The Tradition of Hippias' Expulsion from Athens Forrest, W G Greek, Roman and Byzantine Studies; Winter 1969; 10, 4; ProQuest pg. 277

The Tradition of Hippias' Expulsion 
from Athens 

w. G. Forrest 

THE PURPOSE of this essay is to clarify the tradition about, only 
incidentally the facts of, the expulsion of Hippias from Athens 
in 510 B.C. The main points of the story (Hdt. 5.62-65) are firm 

enough. In 514 Hippias' brother, Hipparchos, was assassinated by 
Harmodios and Aristogeiton. Then, if not earlier, the Alkmeonid 
family went into exile, and after an unsuccessful attempt to establish 
or maintain a foothold in Attika at Leipsydrion they managed, 
through the good offices of Delphi, to win the support of Sparta. One 
Spartan expedition failed but a second succeeded, this in 510, and 
Hippias left Attika. But how did the Alkmeonids win Delphic and 
thence Spartan support ? 

After Herodotos it is nearly a hundred years before another voice is 
heard-two voices but they speak as one, and the reference book 
which both had pulled down from their shelves is readily identified. 
It can hardly be coincidence that Demosthenes in the Meidias (21.144), 
written shortly after 350, and Isokrates in the Antidosis of 353 tell 
precisely the same new story; it must be the story put out a year or so 
before by Kleidemos in his Atthis.1 The story was that the Alkmeonids 
had borrowed money from Delphi with which to raise a Spartan 
mercenary army for the liberation. 

Some twenty-five years later Aristotle had a different tale (Ath.Pol. 
19). After the failure at Leipsydrion the Alkmeonids took up the 
contract for the rebuilding of Apollo's temple at Delphi, burnt down 
some years before (in 548 according to the chronographers). By doing 
so they were able to collect an advance payment, which they mis­
appropriated-oO€v €iJ7TopT)aav XPYJJLrXTWV 7TPOS T~V TWV ilaKWVWV 
f3o~O€tav. 

Where did this story come from? The bulk of the narrative in 

1 Jacoby's terminus post quem for Kleidemos' publication (354) is less firm than he believed 
(cf L. Pearson, CP 51 [1956] 258), but a terminus ante ca. 350 is solid enough. 

277 



278 THE TRADITION OF HIPPIAS' EXPULSION FROM ATHENS 

Aristotle is Herodotean. But there are marks of other sources too, 
with Androtion, whose Atthis had appeared in the 340' s, as the favoured 
candidate for much, Hellanikos and/or Kleidemos somewhere in the 
background behind the vague 8TJ!-,onKol, who are credited with a 
variant view of the courage of Aristogeiton at 18.5.2 Now Herodotos 
had also described the taking up of the contract, but continued (5.62) 
with the words oCa BY] XfY'J!-,a:rwv £~ 7}KOVT£S <being wealthy' (not 
o(}£v dm6pTJO'av XPTJ!-'a:rwv, <having thus acquired wealth') and had then 
launched into the complicated rigmarole of temple-building to which 
we must return, not into the simple TTp6S TiJv 'TWV AaK~vwv {301}(}£Lav. 
Aristotle or Androtion (not here Kleidemos) has surely misunderstood 
Herodotos, has confused established wealth with an immediately 
acquired bonus, and it might be thought that TTp6S 'TY]v 'TWV AaK~VWV 

{301}(}£Lall was no more than shorthand for Herodotean detail. But its 
form points far more readily to Kleidemos' version, to the purchase of 
Spartan aid than to the Herodotean persuasion of the Spartans by the 
Pythia. Rather, the story is a composite one-Hero datos (misunder­
stood) for the first part, Kleidemos for the second. 

Finally there is another Atthidographer, Philochoros (FGrHist 328 
F 115), who, through the scholiast on Pindar who quotes him, seems 
to have had Aristotle's (Androtion's?) misappropriation of the money 
but to have followed it with an account of Alkmeonid gratitude, of 
the repayment of the money by rebuilding the 'Tl!-,£vos with £VXaPLO''T+ 

pLa TTA£lova. 
Philochoros is final so far as traceable tradition goes. But there is 

one more document, the scholiast on Demosthenes 21.144, who gives 
briefly what is in all essentials the Aristotle (Androtion)-Philochoros 
story but stops short at the liberation while adding one or two 
extra details: the Alkmeonids received ten talents all told, of which 
they spent three on the temple and the other seven on their mer­
cenaries.3 

So far, then, the Atthis. Basically two stories; common to both, the 

I On Androtion as a source for the Ath.Pol. see e.g. F. Jacoby, Atthis (Oxford 1949) 207-08. 
On his date, FGrHist 234 Komm. Jacoby (p.75) may be right in arguing that Kleidemos 
alone was Aristotle's 8TJp.oT~KOl, but the favourable view of Aristogeiton at 18.5 would fit 
well with Hellanikos' view of the tyrannicides (not that it necessarily conflicts with that of 
Kleidemos); cf infra pp.279 and 281. 

a The scholiast names the Alkmeonid responsible as Megakles, not Kleisthenes, but this 
I suspect may be simple confusion at some point in the tradition with the hero of Pind. 
Pyth. 7 (who might well from the poem earn false credit as a temple-builder), rather than a 
deliberate, still less a respectable claim that Kleisthenes' father was involved. 



W. G. FORREST 279 

hiring of Spartan help; one a hiring with money legitimately 
borrowed, the other with money that should have been spent on the 
temple. Implicit in the second is that the temple was built after the 
liberation; but it is important to note that there is no sign that the 
building as a fact of history in its own right ever entered the authors' 
heads. Most are silent; Philochoros mentions only the 'TJ/LEVO<; and 
EVXapWT7]pta 7TAELava, the scholiast on Demosthenes might even be 
taken to suggest that the temple, or part of it, was built before the 
liberation (the Alkmeonids spent three talents on the temple and 
seven on the mercenaries). 

To Herodotos then. After their failure at Leipsydrion the Alkmeon­
ids took up the contract and completed it more generously than was 
demanded of them. But although they took up the contract as part 
of their scheming against the Peisistratids, 7Tav €7Tt 'TaLaL IlEwLa'Tpa'TLOnaL 

fLYJxavwfLEvaL, Herodotos does not choose to tell us how it helped them. 
Instead, he branches off into a different story (5.63) : WI) eLv o~ at ' AOYJvaLOt 

MyovaLv, the Alkmeonids bribed the Pythia to win Spartan aid. How 
then did their lavish provision of Parian marble instead of the pre­
scribed limestone help them? Only, so far as I can see, if the fact 
of building won the Pythia's gratitude and so prompted her to 
stir up the Spartans in the same way as the bribery was supposed 
to have done. That is to say, Herodotos gives two versions: in one 
generosity, in the other bribery produce the same result, Delphic 
support. 

But there is a linguistic point. WI) cJJv o~ ot 'A(J-T]vaLoL MyoVGLv, 

bribery. eLv o~ is normally 'resumptive after a digression' (cf Powell, 
Lexicon s.v.); but if the story of temple-building was a digression, 
what was the point of it, introduced as it is as a deliberate and impor­
tant move against the Peisistratids, the last and by implication 
successful move? Is it possible, then, to read some adversative force 
into eLv o7}? At 7.142 the Athenians debate the meaning of the oracle 
about the 'Wooden Wall'. at fL'v argue that the Wooden Wall is the 
defences of the Akropolis, at 8' that it is the ships. at eLv o~ 'Tal) v'al) 

MyaV'TEI) had trouble explaining the final couplet. Here eLv o7} is again 
in part resumptive, but it surely has some adversative force as well­
"Some said this ... , others said that ... , however the latter ... " I 
should hope to read something similar into 5.63.1.4 

4 Seven talents would provide a reasonable mercenary force, but three would not go 
far on a temple. Cf Powell, Lexicon, on 7.142. 



280 THE TRADITION OF I-llPPIAS' EXPULSION FROM ATHENS 

If so, two stories. But the second, the story told by ol ) A 87JVa'ioL , that 
the Alkmeonids eXVlTT'€L8ov T1]v flv8l7Jv xp~f1-aaLV, is not so simple as it 
looks. Athenians, and for that matter Herodotos, knew yet another 
version of the whole affair, that the Alkmeonids and the Spartans 
had nothing to do with the liberation whatsoever, that it was the 
work of Harmodios and Aristogeiton. This was, in a sense, the <official 
version' throughout the fifth century and, most importantly for us, 
was given the blessing of Hellanikos in his Atthis at the end of the 
century,5 though it earned the censure of Thucydides at the same 
time. 

Unfortunately it is not easy to see precisely what Hellanikos blessed. 
It would be typical of Greek historiography for him to have accepted 
the facts of 510 but to have discredited them by giving the Alkmeonids 
unworthy motives or disreputable methods-bribery, peculation or 
the like. But this would not have served his purpose. However dirty 
he might make Alkmeonid methods, the result would still have been 
the same, that the Spartans were stirred to liberate Athens; and 
Hellanikos was as anxious to be rid of the embarrassing Spartans 
as he was to minimise the role of the Alkmeonids, as Thucydides' 
petulant insistence on the importance of Sparta clearly shows 
(6.53.3). 

There is only one clue. The Marmor Parium (FGrHist 239 A 45) dates 
both the murder of Hipparchos and the expulsion of Hippias to the 
same year, naming the expellers only as oi )A87Jva'ioL. Now the 
Marmor did use Hellanikos;6 secondly, rather trivially, it does show 
here a marked lack of chronological eXKplf3ELa, a fault of which Hellani­
kos has been accused elsewhere (Thuc. 1.97.2); more important, its 
story does exactly what Hellanikos wanted to do. It is a reasonable 
guess, then, that this was his account, that he simply allowed 514 and 
510 to coalesce, that he named no Alkmeonids, no Spartans, only 

• 'A8 ~ OL 7Jva£OL. 
What follows? Certainly that the simplicity of Herodotos' second 

story has taken a nasty knock. For who now are his 'AB7JvatOL who 
thought that the Alkmeonids had bribed the Pythia? Not the Alk­
meonids themselves; not the mass of Athenians, who accepted 
Hellanikos. We need men who disliked Alkmeonids, disliked the 

5 For details see the full account by A. J. Podlecki, Historia 15 (1966) 129-41; Thuc. 1.20.2 
and 6.53.3-59. 

• F. Jacoby, Atthis p.227 n.5. 



W. G. FORREST 281 

tyrannicides, and who were at the same time rather pro-Spartan; for 
the story goes on to stress Sparta's virtuous obedience to the Pythia's 
command-Ta yap TOV O€OV 7TpWf3VT€PCX i7TotEilvTo ~ Ta TWV av8pwv. 

High praise for an Athenian to lavish on a Spartan. There may have 
existed some such men in Athens, but I cannot believe that they were 
very many, enough to be called by Herodotos 0: 'AIJryva:ZoL. TLV€S 
perhaps, Ot 'A8'Y]vcx'Lot hardly. 

There were, however, some people who had no love after 508 for 
Alkmeonids, who had no interest in Harmodios and Aristogeiton and 
who were definitely pro-Spartan-the Spartans themselves, who in 
about 503 did indeed tell this very story to an assembly of their allies 
(Hdt. 5.90-91): "They had decided to restore Hippias to his tyranny, 
for Hippias had been their friend and they expelled him only because 
it was the God's command. Now that they knew that it was the 
Alkmeonids, not Apollo, who had inspired the Pythia, they must 
undo the wrong they had done." The argument did not persuade the 
allies-little wonder. It was sheer invention to explain an astounding 
volteJace. No more, surely, can it have persuaded ot 'AO'Y]vcx'Lot. We 
are therefore forced to Schweighauser's emendation at 5.63.1: for 
'AB'Y]vcx'Lot read ACXKEOCXLfLoVWL.7 

Five Athenian accounts of the end of the tyranny have become four, 
and that is some gain. But how did the four arise and which, if any, is 
true? 

Hellanikos, perhaps, may be set aside at once. On a point affecting 
Alkmeonid tradition neither Herodotos nor Thucydides ranks as an 
independent or wholly authoritative witness,S and on the face of it 
Hellanikos is as likely to be right. But it is very striking that, although 
Kleidemos was, like Hellanikos, a demotikos and an Atthidographer, 
he should choose on this point to ignore his predecessor;9 that 
although Aristotle (Androtion ?) has not taken his Herodotos uncriti­
cally and has probably taken over some details from Hellanikos 

7 The change produces a slight awkwardness in the then repeated AUK€3uLP.OVLOL at the 
beginning of 63.2. But there is a similar repetition at 1.174.5-6-Ws- uth-ol Kv{8£OL >..JyovaL .•. 
Kvt8LOL piv ... It is to be noted that with this change all evidence for friendship between 
the Spartans and the Athenian tyrants disappears (Ath.Pol. 19.4 is of course merely an echo 
of Hdt.), and the non-Hero dote an motive added by Aristotle, the friendship between the 
Peisistratids and Argos, can be given its proper weight. 

S For Thuc. 'Philaid' connections see e.g. H. T. Wade-Gery, Essays in Greek History (Oxford 
1958) 246-47; for Hdt. and Alkmeonids, Jacoby, Atthis 160-61; for 'philaids' and Alkmeno­
ids, Forrest, CQ 10 (1960) 232ff; for another agreement between Hdt. and Thuc. against the 
tradition, Jacoby, Atthis 187. 

9 Cf. supra. n.2. 



282 THE TRADITION OF HIPPIAS' EXPULSION FROM ATHENS 

(or Kleidemos), on this point he prefers, without hesitation, to fol­
low Herodotos ;10 that Hellanikos in fact seems to have stood 
alone. 

On the main issue of Alkmeonid responsibility, then, Herodotos is 
probably right. But did the Alkmeonids beg, borrow or steal the cash 
-or did they just happen to have it ? And was the temple built before 
or after the liberation? 

It is puzzling that Kleidemos should have forgotten or suppressed 
the contract and the temple-building, but no surprise that he should 
offer the version that he did of the help which Delphi was able to 
give to the Alkmeonids. Mercenaries and the money to hire them 
were common enough fourth-century preoccupations to make it 
very easy for him to jump to conclusions. How else could Delphi 
have helped? Kleidemos lived in an age when revolution by oracle 
was rare, but borrowing from Delphi a familiar enough idea. Indeed 
it may well have been at the moment he was writing that Philomelos 
and his Phokians were rborrowing' Delphi's treasures precisely to 
furnish themselves with mercenary help (Diod.Sic. 15.25 and 27-30). 
Closer to Herodotos, Aristotle (Androtion ?) has nevertheless, I have 
suggested, misunderstood him and thus provided the Alkmeonids 
with a fund of newly-acquired cash; the use they make of the cash, as 
in Kleidemos, is a fourth-century use.ll 

Some scholars have thought that there was more to be said for the 
idea of financial aid and have produced numismatic argument that 
there was such aid. C. Seltman12 believed that he could see in two 
series of late sixth-century coins a striking confirmation of the 
Atthidographers' tales. The first, an electrum series, carried the letter 
[> on the reverse, [> for Delphi, and on the obverse an owl, a bull 
or a wheel, the last being an Alkmeonid badge; the second series, a 
silver issue with the triskelis, another Alkmeonid badge, and on the 
reverse a <1>, for Phokis. But others no less percipient than Seltman 
have read the [> for Delphi as an A for Athens, have pointed out that 
three-legged wheel-borne Alkmeonids are a myth, and finally that a 

10 Supra nn.2 and 7. His substitution of at 'A8TJvaiot (19.5) for Herodotos' slightly deroga­
tory 'A8TJvalwv 'Tatut {3av}.op.€votUt £lvat J}.w8tpatut (5.64) is particularly striking (and smacks 
ofHellanikos; cf. infra p.282). 

11 Mercenaries, of course, existed in the sixth century (cf. Hdt. 1.61.4). but they were not 
quite such common furniture. 

12 Athens, its History and Coinage (Cambridge 1924) 79fT, regrettably followed by Jacoby 
in his commentary on FGrHist 328 II 115. 



W. G. FORREST 283 

series of coins, i.e. something minted over a period, is not the same as a 
once-for-all issue for a specific purpose, hiring mercenaries. They have 
also argued that the <D might stand for Phleious, say, not Phokis, and 
that the triskelis is by no means an exclusively Athenian, let alone an 
Alkmeonid blazon.13 

The Atthidographers, then, merit little credit unless perchance it 

can be shown-what they do not assertI4 but certainly imply-that 
the temple was built after 510, an act of gratitude for services rendered 
by Delphi, not one of generosity to prompt future services. On this 
question the archaeologists do not present a very helpful, united or 
even welcoming look; it is an issue that has raised much bitterness 
and calls for some restraint. 

Briefly, when the French School at Athens excavated the temple and 
the Athenian Treasury, they found two texts to help them, Herodotos 
and Pausanias (10.11.5). The latter made the Treasury a monument 
for Marathon, the former appeared to place the contract for the 
temple after Leipsydrion in 514. The two buildings were stylistically 
some ten to twenty years apart. The case seemed proved.15 Pausanias 
was soon challenged, but Herodotos has remained throughout a 
threat to anyone who wished to argue on stylistic grounds that the 
temple was earlier.16 The unanswerable question is how far those who 
argue for later dates have been influenced, consciously or unconscious­
ly, by what they have believed to be Herodotos' support. 

This is not a situation in which the layman can reach an opinion by 
counting archaeological heads. All he can do is to state the problem 
and record an impression. 

13 For various criticisms of Sehman's views see E. S. G. Robinson, NC 4 (1924) 337-38; 
J. H. Jongkees, Mnemosyne, SER. ill 12 (1945) 87-98; R. J. Hopper, CQ 10 (1960) 242-47. I am 
very grateful to Dr C. Kraay for his advice on this section. 

H Supra p.278. It is important to insist on this since Jacoby, seduced by Seltman, has 
suggested that the Atthidographers might have established the date of temple-building by 
research in Delphic records. It is indeed possible that they might, but to judge from the 
fragments they were not interested in the temple at all; there is an alternative candidate 
to take its place among Philochoros' wXaplart]pla 7T>.dova, the Treasury. 

15 Apart from Jacoby (on Philochoros, FGrHist 328 F 115) the most important study of 
the literary evidence is that of Th. Homolle, BCH 26 (1902) 597ft'. The most thorough 
archaeological treatment is by P. de la Coste-Messeliere, BCH 70 (1946) 271-87. Jacoby 
gives essential references to discussion up till 1953. Some more appear in the following 
notes. 

16 Gisela M. A. Richterin the first ed. of her Kouroi (Oxford/New York 1942) accepted the 
French scholars' date; in the second ed. (London 1960) pp.129-30, she argued for an earlier 
date. It is significant that she announces this change of heart with the reluctant words. 
"Herodotos may have been misled." 



284 THE TRADITION OF HIPPIAS' EXPULSION FROM ATHENS 

We have the evidence of sculpture, of architecture and, indirectly, 
of an inscription. Of the first there remain some elements of the 
pedimental figures, similar in style to-some have said, by the same 
hand as-the Akropolis kore no.681, which, if rightly associated with 
its base,17 is the work of Antenor. For the kore and the inscription on 
the base, general opinion strongly tends towards a date in or around 
the 520's; on the pediment there is much more hesitation, but some 
powerful voices have claimed that it is nearly contemporary with the 
kore.18 As for the architecture, argument seems only to have estab­
lished that in some features the Delphic temple is older, in others 
younger than the Peisistratid temple on the Akropolis. The safe 
conclusion will be that they are roughly contemporary. The Peisis­
tratid temple was not being built after 510.19 

This is slender ground to build on. Nevertheless, from what has 
been said and in view of the fact that the Delphians were collecting 
money for the rebuilding of the temple at least as early as 526 (Hdt. 
2.180), it seems respectable to suspect, perhaps even more likely than 
not, that the contract was made, the temple begun a decade or so 
before Leipsydrion. With that in mind we can turn at last to the real 
question-what did Herodotos mean? 

The contract was entered into after Leipsydrion, after 514. The 
temple was completed, the Pythia's gratitude earned in time for her 
to bore the Spartans with repeated demands to liberate Athens until, 
in 510, they obliged. If consultation was still only once a year at this 
time, how long would it take to bore a Spartan? Or even if it was once 
a month ?20 And how long would it take to build the temple, or 
enough of the temple to make their generous intentions plain ?21 
Herodotos could not have imagined, if he stopped to think about it, 

17 The association has been doubted (see H. Payne, Archaic Marble Sculpture from the 
Acropolis2 [New York 1951] 31 n.Z), but the doubts are well answered by P. de la Coste­
Messeliere, RA 15 (1940) 5-10. 

18 For the inscription, L. H. Jeffery, The Local Scripts of Archaic Greece (Oxford 1961) 75. 
For the sculpture, see Miss Richter, loc.cit. (supra n.16), and most recently Chr. Karouzos, 
Aristodikos (Stuttgart 1961) 46-49. 

19 See, among many, de la Coste-Messeliere, op.cit. (supra. n.15) and Fouilles ik Delphes 
IV.4 pp.259"-67; L. Shoe, Profiles of Greek MOUldings (Cambridge [Mass.] 1936) 34 and 104; 
K. Schefold, MusHeiv 3 (1946) 59-93; H. Riemann, MdJ 3 (1950) 7-39. 

20 A summary of the question: P. Amandry, La mantique apollinienne Ii Delphes (BEFAR 
170, Paris 1950) ch. vii. 

21 As Professor M. H. Jameson points out to me, a few loads of marble on the site might 
be enough. For the order of construction see de la Coste-Messeliere, op.cit. (supra n.15) 
271-87. 



w. G. FORREST 285 

that the temple was begun any later than, say, 520, or completed to 
the point of gratitude later than, say, 515. But in that case the temple 
was almost finished before the contract was signed-Herodotos must 
be wrong. 

But what kind of error was it? I would suggest that that is the real 
question, that Quellenforschung and archaeology are alike a waste of 
time. The answer surely is that it is a story-teller's error and a venial 
one at that. Crudely, one of the two Herodotean limits has to go; 
either the contract was signed before Leipsydrion, in which case he is 
wrong; or the temple was completed after 510, in which case he is 
wrong. There can be no doubt about the right answer. The story 
demands that the temple be finished in time for gratitude to work; it 
does not matter one whit when it was begun. Nor, I suspect, would 
Herodotos much care. For him the sequence building-gratitude­
liberation was sacred; how that sequence was to be correlated with 
Leipsydrion was not worth worrying about, and it may be a sign of 
this that he chooses to use imperfects and a historic present for 
Leipsydrion and the contracting, 7TPO€XWP€€ ... 7TpouE7TTawv ..• 
IUu(}ovvTaL, thus very slightly blurring the chronological relationships, 
which are brought into sharp focus again by the aorists in which the 
temple is completed, €gEpyauavTo ... €g€7TOL7Juav. Had he been prone 
to the pluperfect he might have kept historians happy by writing, 
"The Alkmeonids had taken up the contract some time before but 
now, after Leipsydrion, they began to exploit the gratitude." But the 
pluperfect is not a good story-telling tense. 

There may be a parallel. In 6.48 Herodotos tells how Aigina sub­
mitted to Dareios in 491. Kleomenes of Sparta intervenes, Kleomenes 
intrigues, there is hostage-taking and attempted hostage-returning, 
there is Kleomenes' death, war between Athens and Aigina and other 
things besides. Yet when we return to the story we are still only in 
the middle of 490. It is virtually impossible that so much should have 
happened in so short a time, and again it seems that the story may 
have got the better of chronology, that one of the limits must go. 
Modern scholarship abandons the lower, and the events are extended 
into the 480's. It is less easy to say here what the story demands and 
modern scholarship may well be right, but, with 510 in mind, it is 
worth considering the possibility that Dareios had begun his activ­
ities some time before 491, that once more a pluperfect has been 
avoided. 



286 THE TRADITION OF HIPPIAS' EXPULSION FROM ATHENS 

To return to the point. My conclusions are two. First that Herodotos 
dated the building of the temple before 510. Second that in doing so 
Herodotos was probably right.22 

W ADHAM COLLEGE, OXFORD 

October, 1969 

21 An earlier version of this essay was read to a seminar at Trinity College, Toronto in 
November, 1968. I am grateful to all those present for their company and their criticism, 
especially to Professors M. E. White and L. Woodbury.