1. ENGLISH - Atlantis: Myths, Ancient and Modern

Atlantis: Myths, Ancient and Modern

Harold Tarrant

Abstract In this paper I show that the story of Atlantis, first sketched in Plato's Timaeus and

Critias, has been artificially shrouded in mystery since antiquity. While it has been thought

from Proclus to the close of the twentieth century that Plato's immediate followers were

divided on the issue of whether th e story was meant to be historic ally true, this results from a

simple misunderstanding of what historia had meant when the early Academic Crantor was

first being cited as an exponent of a literal ra ther than an allegorical interpretation. The term

was then applied to straightforward stories that were told as if they were true. Iamblichus

argued for a deeper meaning that did not exclude the truth, and Proclus' be lief in an inspired

Plato leads him to assume that a Platonic historia must be true. Hence he misreads Crantor

as having been committed to historical trut h and opposes him to allegorical interpreters.

Scholars have continued to see Crantor as a pr oponent of the historical Atlantis without

adequate examination of the evidence, an indi cation of our own need to preserve the

tantalizing uncertainties of such powerful stories.


I. Introduction

Human beings, it seems, have a natural need fo r myth, a need that ha s not passed with the

advent of the written word, but rather mutated. The advent of modern science, offering

strikingly new ways of explaining the workings of our world, has not overridden the need to

hear a plausible account of how this world began, as if we still need a narrative of its birth in

order to understand how it is. It is rather lik e our need to appreciate details of a fellow

human’s childhood if we are to comprehend the personality that even tually emerges. The

biography is the literary vehicle through which most of us seek to understand persons, and the

biography traditionally gives a chronological account that covers parentage, growth,

development, and complex responses to external events. The story that we require when we

try to understand people is not so different from the story that we require when we try to

understand how our world--our village, our landscape, our city, our nation, or our universe.

Beginnings are in particular need of a story-like explanation, but so too are ends. It is not

just death that remains a mystery, but there is a huge fascination with the demise of great

nations and civilisations, and with the end of th ose great forces that we needed in order to

explain how the present has arisen. We feel the need to know not just that dinosaurs or our

own sub-human ancestors became extinct, but why, and hence by what stages . In the case of

any present order we need to know how and wh y it came to triumph over a previous order.

And triumphant moments can be just as much in need of a story to explain them as moments

of defeat. Such moments include those in which the enemy is evil itself, that power which can

seem both real and immediate, but yet is al ways beyond the kind of cognition that usually

offers us that feeling of immediacy. In all thes e cases humans are particul arly ready to resort

to some kind of story, offering a narrative explanation.

When they work correctly ther e is something special about such stories, which allows

them to offer us inspiration whether or not we accept them as true in any straightforward

manner. It is possible for them to acquire considerable personal or communal importance

without ever being an object of belief in th e strong sense of that word. A community’s

identity can depend upon them without any similar dependence on their truth. They can come

to permeate all aspects of our lives without our having gi ven any rational a ssent to them.

Hence to prove them false, to affirm that they must not be believed or followed, can be to

deny others a fragile rock upon which their lives have come to depend. 


II. Plato and Myth

It is not difficult to paint the history of west ern philosophy, and particularly its earlier phases,

as one in which mythical explanation of the un iverse, of moral forces, and of ourselves is

progressively replaced by rational explanation. The movement from myth to logos stands out

as something important from the Presocratics to Hellenistic Philosophy. But if we continue on

to the Neoplatonists, especially Damascius, a nd to Christian philosophy including that of

Pseudo-Dionysius, we shall find that the pr ovince of what might be called scientific

knowledge is distinctly limited, and fails to shed light on what they most wished to

understand.1 The result is a profound re-mythologising of philosophy, in which the alternative

to myth is not logos but silence. The discursive thought with which we are familiar stops short

of our goal, and the noetic vision that extends at least some distance beyond it usually eludes

us. 2

The Neoplatonists will shortly be relevant to this study, but the ambiguity towards myth

pervades even Plato, and makes him at times the source of profound disagreements over the

appropriate way to interpret him. Neither the total demythologising of Plato, nor an uncritical

equation of everything that he tells us in myth-like speech with Platonic belief yields

satisfactory results. His exercises in myth-like writing include studie s of the origins of

civilisation; of the unive rse; of love; and of types of life. It includes studies of what death

means to us, and of the demise of whole civilisa tions. It includes an account of the origin of

the life of hardships, involving the passing of an old and desi rable universal order, and a

world that even with our own care cannot be saved from degenerating. Plato was aware where

myths were appropriate, and he was aware of their huge importance in shaping the education

of the community, regardless of their truth. 3 A great many myths he considered suspect

because they encouraged dangerous beliefs, wh ile he makes free use of others at various

points of his dialogues.

Throughout the first two millennia of Platonism’s 2400-year history the Timaeus has

been considered the most important of his wo rks, though almost certainly it is not a single

work in its own right. It forms a single dramatic sequence with the Critias from which it was

artificially severed in antiquity.4 The long central monologue of th is work is what has often

been thought of as ‘the Timaeus’, and in it Plato offers us what he calls a ‘likely myth’ that

tells of the creation of the world--both its body and its soul--by a benevolent craftsman, and of

the subsequent creation of human beings and of their physical parts by other divine beings to

whom he entrusts the exercise. The creation narrative involves some extremely sophisticated

explanation of the workings of both ourselves and our universe, but ha s been the source of

endless debate about whether its author ever believed in a te mporal creation such as this

narrative account must inevitably imply. The prevalent view among his followers was that the

time sequence was employed for didactic purposes only, 5 but it seems that denying the

validity of a temporal understand ing of its meaning was to ta ke away something that many

others needed. 6 At least one Platonist was able to deny the literal meaning of the creation

process, while still keeping th e Platonic narrative as the basi s of his account of Platonic

physics, allowing one a range of responses to the Platonic narrative depending on one’s

personal perspective, and thus ensuring the preservation of what is mostly a rather routine

work until modern times.7 It is as if the central feature of a successful myth is our inability to

affirm or to deny its literal truth. 


III. Atlantis

On either side of the cosmological speech of Timaeus lies a combination of conversation and

narrative, and in this case the story narrated is described as a logos rather than a muthos, a

meaningful account rather than a mere story. Th e narration is by ‘Critias’, which is itself a

problem, since the first person of that name who comes to mind is the hated leader of the

régime of Thirty Tyrants (in which case his wo rds might seem less than trustworthy), while

serious prosopography8 would require somebody much older to be telling the tale, presumably

the grandfather of this Critias, a senior member of Plato's own family. Before Timaeus’

cosmology we get the claim that Solon heard in Egypt a tale about prehistoric Athens, which

was then governed according to the principles outlined by ‘Socrates’ in the Republic, halting

the expansionist ambitions of a huge island -state located in th e Atlantic. Ultimately

cataclysmic events overtake both warring parties, with Atlantis being entirely lost beneath the

waves and Athens losing her entire army. After the cosmology we meet a detailed description

of ancient Athens and of Atlant is, an account of how the people of Atlantis grew too arrogant

and incurred the displeasure of the gods, and an enigmatic conclusion: ‘Zeus spoke’. What he

said is left, whether by accident or design, to our imagination, and we never meet the

promised account of the glorious deeds of ancien t Athens that had been what ‘Socrates’ was

promised. If Plato had wished to perplex us, then he could scarcely have done a better job.9

Here, however, I want to tackle a more modern myth, a myth about the Atlantis story,

which was perhaps encouraged in late antiquity by the Neoplatonist, Proclus, but has since

been perpetuated by scholars in tent on preserving for us the mystery of the story. While I

witnessed the destruction wrought by the ts unami that wiped out so many coastal

communities in south-east Asia, I had, with th e Australian Research Council's support, been

translating the very part of Proclus’ mighty Commentary on the Timaeus that dealt with

Atlantis. The tsunami had regularly featured in attempts to explain how a civilisation like that

of Atlantis could have been lost, and such theo ries were given added credibility by the fact

that Critias prefaces hi s treatment of the war between Athens and Atlantis by what passes for

an account by Egyptian priests of how human memory is period ically erased by catastrophic

fires or floods. Volcanic activ ity is not specified, but naturally comes to mind. Like most

Platonists I had been intensely suspicious of any attempt to view Plato’s story as an attempt to

write ancient history based upon some exceptionally reliable evidence from Egypt, since

Plato’s so-called myths are usually constructed fr om a variety of tradit ional materials, woven

together into a new fabric designed to fulf il a specific role within a dialogue, and it had

always seemed preferable to me to assume that the story of the destruction of Atlantis and of

its warfare with some prehistoric Athenian martial state fitted into this category.

Even so, I had tended to take on trust th e commonly held view, orthodox throughout the

twentieth century,10 that there had always been c ontroversy among Plato’s immediate

followers about whether the story was intended by Plato as historical tr uth or as some other

kind of offering: cautionary ta le, or allegory, or perhaps as a philosophically apposite

substitute for the rejected text that had hitherto provided the Greek model for heroic action,

i.e. for the Iliad itself. Believing that Plato’s own circle was as undecided about the status of

the story as we were somehow increased the mystique of this lost civilisation. Aristotle can be

reasonably held to have spoken of Atlantis as a fictional city; the earliest known interpreter of

the Timaeus, working within Plato’s school perhaps only half a century after its founder’s

death, is said to have considered the story to be pure history. This is the modern myth, and it

needs both to be challenged and to be understood.

The greater part of the first book of Proclus’ Commentary on the Timaeus is given over

to a discussion of Plato’s treatment of the Atlantis story and re lated material. When I speak of

related material I mean (1) the account of the story’s Egyptian origins and its transmission to

Solon, Critias the elder and Critias the narrato r; (2) the account the Egyptian priest of

recurrent catastrophes caused by fire and flood, and how they ha ve affected Egypt less than

Greece and other lands; (3) his interpretation of the Phaethon myth in this context; (4) his

account of prehistoric Athens and of its similarities with Sais itself; and (5) his account of the

prominence of this Ur-Athens in the repulsion of the militaristic island kingdom of Atlantis.

The valiant deeds of the ancient Athenians have still not brought an en d to the war when all

its army is wiped out by the most seve re flooding, which likewise causes the whole

civilisation of Atlantis to be engulfed by the Atlantic Ocean a nd lost for ever. The story of

Atlantis needs to be remembered as one that wa s not just about a lost civilisation, but also

about the conflict that occurred between that civilisation and the equally mythical prehistoric

Athens: an Athens run largely according to th e principles set about by Socrates in books two

to five of the Republic. 


IV. The Beginnings of Atlantis-Interpretation

Let us go back to the first stage of interpretation of the Atlantis story to find any hint in

Proclus’ commentary. At 1.190.4-8 we find a link with the earliest phase of the interpretation

of Atlantis, seemingly stemming from Aristotle:

Hence one should not say that the one who obliterated the evidence [elenchon] undermines his

subject-matter [hypokeimena], just like Homer in the case of the Phaeacians or of the wall made by

the Greeks. For what has been said has not been invented [peplastai], but is true.

We know from Posidonius11 as reported by Strabo that Plato had once been accused of

explaining the lack of evidence for his fic tional creations by having his narrative destroy

them: like Homer when he accounted for the destruction of the Achaean wall ( Il. 12.1-33, cf.

7.433-63), another case of a destructive flood engi neered by Poseidon. That Aristotle was the

accuser is strongly suggested by Strabo again at XII.1.36, where his name is attached to this

same Homeric example (though Strabo does not here mention Atlantis). The Phaeacian

example given by Proclus is not in Strabo, bu t involves more destructive waters sent by

Poseidon, this time at Od. 13.149-87, so it must surely have belonged to the same original

Atlantean context. It looks as if there had been an early response to this story, which took it as

something that we should understand at face value, but yet as being a simple fiction. Here in

fact we find something that certainly concerned the truth-status of the story, and deriving from

a stage of interpretation that is not properly do cumented in Proclus. It is likely to go right

back to the most celebrated pupil of Plato, who, if anybody, should have been in a position to

know the status of the story. In any case it must be earlier than the first century BCE.

The first interpretation of the story to which Proclus attaches a name comes in Proclus'

introduction to the story:

Some say that the whole of this tale about the Atlantines is [76] straightforward historia, like the

first of Plato’s interpreters, Crantor. He says that [Plato] was actually mocked by his

contemporaries for not having discovered his constitution himself, but having translated the [ideas]

of the Egyptians. He took so little notice of what the mockers said that he actually attributed to the

Egyptians this historia about the Athenians and Atlantines, which says how the Athenians had at

one time lived under that constitution. The prophets of the Egyptians, he says,12 also give evidence,

saying that these things are inscribed on pillars that still survive. (75.30-76.10)

The passage is full of difficulties. Crantor’s account implies that Plato had to make a positive

decision to attribute the Atlantis story to Egyp tian sources, placing its Egyptian origin in

doubt, and so undercutting its only claim to have been faithfully preserved; yet he seems also

to refer to archaeological evidence for it that was still extant in Egypt. Proclus, seemingly

dependent upon Porphyry' s commentary, understood historia here to imply the truth of the

tale, but was that ever Cranto r’s word, and if so had he ha d that intention? Did Porphyry

himself intend his own report of Crantor’s vi ew to be an argument for historical truth rather

than a straightforward story? The key question here is what historia had meant in the

commentary tradition.

Now we have good reason to believe that for several centuries one’s ability to write

historia was not seen in terms of one ’s ability to research and evaluate historical truth.

Certainly Cicero could discuss the merits of those who engaged in historia without any

consideration of truth and accuracy at de Legibus 1.6-7, speaking rather in terms of whether

their stylistic abilities were suited to the historian’s job. Certainly in non-philosophic writings

uses of the term abound that imply only plau sibility and the absence of fantastic or

supernatural elements, in contrast to the implausibility of myth.13 But there is proof even

within Proclus that historia in the commentary tradition had meant something other than

historical fact:

The historia asserts that Phaethon, the son of Helius ..., veered off course when driving his father’s

chariot, and Zeus, in fear for the All, struck him with a thunderbolt. Being struck, he fell down

upon Eridanus, where the fire coming from him, fuelling itself on the ground, set everything alight.

Such is the account from historia. Upon his fall, his sisters, the Heliades, went into mourning. It is

a basic requirement that the conflagration should have happened (for that is the reason for the story

story’s being told), .... (109.9-19)

It is immediately evident that the term historia here means little more than ‘narrative’ or

'tradition' or ‘traditional na rrative’. The story is called historia, but Proclus still treats it here

as what we know as an aetiologi cal myth: as a tale invented to explain an actual occurrence,

the conflagration. Proclus, however, affirms th at he is here offering here a ‘historical’

treatment of the story, marked by the adverb historikôs at 109.8 that will se rve to contrast it

with a ‘physical’ one ( physikôs) offered by those explaining the physical world, 14 and a

philosophic one (philosophôs) involving loftier hidden meanings.15 The actual conflagration is

the sole happening that Proclus needs to postulate in order for this to be classed as a

‘historical’ treatment! The equivalent ‘minimum historical content’ for declaring Herodotus to

be writing historia would be the postulation that Persians really did invade Greece, and the

minimum for the Atlantis-story to be historia would be the actual destruction of a civilisation

by inundation. As long as the extended story was m eant to be offering an explanation of that

inundation then it would be an historical treatment. This is manifestly less than Proclus

himself meant when he postulated an historical truth behind the story, and is perhaps a trace

of an earlier very different understanding of historia that the commentary tradition had

preserved.

Once historia is seen to pertain more to a type of writing than to its truth-status in the

debates that had preserved Cran tor's position, it is obvious that psilê had nothing to do with

any 'unadulterated' truth as the ‘modern myth’ assumes, but signifies rather a bare

'lightweight' story, with no weighty hidden meaning. And wh en one realises that the

information about Crantor's treatment had been selected to show that Plato's story was a

simple story, without the weightier meaning that it had become fashionable to find there, then

both Plato's reason for writing and the Egyptians' cooperation in finding some relevant

inscriptions point in the same direction: r eaders were not required to discover any hidden

meaning.

Modern scholarship on Plato as a dramatic writer makes it natural to add one further

point. It is not clear whether Crantor had been distinguishing between what Cr itias, an early

candidate for the title ‘unreliabl e narrator’, was offering in his own narration of the story and

its provenance, or whether he was slipping into the assumption that all this is somehow ‘what

Plato says’. Even if it were correct that Critias’ adoption of a Herodotean pose created an

expectation that his account aimed at some significant historical truth, it would still not be

safe to expect that our author is offering a ny such truth. Hence there is no reason to suppose

that Crantor had ever denied the fictional character of the Atlantis story. 


V. The Allegorical Interpreters

The interpretation which appealed to Crantor was opposed most obviously to this one:

Others say that it is a myth and an invention, something that never actually happened but gives

an indication of things with have either always been so, or always come to be, in the cosmos.

(76.15-16)

Interpreters then mentioned are Numenius, Origenes, Amelius, and Porphyry. Amelius had

wanted to explain the story as some kind of allegorical pi cture of the movement of the

heavenly bodies, while the others all felt that it was designed to correspond with wars that

were eternally taking place between better and wo rse demonic or psychic entities within the

universe. ‘Myth’ is for them all a rival ty pe of communication, in which a story with a

temporal sequence of events is consciously composed so as to reflect eternal truths: ‘false on

the surface and true in its hidden meaning’. This contrast is brought out again at 129.10:

... it is neither a myth that is being related nor a straightforward historia. While some understand

the account only as historia, others as a myth, ....

In this second passage the contrast between two types of discourse is achieved without

reference to truth-status, suggesting that it wa s not central to the original dichotomy. In

Porphyry’s day, as can be seen from his miniature treatise On the Cave of the Nymphs (which

like his Atlantis-interpretation is indebted to Numenius, lack of a cr edible surface meaning

could be taken as an important proof that an allegorical meaning is to be sought for. 16 In a

historia one may legitimately expect to find what passes for literal truth, and in its absence it

is natural to look for mythical meaning. 


VI. The Iamblichan approach

A third group of interpreters, beginning with Iamblichus, now tr y to reconcile the other two,

claiming that myth and historia are not incompatible, since a powerful underlying meaning

referring to eternal forces doe s not exclude the possibility that these forces have actually

revealed themselves in the very way suggested by a surface meaning. It would seem that

Iamblichus had indeed assumed that an historia would involve a degree of historical truth, and

he seems to have been particularly dependent on Critias' claim that the story was ‘totally true’

(pantapasin alêthês): something that Porphyry and his circle seem not to have taken seriously.

Even so, the new kind of interpretation offered by Iamblichus and to a large degree by Proclus

concentrates on the deeper meaning, while arguing that we should not despair of the presence

of historical truth. It might have happened (76.17-18), and ‘nothing he said happened was

impossible’ (190.9), as Proclus says in relation to the size and inundation of Atlantis. In fact

one might easily suspect that Iamblichus’ attitude is reflected in Proclus’ claim at 182.1-2:

We must not be sceptical of it, even if one took what is being said as historia only.

This cannot now be using the term historia to mean ‘pure historical truth’, for taking it in this

way would actually exclude scepticism. It could perhaps m ean something like ‘an historical

claim’, but the key thing is that such a claim would have no ‘deeper meaning’. So even at this

point, where Proclus seems to speak on his own authority and his interpretation of Atlantis is

almost at an end, the contrast is primarily one of types of discourse. ‘We should not be

sceptical of it, even if it is the type of discourse to lay claim only to factual truth.’

What I believe has happened is something like this. Iamblichus has not entirely mistaken

the terms of the previous debate, bu t has sought to correlate the terms historia and muthos

with two possible levels of truth. In view of his insight into the possibility of deeper meanings

that do not invalidate the surface meaning, he had introduced a distinction between two types

of mythical discourse, similar in some respects to one which appears later in Neoplatonism.

Olympiodorus (On Plato’s Gorgias 46.4-6) distinguishes a poetic kind of mythical discourse

whose surface meaning is objectionable and dangerously false, from a philosophic kind whose

surface meaning is unobjectionable. Iamblichus had perhaps used the Olympiodoran idea that

philosophic myths, unlike poetic myths, need not have ridiculous surface meanings that

compelled the thinking person to look elsewhere for truth (46.4). It is precisely the

ridiculousness of the surface meaning that ha d confirmed for Porphyry and his Numenian

predecessors that Homer's description of the Cave of the Nymphs was an allegory. 17

Iamblichus was clearly not troubled by his predecessors' finding a deep meaning in the

Atlantis story, but by their in sistence on its being pure (and ridiculous) fiction, a concoction

(plasma) without any merit when understood histor ically. Their mistake, he surmised, had

been in applying methods developed for Homeric myths and the like to the more cautious

tales of Plato, whose myths never contained obvious or objectionable falsehood at any level.18

What Iamblichus was claiming was not that the story of Atlantis had been accurately

preserved in all its details, for he seems to ha ve made no effort to claim any more than its

plausibility, and he attaches no importance to the tortuous account of its Egyptian provenance

as if it did not offer an histor ical argument. Further, no Neopla tonist seems to have set much

importance on the contents of the Critias, which fell outside the Iamblichan curriculum.

Rather, Iamblichus was attacking his predecessors’ hasty assumption that myths must lack

any relation to historical tr adition, and saying that in a certain kind of myth the deeper

universal meaning actually implied that historical events of this kind were likely to have taken

place. So the basic story was plausible, yet still there was no commitment to the idea that the

whole of Critias’ narrative was historically true! Right up to this time, in spite of our ‘modern

myth’, there is no evidence that any Platonist in terpreter in antiquity credited Plato with an

attempt to preserve historical truth. 


VII. Longinus’ Philology and the Contrast of Discourses

The ‘modern myth’ arises from mistaking a di stinction between two categories of written

communication, muthos and historia, for quite another distinction between false narrative and

true. It may help to see why it had been categories of communi cation that were relevant. The

key is the figure who set out to stem the ri sing tide of allegori cal interpretation. Though

Crantor (c. 300 BCE) is the only promoter of historia named in Proclus' pages, one must

postulate at least one literalist to revive that view after Numenius (c. 150 CE) had promoted

allegorical interpretation. Consider the argum ents for the literalist position summarised by

Proclus at 1.129.11-23.

The arguments may be summarized as follows:

• Plato’s remarks at Phaedrus 229d discourage ingenious non-literal interpretations;

• Plato’s methods of communicating doctrine ar e not obscure, like those of Pherecydes,

but in most cases direct;

• allegorical interpretation is only necessary if one cannot explain the presence of an

episode otherwise, whereas this episode is adequately explained by the need to seduce

readers into continuing;

• if one tries to offer a non-li teral interpretation of everyt hing, one will end up wasting

as much time as people who explain every detail of Homer.

Clearly somebody was cautioning against the trend towards allegorical interpretation that had

begun with Numenius. Although Proclus represents these points as arguments for the story

being just historia and not muthos, there is nothing here that would amount to an affirmation

of the story’s truth. No mention is made of Egyptians who can corroborate the account and

point to the relevant inscriptions, nor of travel writers who c ould confirm from local sources

that islands off the west coast of Afri ca had disappeared. These are arguments for historia in

the sense of a narrative to be understood in its simple meaning, not as an allegorical muthos.

We may discern who it was that argued against an allegorical interpretation in this way.

There is just one im portant predecessor of Porphyry’s Atlantis-discussions who is not

mentioned in Proclus’ introductory remarks on the interpretation of Atlantis. This is his early

mentor Longinus, who, in spite of his relative lack of interest in th e Atlantis story that

emerges at 204.18-24 (= fr.37), is nevertheless mentioned more th an any other interpreter in

the detailed discussion (= frs 32- 36), mostly as an object of at tack. His overall view of the

story was never openly stated. Gene rally a literalist, he is attacked there by Origenes in the

following passage:

Longinus raises the difficulty of what the presentation of this narrative means for Plato. For it

hasn’t been composed for the relaxation of the audience nor because he requires them to

remember it. He solved this, as he thought, with the observation that he had taken it up prior to

the physical theory to charm the listener on and to make the presentation an early antidote to the

dryness of its style. Origenes used to say that the narrative had been contrived, and to this extent

he agreed with Numenius’ party, but not that it had been contrived in the interests of artificial

pleasure, like Longinus. (1.83.19-28)

Whereas Origenes accepts Numenius’ kind of manufactured story, that is the manufacture of a

vehicle to carry an encrypted meaning, as worthy of Plato, he cannot accept Longinus’

hedonistic kind: the manufacture of a piece of writing that gives enough pleasure to seduce

the reader into persevering. 19 So Longinus can be said with confidence to have postulated

Plato’s desire to seduce the reader as the reas on for the story’s incl usion, not the momentous

events of hitherto unrecorded hi storical facts, and one must doubt whether his insistence on

taking the story literally had anything to do with a belief in its historical truth.

Once this is understood, the four arguments against finding deeper meaning in it can

readily be attributed to Longinus. They exhibit both signs of his language 20 and his reason for

Plato’s having included the Atlant is episode: to sedu ce the reader into persevering. Remarks

about his views on the Atlantis story confirm th at Longinus was resisting the trends that he

detected in Platonic hermeneutics, making much of some details of language and puzzling

over others. Interestingly he th inks that Plato is not committed to the remarks about Solon’s

poetic status offered by the el der Critias at 21c-d (fr. 34 = 90.18-20), and he goes to great

length to explain away Plato’ s seemingly non-factual remark on Athens’ climate (fr. 35 =

162.15-27). This presupposes that Plato should normally be taken literally.

Longinus was the Neoplatonists’ favourite ex ample of how not to do philosophy, being

too fascinated with literary and linguistic matters. Modes of communication were as central to

his teaching as philosophic lessons. So it was above all Longinus, the literary interpreter, who

promoted the debate about the status of the A tlantis story in terms of types of discourse,

contrasting Herodotean-style narrative with allegorizing myth, and opting for the former,

probably claiming the support of Crantor--who thus becomes the champion for his view! In

due course his pupil Porphyry, to whom we u ltimately owe the account of the contrasting

positions, took over from him this discourse-based contrast. The contrast between types of

discourse never wholly disappears, though a ge neration or so later Iamblichus introduces

greater discussion of literal truth and falsehood. Even then there is no necessary connection

between the term historia and historical truth, and hi s innovation was probably more

concerned with countering the view that all myth, like the myths of the poets, must be literally

false. Plato's myths were not lik e that, and could thus be regard ed as potentially true at the

surface level. Proclus, though ofte n preserving uses of the term historia without any

implication of historical trut h, at times appears to make a stronger connection between

historia and literal historical truth. This is not b ecause he misunderstands the significance of

the term historia. Rather it is because he highlights Crit ias’ claim that the story is ‘true in

every sense’, so that, given the story is meant to have a literal and an allegorical sense, it must

also be true in both of these senses. Proclus’ reverence for Plato makes it difficult for him to

imagine the author encouraging a literal reading without insuring its literal truth. 


VIII. Dénouement

So the ‘modern myth’, as I have called it, is false. Until Proclus in the fifth century CE, there

is no evidence that any of Plato’s followers eith er knew of, or argued for, the literal truth of

even the bare outlines of Plato’s story of A tlantis. Proclus himself, and some others who

denied a deep meaning, may have assumed that Crantor had done so. Though Iamblichus was

content to claim that myths were not in all cases of such a kind as to have no plausible

superficial meaning, by Proclus’ time the vene ration for Plato meant that, where the author

gave appropriate indications, the superficial meaning of a myth had to be historically true. 21 It

seems, then, that the concept of myth f ound in Porphyry and his contemporaries was

superseded by a more complex one in Iamblichus, which was taken still further in Proclus.

Hence he at last finds it necessary to make statements such as that found at 1.190.7-8:

‘For what has been said is not invented, but is true.’

In Proclus’ eyes at least Plato has indeed taken the story of the Atlantine war from historia

(193.16-17), a significant claim even where historia only means the investigation of oral

tradition.

So how does Proclus set about proving that the story of Atlantis was indeed historically

true? The primary piece of evidence seems to be the character Critias’ claim that the tale is

‘entirely true’, interpreted as meaning ‘true in every sense’. This really shows us just how

little evidence that Atlantis had existed had survived into late antiquity. Where are the

relevant Egyptian inscriptions, which should have been enough to clear the matter up if one is

to go by the ‘evidence’ of Crantor as most people have understood him. These stones should

have been famous, but apparently, though they had survived for 9000 years before Solon, they

perished over the next millennium. Though know ledge of the world beyond the Pillars of

Heracles had increased considerably in Roman times, it seems that the more reputable ancient

geographers and natural history writers did not have any site for Atlantis on their map. Instead

of citing ‘reputable’ sources for some historical Atlantis Proclus can only make uncommitted

mention of ‘those who write about the outside ocean’, appealing to the statements of an

unknown Marcellus about a group of ten islands, one about a thousand stades long and sacred

to Posidon, where the people preserve the memory of a huge island of Atlantis th at really

existed, and had long contro lled the Atlantic isles. 22 Wisely, Proclus chooses to make little of

such sources, and even there no mention was made of any contact between such a power and

eastern Mediterranean civilisations.

What we have seen in the ‘modern myth’ is how modern scholars, well versed in ancient

Greek and with all the abilities required to interpret historia in Proclus’ reports on Crantor

correctly, have somehow needed to preserve against the odds a real possibility that the

Atlantis myth is true, at least in outline. They have seized on a phrase in Proclus, failed to

analyse it, divorced it from its wider context, usually assumed that it is Crantor’s own phrase,

and taken it as clear evidence that Plato’s earl y successors read the Atlantis story as being

literally true. We need these myths. We need to bring Atlantis to mind as we sip our Greek

coffee on any one of the many cafés on the rim of the caldera of Santorini, the ancient island

of Thera, long linked to the Atlantis story by J. V. Luce and others. 23 We may need to think of

it also when modern tsunamis devastate communities around the Indian or Pacific Oceans.

But our need for it does not entail that we believe it, but rather (as is typically the case with

myths) that we accept it as part of our cultural heritage. Above all, our need for it does not

mean that it was ever intended to be read as history (in our sense).

We have seen also how late antique Platonists had prepared the way for the ‘modern

myth’ by their own unwillingness to write it off as a story to be read as an allegory alone.

They too needed to keep the possibility of hist orical truth open. Proclu s had also encouraged

us to regard the literal interpretation as quite respectable by juxtaposing literal and allegorical

interpreters as if, before Iamblichus, there had been a genuine hermeneutic dilemma. In a

sense there had been in Longinus’ time, though the choice was not between ‘true’ and ‘false’.

Yet Proclus’ dilemma was essentially contrived, contrived to preserve the great mystery that

has always surrounded the story, and contrived so as to represent Iamblichus’ position as

offering the solution to that dilemma.

Human beings need myths, and philosophers need ‘possible worlds ’. Frequently they

can assist in the understanding of our own. Disc ussion of mythical creatures like chimaeras

and unicorns can assist in our efforts to th ink through problems of a much more immediate

nature. I know little of non-Euc lidean geometry, but I know that I shall never us e it to solve

the real problems that confront me, yet I acknowledge that it will be important to

mathematical theory. I have little confidence in the detail of the theo ries of Freud, but he

offered us a new story of ourselves bringing new analytic tools, and his unquestionable genius

was very much needed in order that the foundations of psycho-analysis should be put in place.

In a sense, Freud took over myth in order to keep us supplied with a similar range of possible

worlds to that which myth had offered.

This paper has dealt with just one example of how sophisticated human beings devise

strange ways of preserving myths that inspire or benefit them. What has become culturally

important, as anything fully qualifying for the term ‘myth’ must have done, will always be

extraordinarily difficult for us to overthrow, fo r the urge to defend it in the face of doubts is

powerful indeed. Even if it has not been through a rational assent, engendering belief, that our

myths have come to have that importance, once others challenge them our need to preserve

them takes over, and what we have long been content to accept as ‘ours’ becomes something

for which we seek a new kind of validati on, usually known as proof. Acceptance becomes

‘belief’, and we fail to attend properly to those who do not believe. What this means for

religion is important for us all. What had once been a culturally imbibed attitude to ourselves

and to the greater forces around us becomes a be lief system that requires, in its potentially

most damaging form, strict adherence to the letter. It is no accident that Proclus, who did most

in antiquity to promote the acceptance of the literal truth of the Atlantis story, also excelled

others in his philosophic efforts to establish th e scientific and systematic truth of the dying

Greek pantheon. 


NOTES 

1 See for example R. J. Mortley, From Word to Silence (Bonn, 1986), 2 vols.

2 See especially Sara Rappe, Reading Neoplatonism: Non-Discursive Thinking in the Texts of Plotinus,

Proclus, and Damascius (Cambridge 2000).

3 Republic 376e-378b, 414b-415c; Laws 712b-714a.

4 For a collection of essays that attempts to correct this, see T. Calvo and L. Brisson (eds), Interpreting

the Timaeus-Critias, Proceedings of the IV Symposium Platonicum Selected Papers (Sankt

Augustin, 1997), especially the contribution of Diskin Clay, ‘The Plan of Plato’s Critias’, 49-54.

On the artificiality of ancient arrangements of Plato’s writings see Harold Tarrant, Thrasyllan

Platonism (Ithaca, 1993).

5 Relevant followers were Speusippus, Xenocrates, and Crantor; on them see John Dillon, The Heirs of

Plato (Oxford, 2003), and on this issue Harold Tarrant, Plato’s First Interpreters (London,

2000), 44-46.

6 A temporal interpretation of the Timaeus was demanded by many, including Plutarch, Atticus, and

early Christian authors; see for instance John Dillon, The Middle Platonists (London, 1977),

252-4.

7 Alcinous, Didascalicus chapters 12-21; strictly speaking, Alcinous (10) has the kind of Aristotelian

first god whom one would not associate with creation processes or temporality of any kind, but

the universe comes into being (cyclically?) after the intellect of the world-soul has been stirred

into response.

8 See the entry for Critias in Debra Nails, The People of Plato: a Prosopography of Plato and Other

Socratics (Indianapolis, 2002); the problem is that if ‘Critias’ is making up his story, then the

account of how it was passed down to him is quite as unreliable, and serious prosopography

becomes irrelevant.

9 Serious Platonists have helped to clarify the problem of late. Of these I have particularly used T. K.

Johansen, Plato's Natural Philosophy: a Study of the Timaeus-Critias (Cambridge, 2004); but

see also C. J. Gill, ‘The Genre of the Atlantis Story’, Classical Philology 72 (1976), 287-304;

and also his ‘Plato’s Atlantis Story and he Birth of Fiction’, Philosophy and Literature, 3

(1979), 64-78; Gerald Naddaf, ‘The Atlantis Myth: an Introduction to Plato’s Later Philosophy

of History’, Phoenix 48 (1994), 189-209.

10 See here Phyllis Y. Forsythe, Atlantis: The Making of a Myth (Montreal 1980), 1; Diskin Clay,

‘Plato’s Atlantis: the Anatomy of a Fiction’, Proceedings of the Boston Area Colloquium on

Ancient Philosophy, 15 (1999), 1-21; also John Dillon, ‘Pedantry and Pedestrianism? Some

Reflections on the Middle Platonic Commentary Tradition’, in H. Tarrant and D. Baltzly (eds.),

Reading Plato in Antiquity (London, 2006), 19-31, at 22-23.

11 F49.297-303 Kidd = Strabo II.3.6; note that the same verbs (plassô, aphanizô) are used by Strabo

and Proclus for ‘invent’ and ‘obliterate’, verbs are tailored to remarks about the status of the

Atlantis story, since Plato himself uses aphanizesthai for the vanishing of Atlantis (25d3) and

plassô for the invention of stories (26e4).

12 The subject to be understood with this was taken to be Plato by Alan Cameron, ‘Crantor and

Posidonius on Atlantis’, Classical Quarterly 33 (1983), 81-91; I prefer to see it as Crantor in the

light of the previous ‘he says’; for more on the details of interpretation see H. Tarrant (trans.),

Proclus: Commentary on Plato’s Timaeus, vol. 1 (Cambridge, 2006), 60-70.

13 See Alan Cameron, Greek Mythography in the Roman World (Oxford, 2004); all indexed discussion

of historia is potentially useful, especially 90-91.

14 Note physikôs at 109.9, physikôteron at 109.31, epibolai physikôterai at 110.22. Domninus (110.1) is

hailed as the best interpreter in this category.

15 Note philosophôs at 109.9, which is not further used of the third approach, whose outline commences

at 110.22; the key word here becomes hypsêloteron (110.23).

16 De Antro Nympharum 2.1-4, 4.1-5.

17 See De Antro Nympharurm 4.1-5.

18 Here it might be noticed that the view of Plato’s character ‘Critias’ which Proclus takes at 70.21-

71.15, presumably following Iamblichus, places him at a higher level than the character

‘Hermocrates’, and goes to some lengths to defend the historical person too. This rosier view of

‘Critias’ is opposed to another, held by unnamed interpreters best identified with Porphyry, that

Plato’s ‘Critias’ is inferior to ‘Hermocrates’ (71.27). I infer that Porphyry and his

contemporaries did not respect ‘Critias’ and hence did not need to avoid having him tell lies.

Iamblichus, however, with a tendency to venerate anybody related to Plato, took a more exalted

view of him and so elevated him above the status of Iamboulus.

19 Longinus apparently supported his view that the story was meant to give pleasure with reference to

the term charis at Tim. 21a2.

20 I refer to the language of literary seduction (psuchagôgia), which had occurred at 59.28 (Longinus fr.

28.20) and 83.23 (fr. 32.5). For this term’s place in literary criticism see M. Patillon and L.

Brisson (eds), Longin: Fragments, Art rhétorique (Paris, 2001), 314-15.

21 Even so, it has to be said that even where there seems to be some confirmation from a travel writer

called Marcellus of there having been a great island power in the Atlantic, Proclus seems not to

set much store by it (177.10-21).

22 in Tim. 177.10-21.

23 J. V. Luce, The End of Atlantis (London, 1970); the book followed closely after the excavation of

Thera by Spyridon Marinatos begun in 1967, as did less well known books by A. G.

Galanopoulos and E. Bacon, Atlantis: The Truth behind the Legend (London, 1969), and James

Mavor, Voyage to Atlantis (New York, 1969). 

==