THE CREATIVITY OF DISASTER

Sarah Iles Johnston

One of the many ways in which Jan Bremmer has contributed to our field
is through his publications on myth, which have provided new models for
thinking about the work that myth does and how it does it. In particular,
I have been stimulated by his thoughts on Greek creation myths,1 and it
is in recognition of this that I offer some thoughts on a related topic.2
I begin by relating a story that comes to us from Pausanias (..–),
who heard it while traveling through Arcadia:
About a stade distant from the city of Caphyae is a place called Condylea,
where there are a grove and a temple of Artemis that used to be called
Condyleatis. They say that the name of the goddess was changed for the
following reason. Some children, while playing about the sanctuary found
a rope, tied it around the neck of the cult statue and said that Artemis was
being strangled. The Caphyans, discovering what the children had done,
stoned them to death. After they had done this, their women became ill and
all babies were stillborn, until the Pythian priestess told them to them bury
the murdered children, and sacrifice to them every year as heroes, because
they had wrongly been put to death. The Caphyans obeyed this oracle, and
they began to call the goddess at Condyleae ‘The Strangled Goddess’, as the
oracle told them to.
This is a strange tale, for several reasons—most obviously, it compels
us to wonder what it means for a goddess to be strangled.3 But in one

1 I think particularly of Jan’s “Canonical and Alternative Creation Myths”, in The

Creation of Heaven and Earth (ed. G.H. van Kooten; Leiden: Brill, ), – and
his “Pandora or the Creation of a Greek Eve”, in The Creation of Man and Woman (ed.
G.P. Luttikhuizen; Leiden: Brill, ), –. Both are now reprinted in idem, Greek
Religion and Culture, the Bible and the Ancient Near East (Leiden: Brill, ), – and
–.
2 A preliminary version of the current essay was presented at the meeting ‘The End

of Everything: Catastrophe and Community in the Ancient Mediterranean and Near

Eastern Worlds’ at Indiana University in October ; I thank Bert Harrill for his
invitation and both Bert and other audience members for their suggestions.
3 This question was brilliantly addressed by H. King in “Bound to Bleed: Artemis and

Greek Women”, in Images of Women in Antiquity (eds. A. Cameron and A. Kuhrt; Detroit:
Wayne State University Press, ), –. I discuss the story as part of a larger group
of similar stories in Restless Dead: Encounters between the Living and the Dead in Ancient
Greece (Berkeley: University of California Press, ), –.
 sarah iles johnston

way, at least, the tale is not so strange after all, for it exemplifies a three-
part story pattern that is common in Greek myth: ) disaster befalls a
community (here, an epidemic of stillborn babies); ) the community
makes a ritual response (here, the Caphyans establish hero cult to the
murdered children and rename their local goddess) and ) the disaster
abates.4
This pattern is so common, indeed, that it seems to have been consid-
ered too banal for comment; scholars seldom acknowledge it, preferring
instead to apply other interpretative models to such myths. One of these
is Alan Dundes’ ‘lack / lack liquidated’ pattern (which is in turn developed
from Vladmir Propp’s ‘Quest’ pattern),5 according to which a deficit or
failure of some kind (here, the community’s inability to reproduce suc-
cessfully) is resolved. Walter Burkert is well known for his application
of this model to ancient Mediterranean myths, demonstrating its useful-
ness, for example, in revealing similarities amongst a group of Greek and
Hittite tales about a ‘disappearing deity’ who must be cajoled back into
action.6 As in the tale of the Strangled Goddess, the deity’s withdrawal or
anger typically manifests itself as a famine or an epidemic of some kind.
A second interpretative model, which was developed by Burkert himself,
is what I will call the ‘sacrifice / restitution’ pattern:7 an individual or com-
munity surrenders something important and in return is promised that
it will receive a greater good. Burkert used this pattern to clarify many
different kinds of Greek myths, but particularly famous is his application
of the pattern to a type of myth that he called the ‘maiden sacrifice’, in
which a marriageable girl is sacrificed in order to guarantee some benefit
for the community. In some such myths, the sacrifice is figured as exactly

4 Although the third part of the pattern is left unexpressed in Pausanias’ version of the

story, it is so commonly included in these stories that we can scarcely doubt it is implied
here; compare the very similar stories of maidens’ deaths that I assemble in Restless, –
.
5 A. Dundes, The Morphology of North American Indian Folktales (Helsinki: Suoma-

lainen Tiedeakatemia, ); V. Propp, Morphology of the Folktale (translation L. Scott;

Bloomington, IN: Research Center, Indiana University, ). See also W. Burkert, Struc-
ture and History in Greek Mythology and Ritual (Berkeley: University of California Press,
), –.
6 Burkert, Structure, – (chapter ).
7 Burkert himself does not give the pattern a name. In calling it the ‘sacrifice / res-

titution’ pattern, I draw on his vocabulary of description (or rather, that of Peter Bing,
the English translator of the book where Burkert most extensively develops this thesis,
Homo Necans: The Anthropology of Ancient Greek Sacrificial Ritual and Myth [Berkeley:
University of California Press, ]).
 the creativity of disaster 

that—the girl is killed. In cult, and in some of the aitia that accompanied
cult, the sacrifice is figured instead as the seclusion and service of mar-
riageable girls in the sanctuary of the deity—a ‘temporary’ or ‘symbolic’
death. Sometimes, both sorts of story attach themselves to a single cult.
The myth of Iphigeneia, a maiden who was sacrificed to guarantee that
the Greek fleet would reach Troy, is associated with the cult of Artemis
at Brauron outside of Athens, but so is the following story, which I take
from the Suda (s.v. arkteumenai S, S, S):8
(Girls) used to celebrate a festival to Artemis (at Brauron) by ‘playing
the bear’ and by wearing a saffron robe, being neither older than ten nor
younger than five, and by placating the goddess. Once upon a time a bear
was wandering about the deme of the Philaidae and was wreaking havoc.
They say that the bear, once it had been tamed, became a companion to
humans, and that a parthenos was playing with it, and when the child
treated it roughly, the bear was provoked and scratched the parthenos. And
they say that because of this, her brothers shot it down. And for this reason
a plague-like illness fell upon the Athenians. And when they petitioned
the oracle, it said there would be a release from these misfortunes if as
payments (poinai) for the slain bear they forced their own parthenoi to
‘play the bear’. And the Athenians voted that a parthenos could not live
together with her husband, until she had played the bear for the goddess.
We can see Burkert’s pattern here: girls are compelled to serve Artemis
by ‘playing the bear’ (meaning perhaps, to somehow imitate a bear,
but in any case dedicating themselves to Artemis’ service and thereby
‘sacrificing’ themselves symbolically); the community is compensated by
the lifting of the plague and, implicitly, by the guarantee that girls will
successfully live together with their husbands after they have completed
their service (‘restitution’). We can also see Dundes’ pattern if we look
hard enough for it: a plague strikes the community (‘lack of health’); the
plague is ended (‘lack liquidated’). But we can see the three-part pattern
that I described, as well: ) disaster befalls a community (plague); ) the
community makes a ritual response (it institutes the ritual of ‘playing the
bear’) and ) the disaster abates.
Certainly, both Dundes and Burkert have contributed to our under-
standing of the role that ‘disaster’ can play in myth and the construction
of ritual. Dundes’ pattern, especially as it was applied to Greek myths by

8 I use this version because it offers the most complete continuous narrative, but other

sources indicate the story was an old one: see K. Dowden, Death and the Maiden: Girls’
Initiation Rites in Greek Mythology (London: Routledge, ), – (chapter ).
 sarah iles johnston

Burkert, has helped us to see that disaster, which constitutes one specific
manifestation of ‘lack’, often drives the plot of a story. Burkert’s pattern
has helped us to see that when loss, which constitutes one form of disaster,
is undertaken purposefully by a community, it can represent the human
contribution that sustains a profitable relationship with the gods. Neither
pattern, however, fully represents stories like the two I have narrated.
Applying Dundes’ pattern to such stories (at least in its simplest form,
which is how it usually is applied to Greek myths) requires us to ignore
the specific mechanism through which lack is liquidated (step three in
Chart One). Applying Burkert’s pattern requires us to ignore what are,
narratively, the stories’ beginnings, which describe the situations that
precipitated the new modes of human / divine exchange (step two in
Chart One):

Chart One
Step one Step two Step three Step four
DUNDES’ Lack
PATTERN Lack Liquidated
Story of Children Plague of New cultic Reproduction
strangled strangle stillbirths actions are is restored
Artemis statue & are strikes instituted (hero
murdered by community cult to children
community & new name
for goddess)
Story of girl & Bear scratches Plague strikes New cultic Successful
bear girl & is killed community actions are marriage &
instituted reproduction is
(symbolic ensured
sacrifice of
girls)
BURKERT’S Sacrifice Restitution
PATTERN

Hard-core structuralists might dismiss steps two and three as being rela-
tively unimportant—as merely individualizing flourishes that distinguish
one manifestation of the pattern from others by tailoring it to the specific
group that told the story. The windless days that precipitate the sacrifice of
Iphigeneia (to take an example of step two in one manifestation of Burk-
ert’s pattern); the specific way in which Jason conquers the dragon that
guards the Golden Fleece (to take an example of step three in a ‘Quest’
story, a type in which Propp’s pattern frequently appears); or the precise
strategy by which Demeter is brought out of hiding (step three in another
 the creativity of disaster 

instance of Dundes’ pattern, discussed by Burkert)9—all of these might

be dismissed as details that are necessary for making the story interesting
to listeners, but that can otherwise be viewed as largely interchangeable
and therefore dispensable. I will suggest, however, that not only does
it distort the stories to dismiss one or the other step—the very fact
that each of the two interpretative patterns includes one of the steps
while discarding the other should signal to us that something is not
working—but also that by paying more attention to these steps—by
in effect combining Dundes’ and Burkert’s patterns into the three-part
pattern that I have described—we will lay the groundwork for enhancing
our understanding of how the Greeks presented disaster as a creative
process in their myths. Chart Two shows the relationships between the
three patterns and the two myths:

Chart Two
Step one Step two Step three Step four
DUNDES’ Lack
PATTERN Lack Liquidated
Story of Children Plague of New cultic Reproduction
strangled strangle stillbirths actions are is restored
Artemis statue & are strikes instituted (hero
murdered by community cult to children
community & new name
for goddess)
Story of girl & Bear scratches Plague strikes New cultic Successful
bear girl & is killed community actions are marriage &
instituted reproduction is
(symbolic ensured
sacrifice of
girls)
BURKERT’S Sacrifice Restitution
PATTERN
NEW [Transgression] Disaster Ritual response Disaster abates
PATTERN

I need to make two more comments before continuing. First, although I

have been speaking of a three-part pattern (disaster / ritual response / dis-
aster abates), Chart Two presents my pattern as falling into four parts

9 Burkert, Structure, – (chapter ). Burkert himself, in applying Dundes’ pat-

tern to certain myths about Demeter, acknowledges the significance of such flourishes
but does not pursue them.
 sarah iles johnston

(transgression / disaster / ritual response / disaster abates). For the sake of

initially comparing the model that I will discuss in this essay with the
models presented by Burkert and Dundes, I temporarily repressed step
one of my pattern, because, as a glance at the chart will indicate, it was
ignored by both of the earlier models. I will return to its significance.
Second, I want to emphasize that I am treating only myth in this
essay. That is, my remarks are not meant to suggest much, if anything
at all, about how the Greeks coped with what I will call ‘real-life’ disas-
ters, such as the plague that swept through Athens during the Pelopon-
nesian War, or the sacking of Corinth by the Romans in the second cen-
tury bce. Myth, after all, is a privileged discourse precisely because it need
not play by the rules of real life: it is pre-eminently useful, among other
things, for entertaining paradoxes—for embracing what would look like
bald contradictions in other venues—and because it can do that, it is
ideal for expressing and sometimes resolving tensions that cannot be
addressed in other venues. As I will discuss later, in fact, this distinction
between ‘real life events’ (or what we often also call ‘historical events’) and
‘mythic events’ may be important when we think about certain kinds of
disaster myths in which disaster is paired with progress and even cre-
ation. It is also, we will see, important for understanding why, in spite of
an obsession with disaster in their myths, the Greeks apparently declined
to ruminate mythically on two particular types of disaster: the end of the
world as we know it, and the end of each individual human—that is, the
existence of human death.

The Three-Part (or Rather the Four-Part) Pattern

The pattern that I sketched above can be found in many Greek myths;
in fact, as I noted, it is so prevalent as to have seemed unworthy of
scholarly notice. Nonetheless, I will begin this section by offering two
further examples of the pattern in order to demonstrate the variety of
forms in which it may manifest itself. I have purposely chosen two stories
that come from different parts of Greece and that are linked to cults that
have different concerns; many other, similarly diverse examples could
have been offered instead.10

10 For reasons of space, I will not give all the sources for each of the myths that I discuss.

I will cite the earliest and / or the most influential narrations of each myth; further sources
for many can be found in T. Gantz’s excellent Early Greek Myth: A Guide to Literary and
 the creativity of disaster 

In each case, community-wide disaster strikes. In each case, the com-

munity responds by performing a ritual—sometimes, but not always, the
ritual requires the surrender of something precious to the community as
Burkert would have us expect. In each case, the disaster is resolved by
performance of the ritual, and the community picks up with life as it was
before:

Chart Three
Step one Step two Step three Step four
Transgression Disaster Ritual response Disaster abates
Story of Children Plague of New cultic Reproduction
strangled strangle stillbirths actions are is restored
Artemis at statue & are strikes instituted (hero
Caphyae murdered by community cult to children
community & new name
for goddess)
Story of girl Bear scratches Plague strikes New cultic Successful
& bear in girl & is killed community actions are marriage &
Brauron instituted reproduction is
(symbolic ensured
sacrifice of
girls)
Story of Worshippers Famine New statue Offerings of
Demeter’s lose statue and created; cult grains and fruit
statue in neglect cult revived are made again
Phigalia11
Story of Dionysus woos All the Annual ritual Madness and
Carya12 Carya, but at maidens in is instituted to suicides stop
her sisters’ the area go honor ‘Artemis
urging, she mad and hang Carya’, who
rejects him. He themselves protects girls in
turns Carya on the tree transition
into a nut-tree that used to be
(carya) Carya

Artistic Sources (Baltimore: Johns Hopkins University Press, ). In all cases, I have
taken care to select myths (or particular versions of myths) that we know to have been in
circulation from a fairly early period—not because I see particular merit in Antiquity per
se, but so as to ensure that I am treating Greek versions of these myths rather than Roman
re-tellings of them. As interesting as it would be to ask the questions that I pose in this
essay of the Romans as well, for the moment I focus on the Greeks.
11 Pausanias ..–. See further Johnston, Restless, –.
12 Servius ad Virgil, Ecl. . (the fullest version); see additional sources and discus-

sion at Johnston, Restless, –.

 sarah iles johnston

Or does it? What scholars have failed to appreciate adequately is

that in each case, the ritual response to disaster—the response through
which the disaster is resolved—requires the institution of a wholly new
cult or at least the re-institution of a cult that had been forgotten. The
very invention of the religious lives of the Greeks, then—the invention
of the practices on which their relationships with the gods are built—
is predicated on the occurrence of disaster. From disasters emerge the
means through which humans can communicate with the gods and
receive benefits from them.
In other words, I am suggesting something different from Walter Burk-
ert’s pattern, according to which a scripted, ritualized ‘disaster’ (for exam-
ple, the annual sacrifice of a valuable animal at a festival or the dedication
of marriageable girls to temple service) guarantees the reestablishment or
continuance of a previously existing benefit (for example, the return of
fertility) or the establishment of a benefit that, although new at the time
to the community in question is not newly invented in human history
(for example, victory in a specific battle that about to take place). The
pattern that I am exploring emphasizes the potential of unscripted disas-
ter (disaster that is unforeseen and un-chosen by the humans who suffer
it) to result in the creation of something altogether new: a new cult, a new
ritual, and thus a new relationship with the gods.
One of the interesting things about this pattern is the inseparability of
‘creation’ and ‘disaster’, or of ‘loss’ and ‘gain’ that it presents. If we were
to try to dissect the stories and read them in a way that such stories
should never be read—as sending a (single) message—we would be hard
put to determine what that message is. On the one hand, the stories
could be understood as articulating an optimistic view of the world: even
the worst of misfortunes will lead, eventually, to something beneficial.
But on the other hand, they could be seen to intimate something much
darker: we could choose to conclude that all progress is predicated on
disaster, that humanity never moves forward without concomitant loss.
The inception of hope is entangled with tragedy; fear is transcended by
the creation of something new. Although these stories (and others like
them) end on a positive note, the world-view that drives them requires
the simultaneous investment of both hope and fear, without assurance
that either is justified on anything but the most transitory level. By the
very fact that they are anchored in the unknown terrain of future—
a terrain that humans can control only provisionally, by imploring the
divine—the twinned emotions of hope and fear provoke the assumption
that something beyond humans is watching, listening and reacting to
them, for better or for worse.
 the creativity of disaster 

This mixed message—this entwining of the roots of divine / human

relations with both disaster and advancement—is evident as well in
another myth. According to Hesiod (Theogony –), if Prometheus
had not caused a disaster (that is, if he had not tricked Zeus into taking
the less desirable portion of meat at a sacrifice and thus precipitated Zeus’
punishment of humanity) then there would be no race of women, created
as a woe to men, consuming their livelihood. However, there would
also be no sacrifice as the Greeks knew it—a ritual that provided the
primary means through which humans and gods could communicate—
and certainly no specific form of sacrificial ritual that allows humans
forever to receive the best cuts of the animal
Prometheus’ role in this story brings me to my next point, and finally
thereby, to step one of my pattern. Prometheus purposefully tricked
Zeus (perhaps with Zeus’ acquiescence). Given this, the story can be
read as an example of Burkert’s pattern (‘sacrifice / restitution’) insofar as
Prometheus, acting on behalf of men, knowingly gives up certain advan-
tages (part of the meat, and in the long run, a life free of women) in return
for what is understood as a greater benefit (the institution of sacrifice and
the communication with the gods that it enables). Similarly, in each of the
cases that I presented in Chart Three, a human transgression precipitated
the disaster that in turn precipitated human advancement (the Caphyans
stoned local children, Brauronian men killed a bear sacred to Artemis,
the Phigalians lost Demeter’s statue and neglected her cult, the sisters
of Carya persuaded her to reject Dionysus). In contrast to the story of
Prometheus, however, the transgressions in these stories were commit-
ted without any expectation of eventual remuneration (it is perhaps no
coincidence that the star of the exceptional story that breaks this pat-
tern is a culture hero whose very name conveys foresight). Rather, the
majority of the myths that fit the pattern suggest that human advance-
ment unavoidably depends not only on disaster, but on disaster caused by
sheer human stupidity or even human wickedness. As a race, we stumble
into the future. The closest that Greek culture ever came to an advocate
for humanity—Prometheus himself—may have used deliberation to set
humanity on a certain path at the beginning of their existence, but noth-
ing indicates that he or anyone else was understood to steer the course of
our advancement after that. When oracles give advice in these myths, that
advice represents an intervention at the latest possible stage, instructing
humans to institute new rituals that will clean up the catastrophes that
their unplanned actions have already created.
 sarah iles johnston

Other Disastrous Aitia

The futurity imbued with both disaster and advancement that these
myths articulate can be understood as a comment on the human con-
dition itself: we are creatures of sufficient independence and delibera-
tion to meaningfully affect the world (in contrast to animals) and yet of
insufficient control and foresight to chart a path that is truly of our own
choosing, good or bad. It is fitting that these myths involve the institu-
tion of rituals and cults; the right to address the gods is also uniquely
human—animals have no such outlet for their hopes and fears (if indeed
they consciously possess hopes and fears) and gods have no need of it.
The pattern that I have presented and the world-view that it articulates
make particular sense apropos cultic aitia, then.
A similar pattern, found in other myths, will expand our interpreta-
tion:

Chart Four
FIRST Step one Step two Step three Step four
PATTERN Transgression Disaster Ritual response Disaster abates
SECOND Divine response /
PATTERN Transgression / disaster New creation
Arachne’s Arachne and her brother Phalanx Athena turns them into spiders
story (pre- commit incest that are consumed by their own
Ovidian young
version)13
Phaethon’s Phaethon dares to drive the Races near the equator are ‘burnt’
story14 chariot of the sun and nearly brown, and so on; Phaethon’s
destroys the earth in the process; eternally mournful sisters become
he crashes and dies (the first?) poplar trees that bleed
(the first?) amber; his mourning
lover Cycnus becomes the first
swan
Meleager’s Meleager angers his mother, who Artemis turns his mourning
story15 causes him to die sisters into the first guinea fowl
who forever cry ‘mele, mele’

13 Scholion ad Nicander Ther. a Crugnola, citing the Hellenistic source Theophilus.

See for this story my “A New Web for Arachne”, in Antike Mythen. Medien, Transforma-
tionen und Konstruktionen (eds. U. Dill and C. Walde; Berlin: De Gruyter, ), –.
14 Probably told by Hesiod fr.  Merkelbach-West; Aeschylus fr. – Radt; Euripi-

des fr. – Kannicht; full story at Ovid, Met. .–..

15 Antoninus Liberalis , citing Nicander; full story at Ovid, Met. .–.
 the creativity of disaster 

Battus’ story16 Battus, who sees Hermes steal
cattle from the top of the peak
where he lives, reneges on a deal
to keep his mouth shut
Niobe’s story17 Niobe boasts that she is more
fecund than the goddess Leto;
Leto’s children (Apollo and
Artemis) kill Niobe’s children
Myrrha’s Myrrha desires her father
story18 Cinyras, seduces him and
conceives Adonis
Hermes turns Battus into a rock
formation in Arcadia that was
known as the ‘Lookouts of Battus’
The mournful mother becomes a
rock formation in Lydia
The gods turn her into the first
myrrh tree, which eternally weeps
sap (myrrh)

In these myths, transgression and disaster often collapse into a single

episode (as in the case of Phaethon). In other cases, the emphasis is so
strongly on the transgression as to virtually exclude a separate, distin-
guishable ‘disaster’ (as in the case of Myrrha); in these myths we might
better call this stage a ‘disruption’ in the orderly running of the world
and the proper conduct of its human inhabitants. In yet other myths of
this type, it is difficult to distinguish the ‘disaster’ from the ‘resolution’—
becoming a spider is a disaster for Arachne but it is also a resolution of
the story.
Indeed, in all myths of this type, resolution comes not through the
abatement of disaster by the institution of cult, but rather through what
I will call the ‘normalization’ of the disaster or disruption through the
transformation of its human subjects into features of the natural world—
animals or physical landmarks. I use the term ‘normalization’ because
the point I wish to emphasize here is that, once the transgressor and
those whom his or her transgression have affected have been trans-
formed, they are no longer at odds with the proper running of the cos-
mos.19 Myrrh trees are ‘supposed’ to exude sap; spiders are ‘supposed’
to be consumed by their offspring (not only do modern arachnologists
tell us this, but Aristotle [Historia animalium b] already said so as
well); certain rocks are ‘supposed’ to exist in certain places; guinea fowl
are ‘supposed’ to cry ‘mele mele’. As humans, the characters in these

16 Antoninus Liberalis , citing Hesiod, Nicander and several other earlier authors.
17 Homer, Il. .–.
18 Antoninus Liberalis  (under the name ‘Smyrna’) citing Pherecydes of Athens;

Panyassis F  Kinkel apud Apollodorus ... Further at M. Detienne, The Gardens

of Adonis: Spices in Greek Mythology (Paris: Gallimard, ; translation J. Lloyd; second
edition; Princeton: Princeton University Press, ), –.
19 My thoughts here are indebted to discussions with my student Karen Ravelli, who

first applied the term ‘normalization’ to the pattern.

 sarah iles johnston

stories broke rules of conduct, but once they were removed from the
human realm and thereby exempted from those rules, they were regu-
larized as part of the world as we know it to exist—the ‘normal’ natural
world.
That way of looking at these myths encourages us to read certain moral
messages into them: for example, someone who has sex with a family
member is an animal, and like an animal must expect maltreatment by
her own family members in turn. Of course, these myths also have the
potential (as we all learn in elementary school) to explain where fea-
tures of the natural world came from. As scholars, we tend to downplay
that potential because it seems superficial and simplistic—and indeed it
is, in the form in which most of us first encounter it as children. But
we need to think again about this aitiological function. If we can tem-
porarily divorce it from the myths’ moralizing functions, then we begin
to see another pattern: the natural world—its physical features and the
animals that inhabit it—is composed of the remnants of human disas-
ter. If fact, if we expand our gaze beyond myths in which the protag-
onist is responsible for the given disaster (that is, myths in which the
action is driven by transgression) and include myths in which humans
meet disastrous ends that are not their own faults, then the natural world
becomes even fuller of such remnants. I offer only two for the sake of
illustration in Chart Five, but the number could be multiplied many times
over.

Chart Five
Human disaster or loss Divine response / New creation
Daphne’s story20 Daphne, desperate to avoid losing She is transformed into the first
her virginity to Apollo, pleads for laurel tree (daphnê)
help
Callisto’s story21 Callisto is raped by Zeus; an To save Callisto from being
angry Artemis turns her into a hunted down by her own
bear when the illicit pregnancy is son, Zeus turns them into
discovered constellations (Big and Little
Bear)

20 Phylarchus FGrH  F , apud Parthenius .

21 Hesiod fr.  Merkelbach-West.
 the creativity of disaster 

A World of Remnants

The Greeks who believed in these myths—or better, to put it as Paul

Veyne has suggested we should, the Greeks who honored this mythic
discourse and used it as a means of thinking about the world on certain
occasions or under certain circumstances—would have found it artifi-
cial to sever disaster from creation, disaster from advancement.22 Vir-
tually everything in the world that surrounded them—plants, animals,
rocks, the cults in which they engaged and the rituals performed by their
daughters—had its origin in an incident that had meant disaster for other
people. Viewed through the lens of myth, both the cultural and the nat-
ural landscape of the Greeks constituted a history of human catastrophe.
We need to pause on that word ‘history’. As I noted in the first section
of this essay, we are accustomed to using the term ‘historical’ in oppo-
sition to ‘mythical’. For us, the ‘historical’ Greeks were those like Per-
icles, Peisistratus, Sappho and Hesiod, who our scholarly standards of
reality tell us actually existed; if we are feeling particularly generous, we
may admit a shadowy figure such as Homer to their number as well, but
we never admit Arachne, Daphne, Phaethon or Niobe. The anonymous
Caphyan children who strangled Artemis’ statue are not considered his-
torical, and neither are the girl who was scratched by a bear at Brauron
and her brothers.
And yet, this is not necessarily the way that people think, when they
are thinking mythically. J.Z. Smith, in his study of an Australian people
called the Tjilpa, makes an interesting observation:
Ancestral time—‘their’ time, ‘then’, the time of the Dreaming—is charac-
terized by a sort of perpetual motion as the ancestors freely transform the
featureless landscape (as often by accident as by design) into its present
configuration. All is fluidity, all is process and change; everything is inde-
terminate, everything is exponential, and, to that degree, might be termed
historical. Our time—the ‘now’ what we might identify as the historical
present—is in fact characterized by a sort of a-temporality from the point
of view of the myths. In an inversion of ancestral times, ‘now’ all is determi-
nate and constant; the fluidity of the ancestors has established the forms of
the present. The ancestral motion has been permanently frozen in stabile
memorials.23

22 P. Veyne, Did the Greeks Believe in Their Myths? An Essay on the Constitutive

Imagination (Paris: Seuil, ; translation P. Wissing; Chicago: University of Chicago

Press, ).
23 J.Z. Smith, To Take Place: Toward Theory in Ritual (Chicago: University of Chicago

Press, ), .

 sarah iles johnston

In other words, a mythic view of time suggests that big things used
to happen; the world used to undergo significant changes that are still
worth talking about. Now, in contrast, nothing of true significance can
possibly occur; humans may do some minor tinkering with their empires
and battles, but the world is essentially finished and real history is over;
we are like flies walking on a marble statue whose form was chiseled and
polished many centuries before.
Similarly, a ‘real’ Greek (that is, a neighbor of Pericles or Plato as
opposed to a neighbor of Arachne or Callisto) probably would have been
surprised to see a new flower emerge from the earth where a slain warrior
had bled, or a new constellation take its place in the sky to commem-
orate the latest of Zeus’ tragically transmogrified lovers—not so much
because such things were strictly impossible but because the time for
them was assumed to be over. The natural world was fixed and stabile. In
general, we hear very little from our ancient sources about direct intru-
sions of the divine into the human world during their own or immedi-
ately prior times—the closest we come are exceptional tales of human
encounters with gods, like the one Herodotus (.) tells about Philip-
pides meeting Pan in the mountains between Sparta and Athens. Notably,
moreover, although our catalogue of responses from the Delphic Oracle
includes many that follow the pattern of transgression / disaster / ritual
response / disaster abates, virtually all of them fall into the category that
Joseph Fontenrose calls ‘legendary’—that is, oracular responses that con-
cern figures of myth or that are retrojected, by their narrators, into a dis-
tant, mythic past.24 A few such responses fall into what Fontenrose calls
the ‘quasi-historical’ period but none fall into what he calls the ‘historical’
period. In other words, ‘real’ Greeks did not expect the Delphic Oracle to
tell them what it had supposedly told the Caphyans, for example; disaster
seldom was expected to lead to a new, divinely designed cult.25
Implicit in all of this is the mythic division of time into two parts,
as Smith has posited for the Tjilpa. There had been a heroic age, when

24 J. Fontenrose, The Delphic Oracle (Berkeley: University of California Press, ).

25 There are some exceptions in later periods—we know that the plague that swept
through Asia Minor in the second century ce led to the Clarian Oracle demanding
that some cities institute new cults: Z. Varhélyi, “Magic, Religion and Syncretism at
the Oracle of Claros”, in Beyond Magic and Religion: Interdisciplinary Studies in Ancient
Mediterranean Religion and Society (eds. S.R. Asirvatham, C.O. Pache, and J. Watrous;
Lanham, MD: Rowman & Littlefield, ), –. But as Robin Lane Fox, Pagans and
Christians (New York: Harper & Row, ), esp. – has discussed in depth, this
was a time when, for a variety of reasons, the Greeks and Romans were once again open
to the possibility of direct divine effects on the quotidian world.
 the creativity of disaster 

people such as Arachne, Phaethon and Niobe were assumed to have

existed and their transformations were assumed to have taken place (this
age sometimes was said to have ended with death of the last great heroes,
during the Trojan War and its aftermath) and then there was the current
age. This way of thinking about time may be one reason that the Greeks
were so fond of aitiological myths: nothing truly creative was expected
ever to happen again, but as a second best one could, at least, meditate
upon the remnants of a creative past that littered the landscape of the
current world and by doing so temporarily raise oneself to the level of
the people who inhabited it. Like Smith’s Tjilpa, a Greek gazing at Niobe’s
rock or a laurel tree could choose to quite literally perceive the physical
landscape as being composed of ancestors’ remains.26

Death and the End of Humanity

Given this way of looking at the world—given the idea that they lived in
an a-historical, stabile time—it is perhaps not surprising that the Greeks
were disinterested in thinking mythically about the largest possible dis-
aster of all: the end of humanity. If existence is now stagnant, if noth-
ing significant will ever again happen within the world, then there can
be no mythic postulation of a time to come when we or the world as
we know it will cease to be. The closest we come is Hesiod’s description
(Works and Days –) of five ages of humanity, which suggests that
the current age of humanity, in which Hesiod’s audience lived, will even-
tually be destroyed, after it has become utterly corrupt. But this narra-
tive, which was strongly influenced by similar stories from the Near East,
never seems to have made much of an impact in Greece outside of philo-
sophic discourse.27 Greek myth typically entertains neither the idea of

26 I have been stimulated in these thoughts by C. Calame, Myth and History in

Ancient Greece: The Symbolic Creation of a Colony (Lausanne: Payot, ; translation
D.W. Berman; Princeton: Princeton University Press, ), –, although Calame’s
exploration of the boundary between periods for the Greeks is somewhat different from
mine.
27 I explored some of the trajectories that this myth took in Neoplatonism and the

doctrines of Dionysiac mystery cults in Ritual Texts for the Afterlife: Orpheus and the
Bacchic Gold Tablets (eds. F. Graf and S.I. Johnston; London: Routledge, ), –
 (chapter ). On the influences (Near Eastern and otherwise) upon Hesiod, and his
possible intentions in creating this myth, see G. Most, “Hesiod’s Myth of the Five (or
Three or Four) Races”, Proceedings of the Cambridge Philological Society  (): –
.
 sarah iles johnston

sequential ages (as does Persian myth, for example) nor the idea of the
world’s end (as does Christianity) because such ideas simply did not fit
into a temporal worldview according to which the ‘historical’ period of
major changes in the world was long over.
We should think as well about the fact that the Greeks had no myth
explaining why humans must die, such as we find in Genesis , or in
a South American story recorded by Claude Lévi-Strauss, according to
which the first humans failed to listen carefully to their creator god and
therefore ended up stupidly embracing Death, who deceptively arrived
among them disguised as a pleasant-looking man in a canoe.28 The South
American story may remind us of a Greek story that is the closest we
do come in Greek tradition to any explanation of death: Hesiod’s story
of how Epimetheus stupidly embraced Pandora in spite of Prometheus’
warnings, and how this mistake led to Pandora’s opening of a forbidden
jar, which contained all bad things for humanity. Previous to Pandora’s
blunder, Hesiod says, there were no evil things (kaka) or arduous labors
(chalepoi ponoi) or difficult sicknesses (argaleai nosoi), the last of which,
Hesiod goes on to say, inflict kêres—a word that we should probably
translate here as ‘death’. But after the jar had been opened, the world
became filled with kêdea lygra—with grievous cares (Works and Days –
).29
The similarities between the two stories are striking: they each offer
an explanation for death, or at least for the diseases that bring death, that
situates misfortune within humanity’s own stupidity. But the differences
are important, too. Unlike the South American tale and many other tales
of this type, Hesiod’s tale does not give death per se a particular emphasis.
The diseases that bring some kinds of death are just one problem among
many other kêdea lygra that Pandora releases. Mythologically, then, the
reason that humans die is not given special notice apart from the myriad
of other problems that humans suffer; it was only the death of particu-
lar individuals—if they happened under the right circumstances—that
might constitute disasters from which something new and significant
might arise, something worth meditating upon.

28 C. Lévi-Strauss, The Raw and the Cooked (Paris: Plon, ; translation J. and

D. Weightman; New York: Harper & Row, ), .

29 On Pandora—and on the fact that we almost never hear of her again after Hesiod—

see Bremmer, Greek Religion and Culture, –.