Lucretius on Creation and Evolution: A Commentary on De rerum

natura 5.772--1104 by Gordon Campbell

Oxford/New York: Oxford University Press, 2003. Pp. xii + 385.
ISBN 0--19--926396--5. Cloth $114.80

Reviewed by
Brooke Holmes
University of North Carolina, Chapel Hill
[email protected]

If survival of the fittest is a principle as relevant to ideas as it is to

species, Darwinism has proved to be a hardy breed, especially if we
judge provocation to be a sign of life. Last November alone, voters
in Pennsylvania ousted school board members who had instated poli-
cies that gave Intelligent Design a hearing alongside Darwin in ninth-
grade biology, while the Kansas Board of Education removed obsta-
cles to teaching both perspectives. The Austrian Cardinal Christoph
Schoenborn felt obliged to go on the record once again about the
Church’s views on the debate over evolution in America. And an am-
bitious exhibition simply entitled ‘Darwin’ opened at the American
Museum of Natural History in New York. One reviewer [Rothstein
2005], citing an 1844 letter in which Darwin says that writing about
his ideas was ‘like confessing a murder’, takes the curator to task for
domesticating a ‘bizarre’ and ‘shocking’ idea. What is so unnerving
about Darwinism? The Cardinal is blunt: ‘It’s all about materialism,
that’s the key issue’ [Heneghan 2005]. It is the idea that matter is
the only reality.
Epicureanism is Antiquity’s most infamous promoter of mate-
rialism, as well as the attendant horrors outlined by the Kansan
Board—secular humanism, atheism, and the idea that life is acciden-
tal, both in the everyday and in its genesis. 1 Nowhere, perhaps, is
the defense of these tenets more vivid and calculated than in the fifth
book of Lucretius’ De rerum natura, in which our gaze is shifted from
the primordia rerum, the imperishable first-beginnings of things, to

1 Of course, strictly speaking, the Epicureans are not atheists; but their gods
inhabit another world and care not a whit for humankind, which is to say
that their brand of humanism is aggressively secular.
© 2005 Institute for Research in Classical Philosophy and Science
All rights reserved
issn 1549–4497 (online) issn 1549–4470 (print) issn 1549–4489 (cd-rom)
Aestimatio 2 (2005) 142--162
Brooke Holmes 143

the contingent origins of our mortal cosmos and the organisms that
inhabit it. Lucretius has an axe to grind in this book with Anti-
quity’s teleologists—the Platonists, the Aristotelians, and, most of
all, the Stoics. His stated aim is to disabuse his reader of the idea
that the nature of the universe is owed to any demiurgic blueprint.
From the outset, the most visible cost of this disenchantment—which
for Lucretius, of course, is its greatest gain—is the uncoupling of the
human from any cosmic master plan. The critique of teleology is,
then, at its heart, a critique of the idea that the world was created
hominum causa, for the sake of people: cosmology for Epicurus and
his followers always entails a payoff for ethics.
Human exceptionalism is, in fact, recast as the particular hostil-
ity of the earth towards these creatures. This is best represented by
the helplessness of the human infant, which is contrasted to the ease
with which the other animals are at home in the world. Of course,
birth trauma also matters to a teleologist like Plato. But whereas the
Timaeus sees education as a realignment of our true nature with the
divine order of the universe, in Epicureanism, human development is
a history of creating defenses, good and bad, of which philosophy is
the most noble and effective. This phylogenesis is then restaged, at
least in part, as ontogenesis. The second half of book 5 of De rer. nat.
takes up the task of accounting for human nature in a cosmos short
on divine solicitude and partial to all the other animals, by investi-
gating how anthropogony parts ways with zoogony under the force of
circumstances. That is, it sets out to explain how the human (scil. us)
is produced in time through the interaction of organism and environ-
ment, rather than to describe the pet project of a benevolent creator.
Lucretius’ is a complex story in book 5: he tries to explain the
spontaneous generation of life, speculates about human prehistory,
gives a description of early communities that includes laconic expla-
nations of the origins of justice and language, and traces the develop-
ment of cities and civilization. For years, the mixture of apparently
dystopic and utopic elements at both the early and the late stages
of this story puzzled scholars, who split on whether Lucretius was a
‘Primitivist’ or a ‘Progressivist,’ i.e., whether he idealized the past
or the present. The past few decades, however, have seen an increas-
ing dissatisfaction with this either/or opposition and an attempt to
 144 Aestimatio

engage more carefully with the text and its competing movements. 2
This trend has opened up a space for a new set of questions for read-
ers of book 5: What constitutes the distance between humans and
other creatures—viable life forms and monsters? What is the rela-
tionship between human vulnerability and the pursuit of ars? What
is the role of necessity in producing the human, and where or when
does it give way to a form of self-fashioning? What ensures the sur-
vival of humans in a quasi-Darwinian world of species competition?
How do the stories that Lucretius tells about the spontaneous origins
of life or human exceptionalism relate to our own dominant, albeit
hotly contested, evolutionary narrative?
Lucretius on Creation and Evolution, Gordon Campbell’s new
commentary on some of the most interesting lines in Lucretius’ story
(5.772--1104), stakes out the ground of some of these issues. Camp-
bell has written a thoughtful and timely reassessment of Lucretius’
engagement with his teleological opponents, as well as with Preso-
cratic zoogony—good use is made of the recently published Stras-
bourg fragments of Empedocles—Golden Age myths, and, of course,
the elusive master text of Epicurus himself. Yet, if all roads lead
to Lucretius, they approach him both from the periods prior to the
poem and from our own recent, and sometimes very recent past. Lu-
cretius’ poem was one of the most prominent explanations of the
creation of life in mechanistic and non-teleological terms from the
early Renaissance until the publication of The Origin of the Species
in 1859, and Darwin’s theories bear some striking similarities to those
found in the ancient tradition of thinking about the beginnings of life.
Besides pursuing its natural task of Quellenforschung, then, Camp-
bell’s commentary wagers that De rer. nat. 5 intersects with our
present set of questions and anxieties about evolution in interesting
ways, and sets out to map these points of intersection. ‘Map’ may be
the wrong word here, however, since it is in his joining of past and
present that Campbell is at his most creative.
In fact, it would have been helpful if a better map of Camp-
bell’s own work had been made available to the reader. No attempt
is made in the introduction to integrate the lines chosen into the
poem as a whole or book 5. Ready familiarity with more general

2 E.g., recently, Blickman 1989, Nussbaum 1994, Asmis 1996, and Holmes
2005.
 Brooke Holmes 145

questions of background and context for the poem is assumed, as is

a strong grounding in Epicureanism. Most of the Greek is translated,
while, unsurprisingly, the Latin is not, although Campbell offers, in
addition to the Latin text, a lucid translation of his excerpt. 3 Still,
a reader working in translation on key topics, such as the origins
of language or failed species, could glean much from the series of
section introductions which form the backbone of the commentary
proper. These range from brief transitioning remarks to condensed
versions of working papers that Campbell has published elsewhere
[2002a, 2002b]. Campbell has also added two long appendices—a
‘Table of Themes in Accounts of Creation, Zoogony, and Anthro-
pogony’ and a ‘Table of Themes in Prehistories and Accounts of the
Golden Age’ (including Isles of the Blessed, Ideal States, Noble Sav-
ages, and so forth)—which, while not exhaustive, offer a wealth of
data. The index itself is unfortunately short and rather arbitrary.
But despite limitations on the text’s accessibility to non-specialists,
both students of Epicureanism and those interested in the history of
thinking about the origins of life and the human in the Western tra-
dition now have a stimulating guide to what is fascinating, if difficult,
Lucretian terrain.
The luxury of painstaking attention to the text is the great joy
of a commentary. Such attentiveness in a commentary like this one,
where it is combined with a deep sensitivity to larger issues, often re-
wards. For centuries, scholars have debated the complex relationship
between poetry and science in the epic masterpiece of a philosophical
tradition ostensibly hostile to myth and poetry. One of the strengths
of Campbell’s commentary is its deft negotiation of Lucretius’ poetic
and philosophical strategies, which allows one to observe the details
of this dynamic. Lucretius has recourse to atoms and void as ex-
planatory mechanisms in book 5 less often than elsewhere in the
poem. This is not to say that we lose sight of the atomic under-
pinnings of the visible world, but we do spend more time at the
macrophysical level. As a result, book 5 is rich in topoi on the for-
mation of life, accounts of the Golden Age, and the development of
human civilization, making it an excellent place to observe Lucretius’
‘remorseless appropriation and recontextualization’ [138] of his non-
Epicurean predecessors in action.

3 On Campbell’s text, see Volk 2004.

146 Aestimatio

Central to Campbell’s treatment of Lucretius’ strategy is his use

of Richard Dawkins’ theory of memes. Memes are described as

the sort of generally accepted background ideas whose origins

are untraceable. . . that tend to exist and evolve as if they have
a life of their own independent of any writer. [180--181]
Campbell speaks, for example, of the ‘Darwin meme’, later corrected
to ‘the pack of memes surrounding Darwinism that make up our Bil-
dungsgut of prehistory’ [183]. More commonly, however, memes are
not attached to a proper name. This presumably justifies Campbell’s
decision to introduce them only ad lin. 925--1010, i.e., over halfway
through the commentary, rather than in the context of his discussion
of the Presocratic background to Lucretius’ zoogony. What appears
key to the definition of a meme, then, is that it is a critically unex-
amined concept which travels easily beyond its initial context: the
meme ‘survival of the fittest’, whose peregrination in the 19th and
20th centuries has been problematic, is an excellent example. Thus,
it makes sense that Campbell appropriates memes for his discussion
of how Lucretius integrates material from other sources and shapes it
to suit his purposes, all the while working, whether actively or more
subtly, to head his reader off from the kinds of incorrect inferences
to which those sources fell prey. Memes, for Dawkins and Campbell,
function like viruses against which one may need to be ‘vaccinated’,
although they may still be useful.
Campbell sees the Golden Age ‘as an integral part of Lucretius’
prehistory, and as the very material out of which he builds it. Cer-
tain themes’, he claims, ‘are rationalized and debunked, while others
are allowed to remain untouched and to do their work of vaccination
simply by their recontextualization in Lucretius’ account’ [184]. Ana-
lyzing a descriptive passage on the streams that satisfied the thirst of
early humans, Campbell argues, for example, that Lucretius appro-
priates pastoral poetry in order to advance the principle of ‘cultural
gradualism’ against the myth of divine beneficence. He then goes
on to suggest that Lucretius also uses this idyllic picture ‘to both
legitimate and illustrate Epicurean ethics’ [206], by showing that the
body’s necessary and natural needs are easily met in a world still un-
contaminated by luxury goods. The utility of an intermediary stage
of village life between the erramento ferino (wandering in the wild)
and the formation of cities, which is unique to Epicurean prehistory,
Brooke Holmes 147

thus becomes clear. For it offers a picture of a simpler time when

Epicurean justice could emerge in rudimentary form, without, for all
that, resorting to a Golden Age: ‘in prehistory we find the Epicurean
theory at its most powerful, stripped of the accretions of culture and
civilization, saving the human race itself from extinction’ [254]. Much
of the Golden Age imagery that one does find in the account of spon-
taneous generation, when people were born from great wombs rooted
to the earth, can be seen as helping Lucretius answer the question
of how these beings survived without mothers or tšqnai (artes): al-
though the image of the Earth as mother, with its overtones of Stoic
allegory, has him ‘skating on thin ideological ice’ [60], he needs the
Earth to give forth a milk-like juice if these first creatures are to
be nourished. Campbell’s attention to Lucretius’ tendency to stress
positive or negative aspects of early human life according to what he
is trying to do at a given moment places him firmly within the trend
of moving beyond the Primitivist-Progressivist opposition.
This sensitivity to Lucretius’ strategies extends from Campbell’s
handling of broad themes to his comments on single words and
phrases. He is alert to how Lucretius manipulates language so as
to downplay external agency (e.g., on exclusae, ad 802 [71]), or con-
versely, how it slips into a teleological idiom (e.g., on crerint, ad 782
[47]). This makes him an ideal reader of Lucretius, that is, someone
whom it is ideal to read with. In this sense, Campbell’s is a more
satisfying commentary than Costa 1984, which covers book 5 in its
entirety. The two commentaries, in fact, work nicely together, since
Costa attends more to questions of grammar, but is more reticent
and conservative vis-à-vis the big issues.
But despite the benefits of working with the text in this format,
it is precisely the strength and cohesiveness of Campbell’s interpre-
tation that makes one begin to wonder why he chose to prepare a
commentary rather than a monograph. Lucretius on Creation and
Evolution occasionally feels like one of the hybrid creatures that Lu-
cretius describes in the zoogony. It can seem as though an argument
is being carried on in footnotes [see below]. On the other hand, while
individual lemmata often give rise to wide-ranging and imaginative
discussions, the commentary can begin to resemble a cabinet of cu-
riosities. It is at these moments that one especially feels that the
format allows Campbell to accumulate information without properly
sorting it out, or simply to wander off. Commentaries, it is true,
148 Aestimatio

are built through the accretion of comparanda. But sometimes this

parataxis can be confusing, if not misleading, especially given that
Lucretius’ dense blend of argumentation and never-innocent illustra-
tion requires careful untangling.
Take, for example, Campbell’s treatment of Lucretius’ denial
of the existence of Centaurs. Lucretius moves from the argument
that men and horses age differently to pointing out that they neither
‘burn with the same passion of Venus, nor come together with a sin-
gle lifestyle, nor find the same things pleasant for the bodies’ [De rer.
nat. 897--898, Campbell’s translation]. This last claim opens onto
the observation that hemlock is great for goats, poisonous to humans.
Lucretius is building on the theory of perception outlined in book 4
whereby different bodies have different pores that determine what
they experience as painful or pleasurable. In his notes on this [154],
Campbell responds to this shift from sexual desire to food with an ex-
cursus on the association between food and sex in Lucretius, which he
claims is underwritten by Greek biological thinking on human seed.
Now it is perfectly true, as Campbell explains, that Aristotle and
probably earlier Presocratic writers understood seed as a residue of
concocted food, and that elsewhere Lucretius directly relates diet
to the quality of seed [De rer. nat. 4.1260ff]. Yet this has little
relevance here, where Lucretius’ easy transition from sexual desire
to food hinges on the word iucunda (pleasures). As Campbell points
out ad 897, Lucretius thinks that seed is stimulated through seeing,
and that the only catalytic objects of vision are members of one’s
own species. In a similar way, certain species gain pleasure from
certain foods, although it should be said that those whose bodies
are incompatible with the food in question may be harmed by it,
whereas the seed of a person looking at Black Beauty is presumably
simply indifferent to equine loveliness. It is not that Campbell is
wrong about ancient ideas about the role of food in the production
of seed. But his digression on this topic obscures what really matters
here, namely, the Epicurean lock-and-key perception theories that
can explain both species-specific desire and some species’ pleasure
in foods that are poison for other species. As it stands, the notes
on food and seed seem better suited to the material on diet and
seed in book 4, or, even better, to the discussion of food (pabula)
on p. 117, where, in fact, they would have been quite helpful. It is
not that valuable information is missing: the relevant comments on
Brooke Holmes 149

seed—that there is no viable means of reproduction without visual

stimulation—are made ad 897, while the explanation of the lock-and-
key theory of perception is available ad 899--900. It is simply that
there is so much unorganized information that the notes no longer
clarify Lucretius’ argument. I should say that Campbell’s notes are
chock-full of interesting information, and that they are occasionally
very funny. Indeed, one is often content just to wander with him. It
is simply that at times the line between erramento and error seems
a bit too fine.
While, in principle, Campbell is looking backwards and forwards
equally, from Lucretius’ predecessors to his Nachlëben and our own
conceptual habits, in practice, the fact that these habits are shaped
by modern evolutionary biology means that they have a special claim
to truth. As a result, the motivation for introducing modern evidence
is often ambiguous. Here again one feels that the commentary format
is problematic in that it allows Campbell to remain less than forth-
coming about his own agenda, especially vis-à-vis Lucretius’ anthro-
pology. Campbell reasonably argues that clarifying our own, heavily
Darwinian ideas about what an anti-teleological story of the creation
of the animate world should look like enables us to understand bet-
ter the specific mechanisms of Lucretius’ system. Lucretius, for ex-
ample, does not accept mutation at the genetic level and seems to
accommodate the inheritance of acquired characteristics, at least in
humans [7--8]. At the same time, Lucretius’ difference from Dar-
win sometimes seems presented in such a way as to make him seem
more cutting-edge: his view of species as fixed and bounded entities,
for example, allies him with post-Darwinian notions of species stasis
[124--125]. In such a context, the use of modern parallels appears to
be less about crystallizing our own preconceptions and more about
vindicating Lucretius. On page 221, for example, Neolithic archeol-
ogy is enlisted as support for Lucretius’ picture of moderate violence
among early humans over and against Moschion’s more lurid account
of cannibalism. Ancient bones again work to bolster Campbell’s read-
ing of the role of cooperation in human evolution [280--281] (‘L. is
quite correct to place pity for the weak in prehistory’), although it
should perhaps be noted that the solidity of the bones cannot be
extended to the inferences drawn from them that are used to prove
Lucretius correct. Different things appear to be at stake in differ-
ent comparisons undertaken either casually or at some length in the
150 Aestimatio

commentary, without these stakes always being made clear. More-

over, the interest in showing that Lucretius anticipated the work of
modern science can dull Campbell’s often incisive readings of the
dense nodes of myth, Epicurean philosophy and rhetoric that are so
characteristic of book 5.
In the second half of this review, I want to take a look at the
most ambitious part of Campbell’s interpretation of Lucretius’ evo-
lutionary narrative. Campbell argues that Lucretius’ claims that
humans are the only species to have evolved; that they evolved into
cooperative beings; and that, in doing so, they ensured the survival
of the species and the definitive break between the human and the
animal. How does this work?
Lucretius’ account at 5.1011--1027 of the ‘softening’ of human
nature and the formation of the first communities is notoriously el-
liptical. In recent years, much attention has been paid to how his ap-
parent citation of the principle of Epicurean justice—neither to harm
nor be harmed (nec laedere nec violari)—participates in the story be-
ing told here. Campbell sees the entire passage as a turning point in
the evolution of the species: whatever changes human nature under-
goes here are, henceforth, passed down as inherited characteristics,
making Lucretius a Lamarckian. Yet, in part because Campbell’s
argument is carried on disjointedly across lemmata, it is unclear how
the external environment provokes these changes to human nature
and what these changes are. Some of this vagueness is due to Lu-
cretius himself. At 5.1011, without obvious motivation, Lucretius
introduces a new stage of human development in which humans have
houses, clothing, fire, and nuclear families. From this point on, they
begin to soften (mollescere). Spending time indoors, their bodies can
no longer bear exposure to the elements. Venus has a hand in this
softening, and children break the arrogant natures of their parents
with their ‘winning ways’. Somehow this process creates the condi-
tions for humans to establish pacts with one another neither to harm
nor to be harmed, and to set up the principle of pitying the weak,
which Campbell calls altruism. It is this ‘somehow’, of course, that
matters. Also of crucial importance is how we understand what has
been gained in this development.
Campbell knows what he wants to show, namely, that ‘instead of
being a woolly-minded pipedream, the Epicurean theory is the most
 Brooke Holmes 151

pragmatic and realistic approach to justice’ [281], and that Lucretius

anticipates post-Darwinian work on the evolutionary benefits of co-
operation: ‘our co-operative ability is thus the feature that defines
our humanity, and enables us to survive’ [262]. Lucretius’ presen-
tation of the Epicurean theory, as Campbell understands it, accords
with recent refinements to the model of the prisoner’s dilemma which
was first analyzed in an Epicurean context by Nicholas Denyer [1983].
The basic form of the prisoner’s dilemma takes two prisoners charged
with the same offense. Each is offered a deal by the police chief: if
both cooperate, they each get a year in jail; if one defects by con-
fessing and betraying the other, he gets off scot-free, while his coun-
terpart (the sucker) receives ten years. 4 Conventional wisdom once
held that the rational choice would be defection. When the dilemma
was translated into Darwinian terms, it became hard to see how any
form of cooperation could be plausible in the theater of ‘survival of
the fittest’. As a result, cooperation could only be explained as self-
sacrifice for the good of the community. Campbell introduces the
challenges to these assumptions posed by the research of Robert Ax-
elrod, who worked with computer models of the prisoner’s dilemma
in the 1980s. Axelrod’s research revealed that it is not competition
but cooperation that proves more advantageous when the game is
repeated. More specifically, a strategy called ‘tit-for-tat’, which al-
ways reciprocates the behavior of its opponents—it betrays when be-
trayed, cooperates with those who have cooperated with it—emerged
as the strongest. Moreover, after some time playing with tit-for-tat,
the ‘suckers’ began to thrive and the defectors became nearly extinct.
Campbell uses these results to reformulate Denyer’s conclusions: ‘the
Epicurean model would achieve the best result both for the individual
and for the group, and the individual gives up no direct advantage
by sticking to the friendship/non-aggression/mutual aid pacts of the
Epicurean theory, but receives a direct personal advantage by doing
so’ [258]. This appears a valid application of Axelrod’s work. In
casting early humans’ negotiation of non-aggression pacts (amicitia)
in a utilitarian light, Campbell keeps Lucretius firmly in the realm of
Epicureanism, where ‘natural’ action is always motivated by a desire
to secure the individual’s pleasure.

4 Campbell omits a third scenario in which both defect and receive five years
apiece.
 152 Aestimatio

However, there are two problems with what Campbell does from
here. First, after applying Axelrod’s work to Epicurean mutual non-
aggression pacts, Campbell immediately rephrases his claim thus: ‘it
is a powerful individual survival strategy that all should pity the
weak’ [258]. He presents such pity as ‘learned behavior’ and gives
it pride of place in human evolution: ‘now that the human race has
evolved, they are able to make the conceptual leap from their previ-
ous conviction [i.e., aggressive competition was the only survival tech-
nique] to “it is fair that all pity the weak”.’ But is this pity the same
thing as amicitia? What I find troubling about this easy conflation
of the two ideas is that the prisoner’s dilemma assumes a community
of equals; the tit-for-tat strategy relies on the opportunity for future
retribution in the case of defection. If it is just a question of the
weaker versus the stronger from the outset, game theory seems irrele-
vant: we would seem to be in the world of Plato’s Gorgias, where the
only answer to Callicles’ law of the stronger would be the one that
Plato gives, namely, that doing injustice is worse than suffering it.
Campbell might claim that in a community which includes defectors,
the weak (i.e., the suckers), and tit-for-tat players, altruism (always
cooperating) still emerges as the second-best strategy. Indeed, he
seems to suggest this on pages 277 and 280. But even if a player is
programmed to cooperate always within the game, the rules of the
game still assume the conditions of total reciprocity, that is, that the
weak, at least in theory, have the power, say, to lessen my prison
sentence. In any kind of pragmatic situation, if a player is by defi-
nition weaker, the stronger has no reason not to cooperate with his
equals and dominate the weaker. 5 Thus, asserting that the principle
‘pity the weak’ is strictly utilitarian hardly seems valid. I stress this
because, although Campbell acknowledges that ‘there does seem to
be a huge conceptual gap between the pacts “neither to harm nor be
harmed” and “it is fair that all pity the weak” ’ [277], he nevertheless
groups amicitia and pity for the weak together as cooperation. To
classify pity for the weak as cooperation is a bold move, since coop-

5 It is worth pointing out that Axelrod seems only to admit strict altruism
in cases of kin-relations, where self-sacrifice can be understood in terms of
propagation of the gene pool [1984, 88--89 and 134--135]. At the same time,
he does speculate that it is cooperation within kin groups that leads to the
adoption of cooperative strategies outside of kin relationships, a move that
Campbell reproduces [see above].
Brooke Holmes 153

eration is the human trait that Campbell argues protects the species
from extinction, but it does not appear justified by the argument.
Yet it is clear that there must be some connection between the
softening of humans, the formation of friendship pacts, and pity for
the weak. The second problem with Campbell’s account is that he
is not explicit about how this connection works in Lucretius. Nev-
ertheless, with a little work, his reconstruction of Lucretius’ argu-
ment can, I think, be discerned. Campbell recognizes that any evo-
lutionary change must come from the environment. This means that
the softening of human nature in response to key aspects of the do-
mestic environment—the warmth of the fire, shelter, marriage, and
childcare—enjoys some claim to priority as a cause for all future
changes. What gives rise to this shift towards domesticity is not ev-
ident; Lucretius may be silently assuming that the vulnerability of
humans to other beasts (or at least some humans to other beasts)
could have driven them to extinction, had they not formed families
and communities. In any case, Campbell argues that it is because
men become softer in their relationships with their wives and their
children that they behave favorably to their neighbors: ‘the amicitia
does not seem to arise from utility as in Vat. 23 but from a more
spontaneous and more nearly altruistic motive. The results are prag-
matic’ [273]. Thus, he understands the softening of human nature
as leading directly to the gain of reasoning (logismÒc): the argu-
ment would be that once the benefits of cooperation, initiated not
for utilitarian reasons but as a result of the softening of human na-
ture, were seen within the family, humans were then moved to forge
relationships of amicitia with their neighbors. This would explain
why family life is so important to the account of justice: it generates
the conditions under which humans learn the benefits of cooperation,
whatever we take these to be—again, Campbell is vague on this point,
as well as on how these benefits map onto the benefits of coopera-
tion within the prisoner’s dilemma model. The final stage would be
something that looked like altruism: the first humans ‘are pictured
extrapolating from the lessons they have learned at home, and ap-
plying the results to the women and children of other families in a
positive development of the negative nec laedere nec violari’ [277].
In this version of the argument, pity for non-kin weak thus gets
a more complex explanation than Campbell puts forth in his original
continuous version of the argument [see 252--261], where pity is regu-
154 Aestimatio

larly just substituted for amicitia. Or rather, it gets two and perhaps
even three explanations. On the one hand, Campbell recognizes the
difficulties posed by his earlier inclusion of women and children in
the prisoner’s dilemma: ‘the weak cannot strictly engage in the first
part of the pact nec laedere (do not harm), and so this transcends the
basic non-aggression pacts’ [277]. His solution is to make the soften-
ing of human nature the cause of both the non-aggression pacts and
pity for non-kin weak. In other words, pity for non-kin weak bears
no direct connection to the non-aggression pacts, although both have
their roots in the formation of nuclear families. Such a position makes
Campbell ’s vagueness about the ‘lessons. . . learned at home’ all the
more troubling, since these lessons can no longer been seen as derived
from the utilitarian benefits associated with the non-aggression pacts.
It seems likely that Campbell is close to the Lucretian position
when he says that the softening of human nature ‘is not strictly driven
by utility, and the value of such a psychological change becomes
clear only later’ [272]. In fact, Campbell has not shown the place
of utility at all at this stage of evolution: the utility associated with
his re-reading of the prisoner’s dilemma only becomes clear when
non-aggression pacts are formed. What this means for his argument
is that, first, some fundamental psychological change unrelated to
the intellectual perception of benefit is the only explanation for the
development of amicitia and any utilitarian justification comes later.
It is thus unclear, if ‘the first friendship pacts had to be learned
intellectually’ [278], what experience the first humans are learning
from. Second, since at this stage of the argument Campbell no longer
relates pity for non-kin weak directly to the non-aggression pacts and
the benefits of cooperation that they reveal, the relationship between
utilitarianism and altruism falls apart. Pity for non-kin weak may be
an extension of intrafamilial cooperation, but we do not know what
that is. We are left to conclude that pity is an acquired characteristic
derived from the softening of human nature that never enters into
the utilitarian calculus.
But on the other hand, Campbell is clearly attached to the idea
that altruism for Lucretius holds an evolutionary advantage and is
an important part of the cooperation that saves the species. Having
recognized that women and children are technically out of place in
the prisoner’s dilemma since they do not have the power to harm,
he nevertheless reintroduces the prisoner’s dilemma in an attempt
 Brooke Holmes 155

to reestablish the utilitarian pedigree of altruism. This return to

the earlier arguments seems to offer an explanation of pity that is
incompatible with the one in which altruism and amicitia are parallel
developments arising from changes to the psychological makeup of
human beings. It may be that Campbell understands both altruism
and amicitia as independently revealing their utility over time. But
then we are back where we started: how can we evaluate the utility of
altruism if its benefits are not those derived from cooperation within
a community of equals? If this utility lies elsewhere, why introduce
the prisoner’s dilemma to explain altruism?
There is still a third explanation of pity lurking in Campbell’s
text, one that suggests yet again the underlying problems with the
utilitarian reading of altruism that he offers. On page 278, Campbell
shifts gears and embeds altruism once again in the development of
amicitia, rather than allowing it to develop directly from the nuclear
family. On this view, pity is the extension of the non-aggression
pacts to unequal power relationships with other men’s women and
children. 6 Here, it is as if having learned the benefits of cooperation in
the quasi-political sphere, men no longer want to exploit their power
advantage even when reciprocity is out of the question although, if
domination is taken to be a good thing in itself, pace Plato, aggression
would yield more gain than cooperation with the weaker. Campbell
gets around the problem of why men stop dominating the weak even
when there is no advantage in their restraint by reintroducing the
theory of acquired characteristics. Cooperation is thus ‘learned be-
havior’ which is ‘passed down to offspring as an instinctive response
to women and children’ [278]. That Campbell offers this interpreta-
tion would seem to be a tacit acknowledgement that it is difficult to
construe pity for the weak in utilitarian terms, or at least utilitari-
anism as it appears within game theory. One also wonders whether
pity is the result of a characteristic acquired during the process of
domestication or if it comes after the development of amicitia. Most
importantly, while the position that pity for non-kin weak is an ‘in-
stinctive response’ solves the problem of fitting altruism into the
prisoner’s dilemma model, it moves decisively away from Lucretius’
text, where amicitia and what Campbell calls altruism are closely

6 This is sometimes called the ‘associationist’ argument, which is treated at

length in the context of Epicurean friendship by Philip Mitsis [1988].
156 Aestimatio

linked. As a result, altruism disappears from the species-saving co-

operation that Campbell thinks is so important to Lucretius’ concept
of the human.
Campbell’s overall argument has its merits, and one wishes it
had been laid out more systematically at some point. His practice of
referring the question of pity for the weak to Axelrod’s work on the
prisoner’s dilemma seems to reflect his own desire to make Lucretius’
definition of the human conform to what one suspects is his own.
Unfortunately, his use of game theory distracts him from the text
itself. It would seem that Lucretius is telling two stories in these lines.
One certainly has to do with justice. Equally important, however, is
the fixation on children, which cannot be fully explained by staying
within the parameters of the debate about justice. As Campbell
himself observes, when Lucretius talks about pity for the weak, he
‘deliberately go[es] beyond what Epicurus would consider allowable
in giving only one possible cause of justice’ [276]. Excessive, too,
is 5.1027, a verse on which Campbell is uncharacteristically silent.
Lucretius asserts that the human race would have been destroyed
without concordia, then adds that ‘nor would the offspring have been
able to prolong the race to this day.’
Another oddity is the expression ‘muliebreque saeclum’ (‘female
race’) at 5.1021, found nowhere else in Latin literature. Campbell
insists that Lucretius does not see women as a separate race or species.
He relates ‘muliebreque saeclum’ to similar phrases (e.g., ‘muliebre
secus’) in other authors, although ad 853 he has a long note on
‘saeclum’ where it, with ‘genus’, is clearly Lucretius’ word for species,
whether human or animal [e.g., 5.791, 1169, 1238]. And ‘muliebre
genus’ is found at 5.1355, in a discussion of the first forms of clothing.
Lucretius tells us there that Natura forced men to weave first, since
the genus virile excelled in ars and was far more clever than the
race of women. Not a separate species, then, but also, perhaps, not
entirely at home in the genus humanum. One begins to wonder
whether all the talk of human nature in both Lucretius and Campbell
forgets a difference embedded in the text that needs to be acknow-
ledged if these lines are to seem less opaque.
My own suspicion is that there is a bit of a conceptual lacuna
in the erramento ferino, which exerts pressure on this subsequent
phase. Lucretius goes to great pains, as we have seen, to account
 Brooke Holmes 157

for how the first humans survived by positing a surreally maternal

earth. But this earth becomes hard and brutal after the early days
of spontaneous generation. We know too that care of the young is a
pressing object of concern for Lucretius: as I noted above, earlier in
book 5, he argues against teleological cosmogonies by enumerating
the earth’s faults, foremost among which is its hostility to the human
infant cast out from its mother’s womb. So what is going on in the
period when the earth has turned cruel but families have not yet
formed? At 5.1011ff, children re-enter the picture because men and
women join in marriage. A line is probably lost between 5.1012 and
1013, and it might have shed light on the implications of this move to-
wards the nuclear family. It is intriguing that the manuscripts have
cognita sunt—although Campbell accepts Lachmann’s emendation
(coniubium) and argues against the lacuna—which anticipates the
intellectual act implied by the next words: ‘and they saw the chil-
dren created from them.’ ‘They’, as Campbell rightly notes, must be
the fathers, an indication that this ‘human’ story is, in fact, being
told from the perspective of one sex, whose acquisition of reason may
rely more on the inference of paternity than on the perception of any
intrafamilial cooperation. For I think Campbell is right to under-
stand the verb ‘to see’ (videre) to mean that the fathers realize that
the children are their own, that is, that they are created ex se (from
themselves). This recognition, combined with fire and love, prepares
the way for amicitia and pity for the weak. But how?
In recent years, it has been argued [see van der Waerdt 1988]
that later Epicureans modified Epicurus’ denial that there could be
any natural affection between humans—even between children and
parents 7—by adopting some version of the Stoic concept of o„ke…wsic,
which Long and Sedley [1987 I.351] translate as ‘affectionate owner-
ship’. Campbell accepts these arguments in order to make o„ke…wsic
a factor in the development of pity for non-kin weak [277--278] in
the third explanation which I sketched above. Yet it would make
more sense to locate o„ke…wsic at the moment when fathers recognize
their children as their own and take on wives, especially since Stoic
o„ke…wsic begins with, in addition to a desire for self-preservation,
love of offspring.

7 Epicurus advised the sage to stay single, marriage being full of ‘inconve-
niences’.
158 Aestimatio

If we understand the protection of wives and children as now

implicated in the well-being of the autonomous male individual [cf.
Homer, Od. 9.114--115], we can see that pity for the weak may fall
under the logic of a mutual non-aggression pact without being sim-
ply a knee-jerk extension of it or a parallel development. It is not
completely clear whether the weak are threatened by wild beasts or
other humans (already in the erramento ferino, sex is marked by ei-
ther mutual desire or male violence, and perhaps by the idea, too,
that women need to trade sex for food), or by both. If it is other
humans that are the problem, men may entrust (commendarunt)
women and children, presumably to one another, in these pacts on
the basis of a shared vulnerability—I will not harm your family if
you will not harm mine—thus restoring the conditions of reciprocity
and the promise of retribution required by game theory. This may
explain Lucretius’ emphasis on the importance of man-to-man com-
munication in the passage on altruism (‘signing with cries and halting
gestures that it is right for all to pity the weak’). But if we at least en-
tertain the possibility that the danger lies with wild beasts, pity then
encompassing a commitment to the protection of wives and children
from these beasts, we might get a sense of why Lucretius’ conclusion
to this passage is so apocalyptic.
If Lucretius so readily groups the female race with children as im-
becilli (the weak and the helpless), it is not just a question of how this
‘race’ ever survived, but how children managed to. For, presumably,
women were solely responsible for childcare before. Acknowledging
this tension between the need to offer protection to women in the
first communities and the earlier assumption of gender-neutral self-
sufficiency sheds some light on 5.1027, then, where the worry is not
only that the human race will be wiped out, but that the offspring
will not be able to lead the species (saecla) into the future: to re-
produce, they have to survive their traumatic and vulnerable early
years. There may be a sense here that in such a hostile world, women
and children need men, that the future of the species required fathers
once Natura herself turned cruel.
The Garden, Epicurus’ philosophical school, was revolutionary
in its acceptance of women, and Epicureanism may, as a philosophi-
cal system, leave more room for ‘feminine principals’, as Campbell’s
 Brooke Holmes 159

mentor, the late Don Fowler [1996], has argued. 8 Campbell suggests
that the very process of softening may be taken as the ‘feminiza-
tion’ of the human race. But this only reminds us that what matters
here is the evolution of men [see 267--268 ad 5.1014], whose softening
creates the impetus for ars, undertaken on both their own behalf and
that of a genus that is far less clever. Lucretius, it is clear, excludes
the saeclum muliebre from justice and the political at the origins of
these institutions, and, as I have suggested, inscribes the male pro-
tection of women and the paternal oversight of the family line into
the order of necessity. This order of necessity becomes, for Camp-
bell, the very conditions under which the ‘human’ emerges as such.
At the very least, it is worth observing that when Campbell talks
about the need for humans to learn cooperation to survive, he, like
Lucretius, is talking about a community of men. Thus, the awareness
of difference within the genus humanum is not just some politically
correct orthodoxy. Nor does it mean that we have to reinstate a
binary opposition, this time between Lucretius the misogynist and
Lucretius the feminist; for like Primitivism and Progressivism, these
terms conceal more than they reveal. Rather, it is a question of
marking complications that seem to disrupt the text itself. In mim-
icking Lucretius’ own conflation of human nature and male nature
by reducing everything here to the prisoner’s dilemma and equating
political cooperation with what is human, Campbell’s reading falls
short of full engagement with a complex text.
In short, I rather like the idea that Lucretius could have submit-
ted the winning entry in Axelrod’s first iterated prisoner’s dilemma
competition. My concern is that if we focus on whether Lucretius’
idea of human nature, and particularly that which makes us human,
is validated by modern views of utilitarianism in an evolutionary con-
text, we may miss how his story is implicated in the dynamics of the
poem and its assumptions, something to which Campbell is sensitive
at other points. Moreover, such a strategy fails to recognize that
even if the conclusions of modern science and Lucretius coincide, it
still matters how they arrived at those conclusions: surface similari-
ties betray fundamental differences in the production of knowledge.

8 While Campbell’s citations are often extensive, in noting Fowler on the

‘feminine principal’, he fails to cite the article that sparked it, Nugent 1994,
which presents a less positive picture of gender in Lucretius.
160 Aestimatio

Finally, and perhaps most importantly, Campbell seems to take the

evidence of science—evolutionary biology, game theory, prehistoric
archaeology—as the gold standard for determining what it is that
makes us human, that is, where it is that Lucretius gets it right. I
want to stress that I find Campbell’s revision of Denyer’s account of
Epicurean justice in many ways persuasive. It has changed my under-
standing of this passage and challenged me to rethink it; I have no
doubt it will stimulate further discussion of these lines. Thus, it may
seem perverse to want to restore sexual disequilibrium or a politics
of dominance to a reading of Lucretian prehistory that celebrates
the social utility of cooperation and the ethics of altruism. But it is
precisely the use of modern sources to legitimate Lucretius’ human-
ism (and, perhaps, vice versa) that makes a reading which recognizes
the specificity of the saeclum muliebre end up looking like support
for a socio-biological theory of sex roles and, more insidiously, like
an endorsement of such a theory as a blueprint for defining the hu-
man. That is, the reading which I have adumbrated starts looking
strangely prescriptive. And so, it is not just that I think that Camp-
bell’s fidelity to game theory sets an unnecessary limitation on his
interpretation of these lines. I also find the resulting naturalization
of Lucretius’ story troubling.
Lucretius on Creation and Evolution offers a bold and sophisti-
cated attempt to come to terms with Lucretius’ arguments on evo-
lution in the spirit of the poem’s most ambitious commentators. It
deserves not only consultation but active perusal. I could not agree
more with Campbell’s commitment to putting Lucretius and Epicure-
anism into conversation with the present and with our own attempts
to figure out where humans belong in a world of chance and imper-
sonal necessity. But reading book 5, it seems to me, is an exercise in
tracing the contingency of anthropologies and anthropogonies. We
can locate its contemporary salience in the interaction visible within
it between materialism and the stories which Lucretius tells of what
nature ‘forces’ [e.g., 5.1028, 1354] us to be or become, and in Lu-
cretius’ interpretation of the imperatives which he believes he finds
inscribed into us. As Benjamin writes, ‘It is true that men (Men-
schen) as a species completed their evolution thousands of years ago;
but mankind (Menschheit) as a species is just beginning his’ [1996,
 Brooke Holmes 161

I.487]. 9 In this chronology, the equation of mankind and humankind

remains the least evolved thing of all. 10

bibliography

Asmis, E. 1996. ‘Lucretius on the Growth of Ideas’. See Giannan-

toni and Gigante 1996, 763--778.
Axelrod, R. 1984. The Evolution of Co-operation. New York.
Benjamin, W. 1996. ‘Einbahnstraße’. Pp. 83--148 in T. Rexroth ed.
Walter Benjamin: Gesammelte Schriften IV.1. Frankfurt am
Main. See Jephcott 1996, 444--488.
Blickman, D. R. 1989. ‘Lucretius, Epicurus and Prehistory’. Har-
vard Studies in Classical Philology 92:157--191.
Campbell, G. 2002a. ‘Lucretius 5.1011--27: The Origins of Justice
and the Prisoner’s Dilemma’. Leeds International Classical Stud-
ies 1.3:1--12 (http://www.leeds.ac.uk/classics/lics/2002/200203.
pdf).
2002b. ‘Lucretius and the Memes of Prehistory’. Dis-
cussion Paper 1, November 2002. Leeds International Classi-
cal Studies (http://www.leeds.ac.uk/classics/lics/discussion
/2002dp1.pdf).
Costa, C. D. N. 1984. ed. Lucretius: De rerum natura V. Oxford.
Denyer, N. 1983. ‘The Origins of Justice’. Pp. 133--152 in SUZH-
THSIS: Studi sull’ epicureismo Greco e romano offerti a Mar-
cello Gigante. Naples.
Fowler, D. 1996. ‘The Feminine Principle: Gender in the De rerum
natura’. See Giannantoni and Gigante 1996, 813--22. Reprinted
in Fowler 2002, 444--452.

9 If Benjamin’s own Menschheit, like Lucretius’ genus humanum, functions

as the umbrella term of sexual difference in order to name a single species,
there is a certain honesty to Edmund Jephcott’s politically incorrect English
translation, ‘mankind,’ which lays bare what is concealed by the ostensibly
neutral German and Latin terms.
10 Many thanks to Jim Porter, Gerry Passannante, and Miles Nelligan for
helpful criticisms.
162 Aestimatio

Fowler, D. 2002. Lucretius on Atomic Motion: A Commentary on

Lucretius, De rerum natura 2:1--332. Oxford.
Giannantoni, G. and Gigante, M. edd. Epicureismo Greco e ro-
mano: Atti del Congresso Internazionale, Napoli 19--26 maggio
1993. Naples.
Heneghan, T. 2005. ‘Vienna Cardinal Draws Lines in Intelligent
Design Row.’ New York Times Nov 20.
Holmes, B. 2005. ‘Daedala Lingua: Crafted Speech in De Rerum
Natura’. American Journal of Philology 126:527--585.
Jephcott, E. 1996. ‘One-Way Street’. Pp. 444--488 in M. Bullock
and M. W. Jennings edd. Walter Benjamin: Selected Writings.
Vol. 1: 1913--1926. Cambridge.
Long, A. A. and Sedley, D. N. 1987. The Hellenistic Philosophers.
2 vols. Cambridge.
Mitsis, P. 1988. Epicurus’ Ethical Theory: The Pleasures of Invul-
nerability. Ithaca, NY.
Nugent, S. G. 1994. ‘Master Matters: The Female in Lucretius’ De
rerum natura’. Colby Quarterly 30:179--205
Nussbaum, M. 1994. Therapy of Desire. Princeton.
Rothstein, E. 2005. ‘Enough to Make an Iguana Turn Green: Dar-
win’s Ideas’. New York Times Nov 18.
van der Waerdt, P. A. 1988. ‘Hermarchus and the Epicurean Ge-
nealogy of Morals’. Transactions of the American Philological
Association 118:87--106.
Volk, K. 2004. rev. Campbell 2003. Bryn Mawr Classical Review
2005.06.26.