*aecoq&vcxa9). So although an unmoved mover as a subject of study is
itself outside the province of physics its effect on the world of natural objects is
a proper subject of study for the physicist. Simplicius recognises this paradox:
OGY&p &arL(puaLX' &XECVin
(puaixCv(367. 12-13). So too does Philoponus,
PX, &X)Xo
who explains the point at greater length (304. 11-24).
'1 F. Solmsen, Aristotle's System ol the Physical WVorld
(Ithaca, 1963), p. 113 n. 83.

259

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions

classed as a later insertion; if there is any force in my first argument

above, it suggests the latter hypothesis. But in either event enough has
been said to suggest that the section 7.2 either includes or consists of
later insertions, and that the references to an unmoved mover did not
appear in the original version of Physics B.
I have argued that the four passages in Physics B and Ir where
Aristotle refers to an unmoved mover are later additions, and that
there is no evidence for any mention of the doctrine in the original
version of Physics ABFL\. This suggests that the references to it in our
present version are not references forward, anticipating a doctrine that
Aristotle saw as the culmination of the Physics, but notes added later,
either by Aristotle himself or by an editor, after the doctrine had been
expounded.
If the doctrine of the Unmoved Mover did not appear in the original
version of the Physics, it seems a fortiori unlikely that it occurred in
de Philosophia, which presumably represents an earlier period of
Aristotle's development. But there is one important passage in
Physics B 2 which seems to tell against this thesis; this is the passage
at 194 a 35-36, which is cited by most editors as a fragment of de
Philosophial5.
In the latter part of this chapter Aristotle is discussing the question
should take account of both form and matter (194 a
whether the ypuamx64
is
related
to form as means to end, as is shown by those
Matter
15-16).
TeyvoLwhich make their u`Xninto something and enable us to use it for
our own advantage (a 33-35); but the illustration misfires (Aristotle
is rambling here,) because ,uiCv ?vexm (a 34) exemplifies not the point
that he wanted to make, viz. that form is a re'),o but a different point,
viz. that we, i.e. the users of manufactured goods, are also a re&?og,but
in a different sense'6. He has to explain this difference - or rather he
has to note that there is a difference between these two senses of re'),o
and he refers the reader to de Philosophia for the explanation. This
shows that de Philosophia must have explained the difference between
two sorts of final cause, o'urvexc rLIVo;and Lv[.In several other passages
from the treatises where Aristotle makes this distinction he does so in
order to expound the notion of an unmoved final cause; from this some
scholars have inferred that the unmoved final cause referred to in de
15

Fr. 28 (Ross), fr. 30 (Walzer).

"I Cf. Graeser, 'Aristoteles' Schrift Uber die Philosophie und die zweifache
Bedeutung der causa finalis' (Mus. Helv. 29, (1972) 56ff).

260

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions

Philosophia was the Unmoved Mover, i.e. the prime mover of the
universe17.

This inference is clearly somewhat speculative. There are of course

other sorts of unmoved final cause besides the prime mover of the universe; in Aristotle's theory of causation final cause and formal cause
are only different aspects of the same thing, so that any formal cause
ranks as a re'Xos
and can be classed as a oiu9vxca (more precisely a o6
9ve%XClv6r),

as Aristotle himself points out at 198 b 2-4. Thus it is by no

means certain that the appearance of the term oi 9vexaonv6qin this

quotation from de Philosophia entails the presence in that work of the
doctrine of the Unmoved Mover. It has been pointed out that we have
no knowledge of the context in which Aristotle introduced this distinction between the two meanings of oT?vexa'8; it is true that in Met. A he

uses the distinction as a means of explaining the movement of the

universe"9,but we have no right to assume that he used it for the same
purpose here. On other occasions when he makes this distinction,
it is not always with the Unmoved Moverin mind20.
All the same it is not enough to make this negative point. There may
be no direct evidence to show that the Unmoved Mover figured in de
Philosophia, but we know that the dialogue contained a discussion of
theology2l; this provides an obvious context for the mention of final
causation, and it is difficult to see, in a reconstructionof the dialogue,
any other plausible context in which the topic might naturally have
17 Cf. J. Bernays, Die Dialoge des Aristoteles (Berlin, 1863), pp.
108-110, followed
by H. F. Cherniss, Aristotle's Criticism of Plato and the A cademy (Baltimore, 1944),
pp. 594-595, E. Berti, op cit. pp. 355-356, and M. Untersteiner, op. cit. p. 285.
For the most recent discussions, see K. Gaiser and A. Graeser, op. cit. (above,
note 3).

18 Cf. Solmsen,

op. cit. (above, note 14) p. 113, n. 83, and PNpin, op. cit. (above,

note 3), pp. 468-469.

'"1072 b 1-3-

Cf. de Anima 415 b 15-21, where the quXyis said to be the o6 gvexmfor the body
and its organs. There are three other passages where Aristotle makes the
distinction (de Anima 415 a 26-b7, Met. A 1072 b 1-3, and E.E. 1249 b 9-19).
In only one of these (Met. A 1072 b 1-3) is there an explicit reference to the
Unmoved Mover. The other cases show that the concepts of o5 bvexx and
20

&xivL*rov XLvoZvare not necessarily

connected

in Aristotle's

mind. In both these

passages he speaks of god or T-6&lov as the oTbvexaof things without any mention
of his (or its) effect on the movement of the heavens, or indeed on x[V-aCl as such
at all. For a full discussion of these passages, see Gaiser and Graeser op. cit.,
(above, note 3).
'A In frs. 16 and 26 (Ross) the dialogue is cited by name.

261

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions

been raised. A theological context is one where it would fit most

naturally and appropriately, and anyone who is reluctant to see it in a
theological context is under an obligation to suggest an alternative of
equal plausibility.
However, even if the context of this fragment in de Philosophia was
theological, this does not necessarily mean that Aristotle had evolved
the doctrine of god as the Unmoved Mover at this stage of his development. There are two distinct conceptions in Aristotle of the apyJ'of the
uriiverse: (1) there is god, the supreme being of the universe, whose
perfection makes him serve as a goal of aspiration, i.e. a final cause, for
all other beings in the universe, and (2) there is the Unmoved Mover
responsible for the movement of the heavenly bodies. These two conceptions are of course combined in one in Met. A, but this does not mean
that they must have been combined in Aristotle's mind at other stages
of his development, and in point of fact it hardly needs emphasising
that the account in Physics 0 is in terms of (2) only.
The doctrine of the Unmoved Mover as it is developed in Phlysics0
is based on Aristotle's analysis of xtv-caL; beginning with the thesis
that the eternal movement of the universe must be explained by
reference to a prime mover, which must itself be either self-moving or
unmoved, he examines the nature of xLv-natand concludes that a selfmoving prime mover is impossible; from this it follows that any prime
mover that we postulate must be unmoved. In this argument he is
thinking on a cosmic scale (although this is not stated explicitly at
the outset, and only becomes clear in the course of the argument), and
the unmoved mover that emerges in the course of E)is the Unmoved
Mover, i.e. the prime mover of the universe, conceived as the agency
responsible for the movement of the universe and in particular for the
movement of the heavens. The argument is worked out in mechanical
terms, and Aristotle's conception of the Unmoved Mover still seems
to be mechanical, since he assumes that a mover must be in contact
with the body moved by it, and he locates the Unmoved Mover at the
circumference of the universe. There is no idea of invoking final causation to explain the Unmoved Mover's mode of operation. Moreover
there is no attempt in this book to link mechanical and religious
ideas; there is no suggestion that the agency under discussion deserves
the title &e6;.
There is thus no necessary connexion between the two ideas of (1)
god as the supreme being of the universe, and (2) the celestial Unmoved Mover. Physics 0 contains (2) but not (1), and it is quite
262

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions

possible that de Philosophia, it it dealt with tlheology,may have contained (1) but not (2). In other words, one can accept that fr. 28 must
refer to god's cosmological function as a final cause without necessarily
accepting that this implies a view of god as the Unmoved Mover along
the lines of Met. A.
In point of fact, while (1) figures quite clearly in the fragments of
de Philosophia, there is nothing in them that suggests the presence of
(2). The basic evidence of the fragments, such as it is, is as follows:
Fr. 16 contains the argumenttumex gradibus entium: god is an entity

more perfect than any other existing thing, who stands at the head of a
scale of being; because of his perfection he is exempt from change, since
change in him could only be for the worse, and there is in any case
nothing superior to him that could act upon him and cause him to
change. If we can rely on fr. 17 as a genuine quotation from de Philosophia, Aristotle also spoke of this first principle as that which unites
the world and gives it order. From the jumbled criticism of fr. 26 we
can extract, as a minimum, the fact that Aristotle believed in a god who
in some sense stood at the head of the universe (prae/icit mundo),but it
is doubtful whether we can rely further on the details of this fragment.
The ideas contained in these three passages are strongly reminiscent
of four other passages in Aristotle:
(1) de Caelo 279 a 17 ff. This passage has been claimed by some
scholars as a quotation from de Philosophia.22
Whether or not this is correct, the beings outside the heaven to
which Aristotle refers here (-oxe-Z,a 18) are described as changeless and

enjoying the best possible kind of life. The punctuation of this much
disputed passage is uncertain, and a good deal depends on it, but it
looks as if the universe is said to depend in some sense on these beings
(6SbhvZ'e-trsc,
a 28)23.
(2) de Caelo292 a 22-23 and b 5-7. Here again the supreme being is
said to enjoy perfection and consequently changelessness.The heavenly
bodies move as they do in an attempt to attain the perfection that it
exhibits.
(3) de Anima 415 a 26-b 7 conveys a similar idea. Mortal creatures
cannot enjoy the true immortality of so IZov, and their life cycle is an
22

Fr. 28 (Walzer), fr. 29 (Untersteiner).

This is by no means certain. It is possible that the antecedent to 6Ovvis not
t&xeZor oupmv6q
but mtc'v.But even if this is so, it still implics the same kind of
dependence of the universe on 'cxeE,since the at 'v is the time-span of their
existence.
23

263

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions

attempt to come as close to this ideal as possible and to achieve the

vicarious immortality that is provided by continuation of the species;
in this way, living creatures are striving Eva -roUie' xac ?oi5 &ecou
eTX(AaLV

7cXvT
8&VOWVOCL

yap

?XSLVOU Opeyvrcx.

Here again it is implied that TorbtLovrepresents an ideal state of

perfection which acts as final cause for beings in the sublunary
world.
(4) E.E. 1249 b 9-19. Here again god acts as a final cause, by serving
is directed. The
as the object of contemplation to which rpp6vwaLm

meaning of this puzzling section of E.E. is very obscure, but it is clear

that god is seen as a final cause and that his function as final cause is in
some way connected with his perfection
8ae , b 16).

(Xitd

?)ezvo0

0tV OuAV64

In all these cases Aristotle speaks of some kind of supreme being(s)

whose perfectionis associated with his influence on the universe, and in
passages (3) and (4) this influence is said to be the influence of final
causation. It is striking that on both the occasions when he expounds
this idea Aristotle introduces it by way of the distinction between the
two types of final cause. It is tempting to suppose that his reason for
making the distinction in de Philosophia was the same; it looks as if
this distinction was associated in his mind with the idea of god as final
cause.
This suggests that in de PhilosophiaAristotle had formedthe concept
of god as final cause; thus we can place fr. 28 in a theological context
with some plausibility. But there is nothing in the evidence to suggest
that god was also conceived in the dialogue as the prime mover of the
heavenly bodies; this conception of god springs from a different line
of thought (viz. the analysis of xtv?larL coupled with an interest in

astronomy), for which there is no evidence in de Philosophia.

The thesis put forwardabove is similar to that recently advanced by
W. Potscher in his book, Strukturproblemeder Aristotelischenund
TheophrastischenGottesvorstellung24.
Potscher has argued that at the
period of de PhilosophiaAristotle thought of god as a final cause and as
a divine mind, but that this was only the starting point from which was
developed the later doctrine of god as the Unmoved Mover. Met. A
seems to combine the physical theory of Physics E) with the earlier
final causation of de Philosophia.De CaeloB 12 is perhaps the earliest
place where this combination can be seen in Aristotle's writings.
'"

op. cit. (above, note 3).

264

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions

If this thesis is correct, and if in de Philosophia Aristotle conceived

of god as a final cause, this is a sufficient explanation of fr. 28, and we
have no reason to read into the dialogue the doctrine of the Unmoved
Mover as commonly understood. The Unroved Mover, on this thesis,
was a later development. This squares with my conclusion above
concerningPhysics B and IF,that the doctrine of the Unmoved Mover
as commonly understood was absent from these books and from the
earliest period of Aristotle's thought.
Trinity College,Cambridge.

265

This content downloaded from 134.76.63.66 on Mon, 23 Feb 2015 14:13:35 PM

All use subject to JSTOR Terms and Conditions