Lucretius on Atomic Motion: A Commentary on de Rerum Natura 2.1-332. By Don Fowler.

Lucretius on Atomic Motion: A Commentary on de Rerum Natura 2.1-332.By Don Fowler. Oxford: Oxford University Press, 2002. p.

xiii+ 513. $110.00. 0-19-924358-1.Reviewed by Tim O’Keefe

[Penultimate draft. Please see the published version, Ancient Philosophy, vol. 23no. 2 (2003), 461-468 to cite.]

The main body of this commentary is Don Fowler’s doctoral

thesis, submitted to Oxford University in 1983. His death at

age 46 in 1999 unfortunately prevented him from composing a

commentary on the whole of book two, but his widow, Peta

Fowler, and the other people who helped bring this volume to

light should be commended. It will be an essential reference

work for everyone who works on the areas of Epicurean

philosophy that Lucretius addresses in DRN 2 1-332. In

addition to the usual sort of detailed textual analysis

contained in a commentary, this volume distinguishes itself

by containing extensive argumentation on how to interpret

Lucretius’ position on philosophical issues raised in DRN 2

1-332, in particular, what role the swerve plays in free

action, and what sort of reductionist Epicurus was. These

issues are dealt with in the commentary itself, but also in

three free-standing papers (one previously unpublished)

included as appendices. I’ll begin by discussing the

commentary as a whole, and then turn to evaluating Fowler’s

arguments about the role of the swerve in free action.

The last commentary in English that dealt with this

section of DRN was Bailey (1947), which covers the whole of

DRN. Fowler’s commentary represents a significant advance

on Bailey’s in at least two respects: currency and depth.

First, Fowler is able to take into account decades of

scholarly work on Lucretius not available to Bailey. In

particular, Fowler considers much of the work on the swerve

spurred by Furley’s Two Studies in the Greek Atomists and the

sections of Epicurus’ On Nature 25 concerning freedom and the

development of one’s character. Fowler’s own commentary

contains references only up to 1983, naturally enough, but

the editors have updated the bibliography to include works

since then and inserted footnotes referring to them at

appropriate points. They have done a fine job of including

references to the literature since 1983. (I do have one

editorial complaint: references to most standard

commentaries and editions do not conform to the author

(date) format used in most of the other references, and

these works are not included in either the bibliography or

the list of works cited. Most classicists would know how to

track down a reference like ‘Nisbet and Hubbard on Hor. C.

I. I, p. 3,’ but if OUP wishes a wider audience for this

commentary, it would be more user-friendly to put all

references in a standard format and include all works cited

in the bibliography.)

Secondly, Fowler’s commentary goes into far more detail

than does Bailey’s. The main body of the commentary is 392

pages, i.e., an average of more than a page per line of

Latin. By way of comparison, the section of Bailey’s

commentary on this portion of DRN is 64 pages. Fowler

treats difficulties in the received text, considers the

logical structure and import of Lucretius’ arguments, draws

extensive parallels in Latin and Greek literature to

Lucretius’ metaphors, and discusses the tone and rhythm of

individual words and phrases. Fowler allows himself to roam

widely, and although this lengthens the commentary

considerably, generally this indulgence makes for

interesting and suggestive connections, more so than one

might expect in a typical commentary. For instance, when

discussing Lucretius’ joy in comparing his own situation to

the poor drowning sailors and whether this joy is a form of

schadenfreude, Fowler pulls into the mix Bishop Butler, Edmund

Burke, and Aquinas’ doctrine that the joys of heaven include

contemplating the suffering of the damned.

Also, Fowler includes interesting discussions of many

issues that are dealt with only in passing in this section

of DRN. This volume should not be consulted only by people

interested in issues of atomic motion. Fowler’s excellent

treatment of Lucretius’ ‘semi-personification’ of Nature at

DRN 2 168 would be useful for people working on other

sections of DRN which contain more elaborate

personifications of Nature like 2 594-660 or 3 931-962.

Similarly, Lucretius’ oration on the plight of most people

and his attack on luxury at the beginning of DRN 2 affords

Fowler the opportunity to fruitfully discuss Epicurean

ethics. Fowler makes some brief but intriguing suggestions

along the way that deserve to be picked up and argued for at

more length by somebody, such as his analysis of the

relationships between different types of fears and desires,

represented in diagrammatic form on p. 79, or his remark on

p. 275 that the debate between Epicureans and Aristotelians

on natural motion has ethical implications.

Fowler stresses how coherent, well-written, and

logically well-articulated the DRN is; he sets out each

section of the poem in terms of its argumentative structure,

highlighting what he takes to be the crucial logical

connectives—a useful feature. Although Fowler doesn’t gloss

over the rough patches in Lucretius’ verse and admits that

the arguments are sometimes less than convincing, his

overall approach is highly sympathetic, and I think that

this will prove helpful to most readers.

As I mentioned earlier, this commentary is also notable

for its extended argument for Fowler’s own view about the

role of the swerve in preserving human freedom. The argument

is contained in the body of the commentary on DRN 2 251-

293, and in two of the appended papers: ‘Lucretius on the

Clinamen and ‘Free Will’ (2. 251-293),’ originally published

in 1983, which repeats much of the commentary but is still

useful as a self-contained extract of Fowler’s position, and

‘What Sort of a Reductionist was Epicurus? The Case of the

Swerve,’ a previously unpublished paper delivered in 1993,

which deals in more detail with issues of reductionism and

also has a brief synopsis of his position (440-441), which I

found useful for clarifying some obscurities in his earlier

expositions of it.

For Fowler, voluntas (volition, or will) is the cause of

our actions, but voluntas is not itself causally determined

by the past states of the mind. It is a self-initiated

motion in the mind and is constituted by a swerve (p. 330).

Thus, in broad outline, Fowler’s position (and much of the

argument he gives for it) is a variant on the position of

Bailey, and followers of Bailey including Asmis and

Purinton, that voluntas simply is the swerve of the atom. (This

has been overlooked by previous commentators on Fowler’s

1983 article. Pace Purinton (1999) 256, Fowler does not think

that a volition is a separate event that occurs after a

swerve happens.)

But the exact relationship between swerves and

volitions according to this family of views is still not

entirely clear—perhaps swerves ‘constitute’ volitions (as

Asmis (1991) 291 and Fowler say), and in some sense swerves

‘produce’ volitions (Fowler’s term) or ‘cause volitions from

the bottom up’ (as Purinton puts it), but what exactly does

all this amount to? Fowler starts to address this issue in

‘What Sort of a Reductionist was Epicurus? The Case of the

Swerve.’ Fowler says that a ‘type’ of psychological

phenomenon—an act of volition—is ‘related’ to a type of

atomic event, the swerve. But Fowler realizes that there

cannot be a straight type-type identity or reduction between

volitions and swerves, since only swerves that occur in a

suitable atomic compound—an animus, and not e.g., a coffee

cup—will be volitions. (440) In response to this difficulty,

Fowler hedges, saying “It is not quite true that the swerve

itself is libera voluntas, but rather the alteration of the

compound’s atomic constitution by the swerve is or

represents or is reducible to—or whatever—the act of will.”

(440-1) Despite the vagueness, I think that Fowler’s

suggestion is worthwhile, and if one wishes to advance a

Bailey-type view—which I do not—something like it could be

further developed. Here is my suggestion on how to do so, by

emphasizing relational features of the swerve, as does

Fowler:

Individual volitions are identical with particular

atomic swerves—that is, there is a token-token identity

between each volition and the swerve that constitutes it.

However, what makes a libera voluntas that type of thing (a free

volition, and not merely a random swerve) is that it is the

sort of thing that is apt to be a self-initiated start of an

animal action: volitions (typically) result in horses breaking

from starting gates, people walking, and the like. In order

to account for what allows some swerves to initiate action,

it is insufficient to look to the intrinsic features of the

swerving atom and its trajectory; one must also include in

one’s account the relationships of that swerving atom to the

atoms around it in the animus. As Fowler rightly notes,

Epicureans insist that these sorts of relational features

must be included in a full atomic explanation of many

accidental properties, especially dispositional ones like

‘being poisonous’, and that this insistence is compatible

with a robust reductionism about what it is for something to

be poisonous. (437-9, see also O’Keefe (2002) 158-60) Once

we specify which features of the atoms and their

relationships to one another allow some swerves to act as

self-initiated origins of action, this opens up the

possibility of type-type, and not merely token-token,

identity: volitions are not type-identical to swerves tout

court, but to the class of swerves that have the proper

relationships to other atoms that make them suitable as

(self-)initiators of action.

Fowler’s view is not merely a recapitulation of

Bailey’s. Fowler puts forward an analysis of exactly where

the swerve works into the account of action Lucretius gives

at DRN 4 877-906 (esp. 881-888), even though he doesn’t

mention the swerve in this passage. Fowler’s view (441) is

that volitions work “through a focusing the mind,” as

Lucretius describes in DRN 4 802-17. In DRN 4 802 ff.,

Lucretius says that our minds are constantly bombarded by

fine atomic simulacra, and this is what allows us to

visualize, at will, what we decide to visualize: we can

focus the mind on those simulacra we wish to in order to

produce an image (simulacra of countless other things are

‘available’ to us, but since we don’t focus on those, we

don’t notice them). In DRN 4 881-888, Lucretius also starts

by asserting that our minds are constantly bombarded by all

sorts of simulacra, and that this allows us to ‘foresee’

(provideo) what we wish to do. Fowler thinks Lucretius’ view

is that when one decides to act, this volition (which is

also a swerve) involves a “voluntary focusing of the mind”

upon simulacra of what one wishes to do. This focusing forms

an image of what one wishes to do, and this image then

causes one to act.

This account faces many serious difficulties. The first

is textual: Fowler explicitly says that the voluntas precedes

the production of the image (330) (e.g., you decide to walk,

and this decision/focusing of the mind on simulacra of

walking produces an image of oneself walking), but the

passage seems to put the relationship the other way around.

The problem for Fowler isn’t that Lucretius says (in 881-3)

that simulacra of walking strike the mind before voluntas

occurs (pace Englert (1987) 183 n. 9), since Fowler admits

this, and says instead that it is the willing to walk, which

involves focusing on those images which have been bombarding

the mind, that precedes the production of the image.

However, in 883-4, Lucretius seems to say that ‘foreseeing’

what one wishes to do occurs before one initiates any action:

inde voluntas fit; neque enim facere incipit ullam

rem quisquam, quam mens providit quid velit ante.

(After that voluntas arises; for nobody begins to do

anything, until the mind has previously foreseen what

it wants.)

For Fowler’s position to stand, he has to read Lucretius as

identifying the ‘foreseeing’ of walking with the voluntas to

walk, and then as stating that this dual action of

‘foreseeing’ and willing what one wants must occur before

any action is initiated. (p. 330) But this reading of the

passage is quite strained. Lucretius is trying to explain

why voluntas occurs after images strike the mind, and he says

that ‘foreseeing’ what one is going to do must occur before

one initiates any action. Since voluntas is what initiates

action, this suggests that ‘foreseeing’ what one wishes to

do occurs before the voluntas. Therefore, I think that the

more natural reading of the passage is as follows. We decide

to walk. But to reach this decision, we must be able to

think of ourselves walking before reaching this decision (we

must ‘foresee’ it), and we are able to do this only because

there are images of walking available to our minds, subtle

images constantly bombarding the mind from without.

I agree with Fowler that there is an important parallel

between DRN 4 877 ff. and the earlier discussion of how one

visualizes what one wants, but I think that he misidentifies

this parallel. In effect, Fowler wishes to read the earlier

discussion of visualization into the later discussion of

action as an extra step which he interpolates: making the

decision to walk involves picturing ourselves walking—when

we decide to walk, we “clearly visualize the image of

walking,” and after doing this, “then the bodily reactions

of walking follow.” (p. 329) (This is why Fowler chooses to

translate provideo with the clunky ‘previsioning,’ which I

think is too literal; ‘foreseeing,’ which may involve just

thinking ahead, is preferable.) This theory is quite

implausible in itself: it might make sense to say that in

order to decide to walk, I have to think of walking

beforehand, but I do not normally visualize myself picking

up a coffee cup before I pick up the cup, and deciding to

pick up the cup doesn’t involve visualizing myself doing so.

I think that the two passages, instead, are describing

how the fine images that are constantly and directly

striking the animus are necessary to explain, respectively,

how we can both decide to picture to ourselves what we wish to picture to

ourselves (DRN 4 781 ff.) and decide to move ourselves as we wish to

move ourselves (DRN 4 877 ff.). Each passage describes how we

can make a different type of decision. In the earlier

passage there is a decision such as to visualize oneself walking,

whereas in the later passage there is a decision to walk. In

both cases, there must be images of oneself walking available

for the animus to focus on in order to perform the operation

described. But I don’t think Lucretius believes that, in

order to decide to walk, one must first decide to visualize oneself

walking or that deciding to walk involves visualizing oneself

walking. (Lucretius’ assertion in DRN 885 that an image

(imago) of what the mind foresees is produced and his

discussion in DRN 4 877 ff. both show that Lucretius

probably pictured thinking in overly imagistic terms. But

this doesn’t effect the point that the ‘foreseeing’ seems to

occur before the decision to walk, or that the parallel

between the two passages doesn’t strongly support Fowler’s

identification of making a decision with a kind of

visualization.)

A second textual problem for Fowler is that the swerve

isn’t mentioned where it should be mentioned, if he is right

about its role. Fowler says that every voluntas is

constituted by an atomic swerve in the animus, but Lucretius

says nothing in this part of DRN (or indeed, anywhere other

than DRN 2 216-293) about swerves occurring, even though

(pace Purinton (1999) 279-80) this passage is supposed to

describe what goes on at the atomic level when a person

engages in voluntary action.

Fowler recognizes this problem and responds that it

isn’t surprising that Lucretius doesn’t mention the swerve,

since “Lucretius is concerned in this passage with how we

move when we wish to, not with how we come to wish to move;

hence there is no explanation of how voluntas occurs.” But I

find this inadequate. Fowler quite clearly distinguishes

between desiring to do something and willing to do it. He

writes, of the example of the horses bursting from the gates

in DRN 2 263ff., that the horses are eager (cupidam, 265) to

run, but they do not move until the command is given by the

voluntas. Since (according Fowler) the willing to move is a

distinct event from the preceding desire to walk, Lucretius

cannot in the passage in DRN 4 simply be assuming that one

comes into the situation being described with one’s voluntas

to walk already formed, so that its formation needs no

explanation in this context. Simply to assert that the

voluntas occurs after the images strike the mind without

mentioning the swerve at all would, at the least, be highly

misleading if the voluntas really is a swerve, and it would

give the unwary reader the strong impression that the

voluntas is caused by the reception of the images.

Fowler believes that the example of the horses breaking

from the stalls helps illustrate his thesis. According to

Lucretius, there is a perceptible, although slight, delay in

the horses getting started, although they’re eager to go.

Fowler thinks this shows that, according to Lucretius, we’d

expect the stimulus (the opening of the gates) to

immediately produce the effect of the body moving, since the

horses are already eager. This doesn’t occur immediately

because the horse’s movement awaits the decision of the

animus to assent, so the horse delays until one of these

unpredictable swerves (which constitutes a voluntas) occurs.

There are two major problems with this. The first is

that Lucretius isn’t using the example to illustrate a delay

while the horse’s animus has to wait for a swerve to happen.

His point, instead, is to illustrate his contention that

there exists motion initiated by the animus, in which case

it takes some time for the decision of the animus to stir

together and move all of the matter of the horse in a

coordinated manner, which he contrasts with motion caused by

external shoves, which don’t require the time for such

internal processing. (2 266-276).

The second is that the clinamen, on Fowler’s view, makes

us ‘free’ from causal determinism by placing a potential

roadblock between our desiring to do something and our doing

it. Fowler says that usually an animal will be able to do

what it wants to pretty soon, since swerves probably occur

fairly frequently. But, he admits, the delay time is

unpredictable, and occasionally no atoms at all will swerve

in the relevant period. (331-332) This sort of ‘freedom,’ it

seems, could not help us in our pursuit of pleasure; the

best we could hope for is that it doesn’t get in the way too

often. So it’s difficult to see why any good Epicurean would

value ‘freedom’ on this account.

The usual sort of maneuver here (which Fowler also

makes) is to say that this is a problem with Epicurus’

position, not his interpretation of Epicurus’ position.

Although space limitations prevent me from dealing with this

response in detail, let me note that I don’t find it

convincing. Lucretius’ concern in DRN 2 251-293 is to show

how we can have a voluntas which allows us to go where

“pleasure leads us” (DRN 2 257-258), and in Ep. Men. 133-4, De

Fato, and On Nature 25, as I read them (see O’Keefe (2002)),

Epicurus is concerned to preserve our agency from the threat

of a necessity which would render us helpless and prevent us

from being able to do what we wish to do. Although I

disagree with some points of detail, I think that Bobzien

(2000) is largely right that nothing in Epicurus or

Lucretius should make us think that they are concerned with

the “problem of free will” in some modern sense, or that

they propose a libertarian position to solve this problem.

Their concern is the autonomy of the agent, not the freedom

of the will. Any interpretation of the role of the swerve

should show why Epicureans might think that it helps

preserve a sort of freedom that would matter within their

ethics—the sort of freedom that we need in order to be able

to achieve ataraxia.

The main focus of the previously unpublished ‘What

Sort of a Reductionist was Epicurus? The Case of the Swerve’

is Epicurus’ ‘reductionism’ generally, not the role of the

swerve in particular. This essay, along with Jonathan

Barnes’ ‘Bits and Pieces,’ should be required reading for

anybody who wishes to discuss Epicurus’ ‘reductionism’ or

‘non-reductionism.’ Fowler rightly notes, “Before assessing

whether Epicurus was a reductionist, we have to get clear

the matrix of possible positions against which we shall be

trying to situate his thought. It may all depend on what we

mean by reductionist.” (429-430) This may seem like obvious

advice, but it is often ignored, and Fowler nicely lays out

different types of reductionism before arguing (I think

successfully) that many of the examples often cited to show

that Epicurus is not a reductionist fail to do so, for at

least some interesting varieties of reductionism. He

concludes that Epicurus does have a “sort-of reductive

project,” which is an aspiration to “find an explanation at

the atomic level for all macroscopic phenomena.”

The final paper in the appendix is ‘The Feminine

Principal: Gender in the De Rerum Natura,’ reprinted from G.

Giannantoni and M. Gigante (eds.), Epicureismo Greco e romano: Atti

del Congresso Internazionale, Napoli, 19-26 Maggio 1993 (Naples 1996),

813-22. Here, Fowler argues that Lucretius’ discussions of

Nature in feminine terms do not embody objectionable

patriarchal views towards women and nature. Fowler lucidly

lays out the case that has been made against Lucretius on

this score, before ably defending him against it, and this

essay engagingly opens up issues about Lucretius’ symbolism

that I had never considered before.

Despite my serious reservations about Fowler’s views on

the role of the swerve, let me emphasize again that this

book will be indispensable for researchers who work on DRN

2 1-332. I would not be surprised if more than 55 years pass

again before another commentary on this part of the De Rerum

Natura appears that improves significantly on it.

Works Cited

Asmis, E. 1990. ‘Free Action and the Swerve,’ Oxford Studies in

Ancient Philosophy 8, 275-290.

Bailey, C. 1928. The Greek Atomists and Epicurus. Oxford:

Clarendon Press. (Reprinted 1964, New York: Russell and

Russell).

_______. 1947. Titi Lvcreti Cari. De Rervm Natvra: Libri Sex. Oxford:

Clarendon Press.

Barnes, J. 1988. ‘Bits and Pieces,’ in Matter and Metaphysics,

edd. J. Barnes and M. Mignucci, Naples: Bibliopolis,

225-294.

Bobzien, S. 2000. ‘Did Epicurus discover the free will

problem?’ Oxford Studies in Ancient Philosophy 19: 287-337.

Englert, W. 1987. Epicurus on the Swerve and Voluntary Action, American

Classical Studies, 16. Atlanta: Scholar’s Press.

Furley, D. 1967. Two Studies in the Greek Atomists, Princeton:

Princeton University Press.

O’Keefe, T. 2002. ‘The Reductionist and Compatibilist

Argument of Epicurus' On Nature Book 25,’ Phronesis, vol.

47 no. 2, 153-186.

Purinton, J. 1999. ‘Epicurus on ‘Free Volition’ and the

Atomic Swerve,’ Phronesis vol. 44 no. 4, 253-299.