372 34 16MB
English Pages 364 [201] Year 1968
STUDIES
IN GREEK GENEALOGY BY
MOLLY BROADBENT ·
LEIDEN
E. J. BRILL.
CONTENTS LIST OF TABLES
VI
,ACKNOWLEDGEMENTS
VII
I. INTRODUCTION
I
II. FouR SIMPLEGENEALOGIES · I. II. III. IV.
18
r8 23 27
Genealogy of Local Epidaurian Gentry. The Priests of Poseidon at Halik:arnassos Hellanikos on the Queens of Troy The Kings of Corinth . . . . . .
39
III. LAw-couRT SPEECHES:THE BousELIDAI
61
I. II. III. IV.
Materials of the Genealogy . . . . General Problems of the Materials . The Informants . . . . . . . . . Chronological Analysis of the Principal Body of Information . . . . . . . . . . . . . . . . . . . . . V. The Principal Problems of Relationships . . . . . . VI. Characteristics of the Genealogy and Genealogical Information . . . . . . . . . .
IV. CLASSICAL ATHENIANFAMILYLAW.
I. The Family Unit . . . . . . II. The Interrelations of Families . III. Laws on the Internal Structure of the Family and Kindred . . . . . . . . . . . . . . . . . . .
61 63 69
79 96 105
II3 n9 150
V. AN ATTICDESCENTGROUPAND THE LITERARYGENEALOGY OF ITSEPONYM. · . . Copyright 1968 by E. ]. Brill, Leiden, Netherlands All rights reserved. No part of this book may be reproduced or translated in any form, by print, photoprint, microfilm or any other means without written permission from the publisher
A. The Attic Genos. B. The Literary Tradition ANCIENTSOURCES SELECTIVEINDEX
197
ACKNOWLEDGEMENTS
LIST OF TABLES I. II. III. IV. V. VI. VII. VIII. IX. X. XI. XII. XIII. XIV. XV.
facing page The Epidaurian gentry . . The Halikarnassian priests. The dynasty of Troy . . . Ruhemann grid of Hellanikan Troy The kings of Corinth . . . . . . . Theopompos' family economy . . . The Bouselidai: fourth chronological approximation . The lexicographic tradition of the Law on Heiresses . . . . . . . . . . . . . . . . . . . . . facing page Stemma Kephalidarum: the kinsmen of Andokides . facing page The Pherekydean Kephalos . . . . . . . . . The Homeric genealogy of Odysseus . . . . . Synoptic table of the fragments of Hellanikos' Atlantis . . . . . . . . . . . . . . . . . . . . . facing page Kindred of Kephalos and Odysseus in early romance Kindred of Kephalos and Odysseus in late romance . . . . . . . . . . . . . . . . . . . facing page The 'non-Hellemc' genealogical doctrine . . facing page 1
•
•
•
24 25
29 35 43 70 94 20
8
248 264 322
328 332 334, 336
\~}MY debts of gratitude
are due to: the University of Glasgow and the __egie Trust for the Universities of. Scotland, who have provided q,n.ey for the work on this book: to the Library of Glasgow University, J{o..,have borne my plaguing them for years (and I• am very stupid :,.t}ilibrary catalogu~s) with more than courtesy; and to Miss Maureen a:e; whose hands have been busy with many things in order to set mine ~~-. These have been the indispensable; for the imponderable-and ,. ·.e I know the encouragement received I can only guess at the patience ~quired-Mr. A. R. Burn, Mr. T. Price, and Dr. R. Beerman have been hose upon whom propinquity has allowed the greatest demands; only and distance have, in their essentially insignificant ways, abridged fu,1:iatI owe and have owed to Professor Arnold Gomme, Dr. Felix . f]a~oby, and many others who have confirmed or castigated at various .tf,tages. Indispensable again: Miss Jean Fyfe who typed from my handrtwnting, and Dr. van Proosdij of Messrs Brill, who printed Greek from it. \{A,nd finally, to Mr. Davis Taylor, who by making publication possible, {{dedit his quoque·jinem.
,,,..e
CHAPTER ONE
INTRODUCTION of the more enduring characteristics of Greek literature is its '-~;.of genealogising. This is found already in Homer, together with the /_qus magic of lists of proper names which are not genealogical; in a,,rchaic epic the interest grows, so that complete poems were collecJojs(of genealogies; it is constant in the mythographers and historians ,,,\alJ. .•ages and levels of competence, in tragedy, some comedy, in romance Jf~ensationalist writing, in Hellenistic scholarship and commentary, /ge,ography, ethnology and social science. If we count its literary gwp.irig in Homer, we must recognise its end in his commentator J-1~,tathiussome two thousand years later: and we may reflect with some •cr~d11lity not only on the· thought and time spent on genealogy in \~iting the original works, but the immeasurable labour involved in copying of these works (or their fragments) throughout the manuscript ~µturies. JI'bis genealogising is a literary pursuit, a work of ordered imagination. · igther arts,_ criticism which concerns itself with the larger and smaller. {ails of workmanship, styles and schools, helps to explain something gfJ:1:i~irpleasure, and informs our appreciation: the same should be true g#C:the genealogies. Moreover at certain periods genealogy created by \t~~?rdered imagination becomes the principle of unity for historiography, ·:''.jfth.us combines and expresses the orders of both art and science: this Jiates, a new tension and equilibrium within genealogy itself, and where #~/tension becomes too great to be sustained, sensationalist writing ~~c_tures the larger creations and regales the disorderly with tales of :ti~stand other horrors. ;jl]ie relationship of Literature to Life i~ this department of genealogy ~$;been much debated. The outstanding fact on which the debate rests _s{tliatthe immeasurable mass of genealogy is concerned with the Heroic· Je'; historical genealogies of more than three generations' depth in ·are few, and none can be regarded as complete or indubitable in llidetails. The Athenian law-courts could be asked to decide a series of ~~s arising out of the single question of the identity of a wife's paternal .:~µd.rnother: and there is no suggestion that this ignorance was at all :;septional.
}-ti
:~e
',iiA~
2
INTRODUCTION
The passion for genealogising is then Literary: in Life the Greeks were by no means so concerned. So we must not compare the dispute about Phylomache's grandmother with the variation of statement about Agamemnon's father: in the former case, drawn from Life, we can properly entertain the supposition of ignorance or doubt reinforced by conflicting interests; in the latter we can suppose nothing of the kind. Homer 'knows' that Agamemnon's father was Atreus; Hesiod (F 98) asserts that he was Pleisthenes; the commentators add that Pleisthenes was Atreus' son, and predeceased hi; father, so that Agamemnon was known as son to Atreus. Here plainly there is no question of 'ignorance' : there was some reason which compelled Hesiod to change the accepted genealogy, and change it in a certain way. The instances of variation in the Literature are innumerable: the task is to explain the reasons for them, and the natural way to attempt a beginning of explanation is to try to group and classify these reasons. A genealogical variation consists of two parts: an objection to the received tradition, and the innovation itself: and the compelling reason may lie in either part. For example, Hesiod may have said to himself that it was impossible (for a compelling reason) that Agamemnon's father was Atreus, but that we could meet this compelling reason and reconcile it with the Homeric statement by making Agamemnon's father son to Atreus-let us therefore call him Pleisthenes (More Strength}, to intimate how much greater Agamemnon was than Atreus. Or Hesiod may have argued that (for a compelling reason) Agamemnon's father was Pleisthenes, but that we could meet Homer's assertion and reconcile it with the compelling reason by making Pleisthenes son to Atreus. Each genealogical variation has these two parts; but many have more than two alternative persons-we know, for example, of some five mothers for Penelope, four fathers for Kephalos, and so forth. In a small minority of examples we can say that the variation was propagandist, to the greater glory of some interest concerned in the possession of a heroic past. Most variations are not however susceptible of such an explanation, but are either arguably products of learned doctrine, or may be suspected of being so: others have no discernible reason at all, and. are often written off to a passion for 'originality'. There is however a difficulty in this thesis, namely that the supposed originality works very erratically-that while some heroic persons have six mothers, others have none (surviving in our sources). In other words, some genealogies were of more interest than others to the ancient genealogists, and the question is why?
INTRODUCTION
'ti{::has
3
been the custom to look outside Literature and Learning, to ;f~;Lfor answers to these questions. Where propaganda is observably #c~rned, this is the correct methodology: some variations, preserved ,,9nymously, can be dated and well understood in this manner, and in · }:;-wider field of general mythography, Nilsson has collected and disf~yed a number of examples in Cults, Myths, Oracles and Politics in ·tn'aient Greece (Lund I95I}, which considers, in the field of genealogy, ]j.finstructive examples of the Macedonians and Epeirotes. Following 11.i~;~:x:ample it is natural to ask what department of Life can have given ··;{f.o the mass of heroic genealogising and its innumerable variants. _'f;lie'answer has been sought in the department of kinship structure, known to modern anthropology in both description and theory. The afu6us work is Bachofen's Mutterrecht, which treated Greek heroic ,,~~dfogies as accurate or slightly derivative descriptions of real states i:[ffairs correctly (or almost so) remembered in literature. Bachofen's ~ir1contention was that matriliny (descent reckoned through women) µcl'aslii depa:ttmentroflhist6ry in', the work of Hekataios of Miletosf a!,'6und,1-i:hettl:tn{:@fd:he:sixth,,anq.rfifth: centuries; and its early period triayrbe::convenientlyJ:stu'died:lin", the::ethnogtaphic between portions of Herodotus., !W1:f1fin'd.1 1aJsystemati6:ilifferentiation· what Herodotus saw £di himselfj!andi,w:hati was~toldc[him; ·and within the latter kind of repotit·edi$tinctioris·,are. :nradeubetween !informants'. of different kinds. Herouotg.s is'athis•.woi-st;it1;1his-,a~count ofiEgypt,:wheie he seems to have rhisapprehendedrthe:ha'ture· of the• ilariguage ,difficulty . and been fundame,ntally:,misledhby-inform:ation;;,rec~ived from Greeks domiciled in Egypt ;J on;,peoples'. by~whose,idyilisationsi he, was :less ,impressed he is a muchr.bette1nf~porter;;).ii1
. I ············1 Hieron Aristeas I I (z4) Althephos: 14y (I5) Poseidonios:
I
(z7) Hipparchos: 4y
19y
(zo) Alkyoneus: 17y
Dioskourides
I
12y
I
19y
I 22y is: 3oy
· (7) Anthas:
I
!
Phyleus 23y
I
I Aithaleus
i
7. (z6) Androsthenes:
I
(4) Alkyoneus:
9Y
(9) Hipparchos:
25y
6. (IJ) Andron: 25y
I
12y
I I
I···················~-----. I Athenippos: 5oy Androstheries
0
en
I
( x z) Poleites:, 5y
I
(x 3) Euaion:
I
28y
( X 4) Polites: 27y (adopted son of Apollonides)
t,;l
c.n
26
FOUR
SIMPLE
GENEALOGIES
first cousins of Andron instead of first cousins of his son or second cousins of either. . (I7) Hipparchos II, (I8) Demetrios, and his son (Ig) Philistos come between (I6) Androsthenes an9- his (inferred) grandson (20) Euandros: Euandros and Philistos therefore cannot, ex hypothesi, be more than 7 degrees apart. Since in the previous indeterminable case the nearest possible relationship was given in the table, here (also exempli gratia) the furthest possible is given, so that Euandros is second cousin's son to Philistos; instead we cotild have made Dioskourides brother to Androsthenes, and Philistos and Euandros second cousins. . (2I) Demophilos and. (22) Eukrates succeed (20) Euandros, but the name and patronymic of the twenty-third priest are lost, so that we. have no lower limit. The table makes them first cousins of Euandros , as a guess. We have now to consider the authorship of this genealogy, and its relation to the genealogical source-material or family tradition. The obvious candidate for authorship is the fourth century genealogist Andron of Halikarnassos, whose name suggests that he belonged to this family. His fragments (Jae rn) show that he often agreed with Hellanikos, and perhaps as often innovated; he also subscribed to Skamon's contention that writing was invented by, or ascribed to, Kekrops' sister-inlaw. This claim to the possibility, or possessiorr, of documentary evidence for the greater part of Heroic and all post-Heroic hi.story throws a somewhat sinister light upon our present inscription: if it was intended to be parf of this documentary evidence, it is a forgery. If on the other hand it is intended as a reconstruction based on traditional materials it is that and nothing worse. ' Unfortunately, the names are not very convincing. We shotild expect a genuine genealogy to resemble, say, that of the ancestors of the Asklepiad Hippokrates, in that it would begin with 'mythic' or at least oldfashioned names, contain epichoric names (like Krisamis) in its middle portion, and come to names of the normal historical type rather later. But this genealogy continues to use 'mythic' names down to the time of Alkyoneus II, and two generations afterwards already possesses names of the historical type: Dioskourides, Poseidonios, Demetrios; and there are no epichoric names. It would therefore seem doubtfril whether the author of this list had much material other than his own inv~ntions for these early generations.
It will of course follow that, if he intended limitation of the right
THE
PRIESTS
OF POSEIDON
AT HAL~KARNASSOS
27
. of inheritance to agnates of the seventh degree; the author will have been using either the inheritance law of his own time, or the oldest remembered in Halikarnassos, and attributing it to this early period, ..unless there were already established, in Universal History,a doctrine of early Dorian inheritance law. We shall discuss this question in treating the Corinthian genealogy. · III.
HELLANIKOS
ON THE
QUEENS
OF. TROY
We have now to consider not a sort of forgery, but a sort of fiction, a mythic genealogy. And the same kind of question arises here as in the Halikarnassian case: did Andron invent an inheritance law; did Hellanikos invent a marriage system? I have chosen the Trojan genealogy to represent the type of mythic genealogy because it is the simplest for several reasons. No Greeks claimed descent from it, and therefore we need not suspect the operation of family interest in the variants. The agnatic pedigree is established by Homer: it is a patent fiction partly composed of eponyms, b.ut for our purposes we can accept it as a datum. Otl-re:ae:&noµ.1J-rpo£; 't'o -yevo£; ''Oµ.'Y]po£; O':X.'YJ[LOC-rll:e:L as his scholiast very properly remarks (schol M Eust a 343= Jae 4 F I5o), but Hellanikos and other later authorities do, and it is reasonably possible to isolate Hellanikos' contribution. Our concern here will be to discover and consider the principles of that contribution. The isolation and identification of Hellanikos' contribution has already been largely carried out by Jacoby in his commentary on the fragments of the Troika, upon which Table III is based. This shows the certainly Hellanikan names in capitals: that is, all the Homeric names, and those explicitly q:-edited to Hellanik(?s in our sources. Those which are not Hellanikan are in ordinary print; and those which can be attributed to Hellanikos with a substantial degr~e of probability are in italics. The number of the fragment (J ac 4 F) or the narrie of the alternative source is added for the non-Homeric names. The reasons for accepting or rejecting certain names or relationships as Hellanikan are: I. Kalirrhoe daughter of Skamandros. as wife of Erichthonios (Dion. Hal.) cannot be Hellanikan, since (F I38) she is the wife of Tros: this leaves Astyoche daughter of Simoeis as the Hellanikan wife of Erichthoiiios. 2. Akallaris daughter of Eumedes as wife of Tros (Dion. Hal.) is not Hellanikan, for (F I38) Tros already has a wife.
28
FOUR SIMPLE
GENEALOGIES
3. Bateia daughter of Teukros as wife .of Ilos is not Hellanikan, for (F 24) she .is wife to Dardanos: this leaves Eurydike daughte~ of Adrastos as the Hellanikan wife of Ilos. 4. Klytodora daughter of Laodamas (Dion. Hal.) appears in a .version which Kalirrhoe and Akallaris have already shown to be non-Hellanikan; if she is in consequence excluded, Hieromneme daughter of Simoeis is the Hellanikan wife of Assarakos: she (or rather her homonym the naiad Hieromneme} cannot then be wife to Kapys (Dion Hal.); his Hellanikan wife will be the daughter of Ilos, named Themiste or Erytheia. 5. Strymo, .the Hellanikan wife of Laomedon, will be daughter of Skamandros, for she has no other father .in our sources. 6. Arisbe, a wife of Priam, may have been daughter either toTeukros or to Merops in Hellanikos.
7•. Since the daughters of the rivers Skamandros and Simoeis are so important for Hella.:nikos, he may also have made Oinone daughter of the Kebren. But the statement that Hekabe. was daughter· to the Sangarios is probably the work of an imitator: Hellanikos probably followed Homer. 8. Hellanikos presumably made Bateia (F 24) the dg.ughter of Teukros, as did everyone else; it is not certain that he also made Teukros son to Skamandros, while others made him. a Cretan. But since Hellanikos made Dardanos an immigrant, it is likely that he made Teukros an autochthon, son of the local river and mountain. ··· ···From this point onwards we shall assume for convenience' sake that the textual and source criticism are sound, and that the names in capitals and italics in Table III are in fa~t the Hellanikan genealogy of Troy. The first characteristic of this mythic genealogy is its isolation from the other Greek heroic pedigrees, which are, typically, inextricably intertwined. The Trojan is connected with the others in the generation of Dardanos, and through the sisters of the Pleiad Elektra, and much later through Telamon and Hesione, but it remains true in general that the genealogy as a whole is isolated in a quite exceptional way. Hellanikos in one case even increases this isolation: Hesiod (Theog. 954) had made Laomedon's son Tithonos marry the goddess Eos, who had also married other heroes (e.g. Kephalos) and thus constituted a connection with the rest of the mythic genealogical corpus; Hellanikos (Jac 4 F I40) changes Eos to Hemera, and thus precludes this connection.
HELLANIKOS
ON THE QUEENS
Table III
OF TROY
29
FOUR SIMPLE
GENEALOGIES
Hellanikos achieves this effect, in, the main, by making the queens of Troy come of divine stock. Homer had already shown the way by naming Aphrodite as ,the mother of Aineias; Hellanikos adds the daughters of the Simoeis and daughters and grand-daughter of the Skamandros; and he adds also Iasion the consort of Demeter, and Harmonia the immortal ancestress of the Kadmeians. All this is quite obviously deliberate, and works out a thesis: the Trojan heroes were ijµl0e:OL,of course; and their divine element is continually reinforced through their mothers, from Elektra to Aphrodite. It is a part of this thesis, and no less deliberate, that Hellanikos makes Assarakos and Laomedon marry their grandmothers' sisters; equally deliberately, the source of Dionysios of Halikarnassos changes the marriages so that this phenomenon does not occur. Plainly, we have come upon· a controversy of genealogical principle, and the question is whether we can discover anything of what was at issue. The obvious part of Hellanikos' statement of principle is that Trojan kings marry daughters of the Simoeis and Skamandros in alternate generations, or, in other words, that there is a two-generation cycle in the direct male line. Such cycles of nomenclature are familiar from historical genealogies, as in the Epidaurian lineages; but we have so little information ·about historical marriages that to say we have no evidence for two-generation cycles of marriage means very little. It is, nevertheless, true: the question has not arisen, and Hellanikos' statement . finds us without critical resources. It is therefore useful to begin by examining Hellanikos' notation, his means of making statements. His alternate generations who marry daughters of Skamandros and Simoeis extend from the second to the fifth kings; the first king marries a grand-daughter of Skamandros, the sixth has a grandson called Skamandrios: the notation plainly extends throughout the genealogy, with whatever other elements it may be combined. We need then to create a notation more abstract than the Hellanikan, in order to discover and consider his statements undistracted by his way of making them. For this purpose I shall employ the principles of a Ruhemann grid, 1 which is a device for · displaying the abstract geometry of kinship systems. It assumes a closed community (for purposes .· ·· of marriage): and Hellanikos' isolated genealogy suits well. Given this assumption, it is possible to construct a grid which will provide room for B. Ruhemann: 'A method for analyzing classificatory relationship systems' Sonth-western Journal of Anthropology (Albuquerque) I (1945) pp. 531 ff. 1
HELLANIKOS
ON THE QUEENS
3I
OF TROY
every actual or .recognised category of relative by blood. or marriage: · and the grid is designed to lead from reports of fact to questions of cate:gory. We have, from Erichthonios to Laomedon, four Trojan kings in two two-generation cycles in the direct male line. They constitute a single lineage, and we can write the names of the sons under those of the fathers, designating ·1:hegenerations of the cycle by the letters A and B. If there is more than one sori in a generation, the brothers or cousins are entered together. We immediately see that, if Ilos I belongs to the Hellanikan genealogy, this notation is as suitable to determining the regularities of, a two-generation cycle of nomenclature as to determining those of marriage. A Erichthonios Idaios, brother Ilos I, brother
B Tros
A Ilos II Assarakos, brother
B Laomedon Kapys, brother's son
Now let us write the daughters of Skamandros to the right, and those of Simoeis to the left, of this portion of the Trojan lineage, first as an ordinary genealogical table: ' Simoeis
I
Astyoche
Erichthonios (A) (Ilos I, brother)
Simoeis
Tros (B)
Kalirrhoe
Ilos II (A) Assarakos (A)
Skamandros
I
Hieromneme
I I
I
Laomedon (B)
Skamandros
I
·
I
Strymo
Written in this way, it is obvious that, instead of saying that Assarakos and Laomedon married their grandmothers' sisters, we could as well say· that Simoeis acts both as grandfather and his own grandson: that his
32
FOUR. SIMPLE
HELLANIKOS
GENEALOGIES
name recurs just as does that of Ilos; and the same is true of Skamandros. For Hellanikos, they were conveniently immortal, and therefore he could use the names in this way without multiplying the persons. Our problem is to produce an equally simple but more abstract nota. tion. If we apply those we have already used for the royal lineage, it is plain that just as brothers of the kings are entered with the kings, so sisters of the queens would be entered with the queens; and consequently. that the husbands of these sisters would be entered with the kings. We should then have a number of categories of relatives, and the question would be whether these categories were relevant, that is (in thls case) represented Hellanikos' mind, or whether a less simple geometry, separating out the lineages, would be required: in other words, our problem would be 'how few symbols, how disposed, do we need?' Our problem at the moment is the same. We can analyse the royal lineage with two symbols, A and B, disposed vertically. With how few symbols can we analyse the other relationships? Hellanikos alone gives no direct evidence, but in other sources (r) a brother of Erichthonios is Idaios; (z) a sister is Idaia (who married Phineus); and (3) probably from Hellanikos, Skamandros married an Idaia: I Erichthonios (A) Idaios (A)
I Idaia an Idaia
I
Phineus Skamandros a Skamandros I
Astyoche
=
I Erichthonios
I Tros (B)
(A)
I . Idaia (a)
I
Skamandros
I
Kalirrhoe
But the whole marriage cycle of the royal lineage can be expressed in two symbols, so the first trial should be to see whether these two symbols are sufficient for the other relatives:
I
Erichthonios
(A)
I
I
=
Idaia (a) I Kalirrhoe
Skamandros
(B)
l
(a)
They are sufficient, at the cost of placing Skamandros in the same category as Astyoche, as if they were brother and sister; of placing 'fros in the same category as his mother Astyoche, as though he inherited that category from her; and of placing Kalirrhoe in the same category as Idaia, as though she were her mother. We can only discover whether these categorisations are justified (that is, represent Hellanikos' mind) by a wider trial; and in fact we are now ready to construct our grid. If we need only the two symbols (with their variants for sex) we can write the first line of our grid as above: b = A a = B. In the second line we.already have one son written beneath his father, and a woman written beneath her mother-by-category, so we need only to fill the vacant places accordingly: b B = a A: and.b will be the wife-by-category of A, asif the paper on which the diagram is written were tubular and not flat ..The third generation, following the same principles, will be identical in symbolism with the first: (b)
I Erichthonios
{b)
Tros (B)
(b)
As we have seen, for our geometry, and apparently also for Hellanikos' notation, the nomenclature cycle and the marriage cycle are identical: Idaia must then share her brothers' symbol, and to record the difference of sex we will call her (a). She cannot share her symbol with her brother's wife, who is a daughter of Simoeis, not a relative of Skamandros:
(b)
Tros {B)
Astyoche
I Kalirrhoe
Tros (B)
Astyoche
33
ON THE QUEE:NS OF TROY
I
I Ilos. (A)
(A)
Phineus (B) Skamandros (B) I
I Idaia (a) . I Kalirrhoe Kleopatra
I
(a)
(a)
(A) (B)
We are not told whom Kleopatra married; but a Kleopatra was the first .· wife of the Phineus whom Idai;1 married, and our symbolism would be adequate to include this also. We are now ready to give the genealogy and the grid their test of one another, and in case the reader would like to check the experiment step by step, I set out the procedure in full: On a large sheet of paper, in the m:;i.rgin,set out the figures r-ro wellspaced vertically. These number the generations, and must leave room for multiple entries. At each odd-numbered generation write the line of symbols (conveniently spaced horizontally) = b .B - a A=. At each even- numbered generation write the line of symbols b = A a = B, so that four vertical columns of symbols are formed. These columns are a direct female descent line (by fact or category), a direct male descent line (the same), a direct female and a direct male descent line (by fact or category) respec-
34
FOUR SIMPLE
HELLANIKOS
GENEALOGIES
tively. The grid is now ready for use, and the test consists in writing in the genealogy in the appropriate form, and examining the fit, thus: in generation 2, place (b) write the name of Elektra-Elektryone, and in place (A) the name of her consort Zeus. under Elektra's name, in generation 3, place (b) write the name of her daughter Harmonia, and under Zeus' name, in generation 3, place (B) write the name of his sons, Dardanos and Eetion-Iasion. In the same generation, in place (A) write the name of Kadmos, Harmonia's husband. in generation 4, place (b), write the name of Astyoche, as though she were the daughter of Harmonia; and in generation 3, place (A), write the name of Astyoche's father Simoeis, as though he were brother to Kadmos. in generation 4, place (A) write the names of the brothers Erichthonios, Ilos I and Idaios; in place (a) write the name of Idaia, daughter of Dardanos, and in place (B) write the name of her husband Phineus. in generation 5, place (B) write the name of Tros; in place (a) write that of his wife Kalirrhoe, and in generation 4, place (B) that of her father Skamandros. · · · in generation 6, place (A) write the names of Ilos II and his brother Assarakos, and in place (a) that of their sister Kleopatra. In place (b) write the names of Bateia and Eurydike, wives of Ilos, and of Hieromneme wife of Assarakos, and enter their fathers' names in generation 5, place (A) : Teukros, Adrastos and Simoeis. in generation 7, place (B) write Laomedonand his cousin Kapys, and in place (b) Laomedon's sister Themiste-Erytheia. In place (a) write Strymo's name, and in generation 6, place (B), the name of her father, Skamandros. in generation 8, place (A) write Priam and Anchises, and in place (b) their wives: Arisbe, Hekabe and Aphrodite. In generation 7, place (A), write the wives' fathers: Teukros or Merops, Sangarios or Dymas, and Zeus. in generation 8, place (a) write Hesione's name, and in place (B) that of . Telamon: in generation 9, place (A) write the name of their son Teukros. in generation 9, place (B), write Alexandros, Hektor and Aineias, in place (a) their wives, Oinone and Andromache; and in generation 8, place (B) the names of Kebren father of Oinone and Eetion father of Andromache: in generation ro, place (A) write the name of Hektor's son Skamandrios. Finally, since Teukros appears in 5A, 7A, and 9A write Teukros, the father-in-law of Dardanos in generation 2, place (A); Skamandros his
ON THE QUEENS
35
OF TROY
father in generation r, place (B); Idaia his mother in .generation r, place (a), and Bateia his daughter in generation 3, place (b). The diagram should now look like Table IV.
Table IV b
B Skamandros A Zeus Teukros
a
·Elektryone b Harmonia Bateia
B Dardanos Eetion-Iasion
a
b Astyoche
A Erichthonios Idaios, Ilos
b
B
b
=
a
.
Idaia
Kleopatra
b ThemisteErytheia
B Laomedon Kapys
Strymo
.b Arisbe Hekabe Aphrodite
A Priamos Anchises
Hesione
B Alexandros Hektor Aineias
Oinone Andromache
b
A Skamandrios
2.
A
3.
B Phineus Skamandros
A
A Ilos Assarakos·
b
B
a
Kalirrhoe
b
r.
Kadmos Simoeis
Tros
Bateia Eurydike Hieromneme
A Idaia
a
.
a
B Skamandros (Phineus?)
6.
A Teukros or Merops Sangarios or Dymas Zeus
7·
B
8.
A
9.
a
Telamon Kebren Eetion II
a
a
5.
Teukros Adrastos Simoeis
Teukros
B
Two difficulties appear in the course of constructing this table: Dardanos in 3B marries Bateiain 3b (wherea!>his wife 'ought'. to be in 3a); and Kapys in 7B marries Themiste-Erytheia in 7b. But from the point of view of Hellanikos' contemporaries (so far as we know)these are the most
FOUR SIMPLE
GENEALOGIES
HELLANIKOS
normal marriages in the genealogy: Dardanos as the heir of Teukros is naturally his adoptive son as well as husband of his daughter; and as for Kapys, marriage to an agnate's daughter was preferred (marriage to a brother's daughter was common). These.two instances should not, therefore, inte;fere with our consideration of the remainder of the table, even though later we shall need to ask.how Hellanikos related them to the rest of his material. Our geometry was designed to accommodate categories of marriage and nomenclature, and by entering the names as connected by marriage and blood, the nomenclature coincides with the marriage-category in these cases: Skamandros:. rB (through Dardanos' affiliation to Teukros); 4B, 6B; the patronymic Skamandrios 'son of Skamandros' falls correctly in roA, and Skamandrios, being Homeric, would be Hellanikos' startingpoint for the other placings. Idaia: ra (through Dardanos' affiliation to Teukros) 4a: probably not from Hellanikos: the geometry is not confined to his personal work. Zeus: 2A and 7A: the latter was a datum for Hellanikos Teukros: 2A (through the affiliation of Dardanos) 5A (not Hellanikan); 7A (uncertainly Hellanikan); 9A (presumably the datum from which Hellanikos started these calculations). Bateia: 3b (through the affiliation of Dardanos to Teukros) 6b (not Hellanikan) Eetion: 3B and SB, the latter Homeric. Simoeis: 3A and 5A; Idaios: 4A and Idaia: ra and 4a Ilos: 4A and 6A Phineus: 4B and ? 6B. In other words, wherever a name is repeated, all the repetitions belong to the same marriage category: our notation repeats symbols, Hellanikos repeated personal names. It would therefore seem that our notation in Table IV sufficiently represents Hellanikos' intentions for us to consider more narrowly what these were. _ We have already seen that he clearly wished to keep the Trojan genealogy separate from the Greek, and to derive the element of divinity from the mothers of the ~µ.W~oL,following Homer's account of Aphrodite the mother of Aineias. In view of the long history of speculations about I
ON THE QUEENS
37
OF TROY
'matriarchy' and allied matters in the Greek mythic genealogies, we should explore· Hellanikos' mind with some care. In the first place, we are exploring Hellanikos' mind, ·and not the history .of the citadel of Hissarlik, about which Hellanikos, we -may assume, knew nothing except its topographical relations to the Skamandros and Simoeis, and mount Ida. The roughest formulation of the question must therefore be, not did the Trojans practise 'matriarchy' vel sim., but did Hellanikos, for whatever reason, believe they did. Two cases in the genealogy enable us to give a decided answer: Priam inherits an element of divinity through his mother, but his kingdom from his father; Erichthonios inherits his kingdom/ram his father, and his membership of the A-category through his father. Inheritance is then bilateral, with by far the greater emphasis on the father's side. The disposition of our symbols in Table IV showing a two-generation cycle in the male lines and no cycles in the female lines is not therefore to be interpreted as meaning that membership of the symbolised categories is inherited from the mother: the two-generation male cycle is a property of the male descent line as such, not derived from its mothers: the marriages of Dardanos to his adoptive sister, and of Kapys to his ortho.,. cousin show the cycle undisturbed. Such marriages, and a two-generation cycle in nomenclature, are so normal to, say, fourth-century Athen~an families, tµ.at it is very doubtful whether Hellanikos had anythmg 'prehistoric' or 'Asiatic' in mind at all, beyond the element of divinity necessary to ~µ.We:oL. The river-gods Skamandros and Simoeis were, of course, brothers; and if the names were less 'mythical' in type, there would be ,nothing surprising in finding this on a page of Kirchner, or of Hesperia: Skamandros
I Simoeis
Teukros
X
I
~,----!
X
I
I
Skamandros
Dardanos ~,----~j_·
married"Idaia
-,-I Erichthonios
I
I
Simoeis
I = Astyoche
and so forth: we might be mildly curious as to why the family had not recorded earlier marriages within the agnatic kindred, but otherwise the stemma would appear quite normal.
FOUR SIMPLE
HELLANIKOS
GENEALOGIES
The difference between this stemma and the Trojan genealogy given by Hellanikos is that Hellanikos' river-gods are immortal, so that he does not need to multiply persons: he can state principles and regularities. We have translated bis notation into one still more abstract, in which the principles can be· even more generally stated; and now the question is, what were the principles which Hellanikos had isolated? The pure geometry is this: b my mother
b b. daughter in law
A my father
mother in law
B father in law
A
a
B
a
EGO
my wife
A my son
my daughter
a
B son in law
The model is of a closed community with two patrilineages which intermarry: their marriage law can be expressed as exchange of sisters, or cross-cousin marriage. Neither of these descriptions suits the persons (as distinct from the geometry) of the Hellanikan genealogy: the kings do not exchange sisters with the river-gods, nor do they marry their fathers' sisters' daughters: on the contrary, Kapys marries bis father's brother's daughter, against the geometry. There is however a third way of stating the marriage law: rriy wife's daughter marries back into my wife's patrilineage. This is a description permitting a far more flexible practice, for as I have received one woman, honour and justice are satisfied if I repay one: and niy other daughters are available as, so to speak, investments on behalf of my grandsons. Hellanikos, by this law, is not required to specify the relations between the various fathers-in-law of the kings: if some are brothers but head separate lineages, and if some are not related to the others at all (like Adrastos) this does not matter in the way it would matter if the marriage law Vl'.ereexchange of sisters or cross-cousin marriage. Moreover, the fission of the royal lineage itself into two minor lineages may be recorded by intermarriage: Kapys on this reckoning owed a daughter to Priam or one of bis brothers. It remains a question why the source of Dionysios .of Halikarnassos so completely. disagreed with Hellanikos' view. He was, apparently,, more sophisticated in a certain sense than Hellanikos, for he removes from the ancestry of Rome the more inconsiderable of the two river-gods, and he makes Hieromneme, as well as her daughter-in-law, a goddess in her own right. The parallel cousin marriage disappears, and the Simoeis is replaced by two unrelated unknowns, Eumedes and Laodamas, whose names are undistinguished and from the ordinary heroic stock. This source
ON THE QUEENS
OF TROY
39
then had (so far as we can see) nothing very special to put in place of Hellanikos' geometry: apparently he mainly disliked the geometry, or held Hellanikos' implicit arguments to be ill-founded. So far as we can see, Hellanikos' contention is that the nomenclature and marriage-cycles are identical, and lie had as data the names of Skamandrios, Teukros the son of Hesione and Telamon, Andromache the daughter of Eetion, and Aphrodite, apart from the .agnatic genealogy. It is also extremely probable that he inherited from the Hesiodic or even from recent ethnographic or geographical tradition some 'reliable' assertions about Dardanos and Teukros: the statement that Dardanos succeeded Teukros would be sufficient to give him the pure mathematics of bis calculation. Th.e rest of the genealogy is provided by bis thesis of maternally~inherited divinity: ·But this argument has a weakness of evidence: the names of the kings as given by Homer do not in fact show any cycle of nomenclature, and this was surely enough reason for a critic·to reject the proposed marriage-cycle, while Kapys' marriage within the agnatic kindred would be quite repugnant to bis Roman descendants, among whom spouses do not share the same nomen. We can then to some extent understand the nature of the genealogical controversy; but this understanding is not enough to tell us the source of Hellanikos' ideas. The most we can say is that it is almost completely certain that Hellanikos did not invent the two-generation marriage-cycle: he must have known it either in contemporary practice or in the detailed · tradition of some Greek community.
IV.
THE KINGS
OF CORINTH
(a) the la'lf!pj inheritance
Our fourth genealogy is in itself very simple, but is associated with some narratives which require a complex of evidence for their interpretation. The historical value of the genealogy is probably, within certain limits, good: we should probably regard it as a document of bistoriographic sociology-not, that is to say, as the result of social science of the fourth . century type, but the result of the need in Corinth after. the fall of _the tyranny for unlearned but responsible Corinthians to come to terms with their own history. In other words, this genealogy probably represents, for us, the only type of unlearned and authentically local historiography which would be admitted to ancient Universal History: it embodies a thesis, partial but well-conceived and consistently carried through, on the
FOUR SIMPLE
GENEALOGIES
organisation of pre-republican rulership in Corinth which explains some of the peculiar features of pre-republican history. The document on which we have to work is quoted from Diodorus, by Eusebius in his Chronographia, whence it comes to us in the Armenian version, and, fortunately, the Greek text was quoted by Synkellos, and so is preserved. · The document, unusually, gives not only the genealogy but also (summarily) the law of inheritance: Synkellos p. 337.7 f. reports µr::'t'ix b 1tpscr~6-roc-roc; &d 't'WV oe 't'1JV't'OU't'OU (Aletes, the first King) 't'f::ASUTYJV exy6vwv &~occr£Ar::Ucrr:: µexpt -r'yjc;KmjieAOU-rupocw(ooc;.'after his death, the eldest of his posterity always reigned, up to the tyranny of Kypselos'. The Armenian translates this sentence, according to Petermann, as 'Post cuius obitum semper filius natu major regnum a parentibus accipiebat', and according to Karst as, 'Nach dessen .Ende stets nur der erstgeborene Sohn in einem fort van den Eltern das Konigtum erlangte' : 'after his death the eldest (elder Petermann) son always received the kingdom from his parents', on which Schoene observes that a parentibus is apparently intended to translate an &Xyovewvread for the exy6vwv of Synkellos. In other words, the Armenian was convinced that b 1tpr::cr~6't'oc't'oc; must mean the eldest son, and then had to obtain some meaning for 't'WV exy6vwv,which he did by reading it as &Xyovewv.But b 1tpr::cr~u-roc-roc; -r&v exy6vwv should mean inheritance like that of the Halikarnassian priests, the eldest survivor within the generation always succeeding. As we have seen from the Halikarnassian genealogy, this system of inheritance could work well enough while office was held for life; the Corinthian document however goes further, and says that this rule obtained down to the tyranny, including therefore the annual magistrates at o' CX7t0 .who succeeded the kings: A1hoµev'Yjc;µev ~p~r::v&Vtocu-r6v, 'HpocxMouc;Bocxx£ooct7tAr::£ouc; ()V't'SI,; otocxocr£wv xri-recrxov't'~V &px~v XOCL xowfi µe:v 7tf)Of::LO''t'~Xf::LO'OCV -r'ij,;;7t0Af::WI,; &.1tOCV't'f::I,;, &~ oc1hwv oe: lvoc xoc-r' evtocu-rov1ipouv-ro1tp6't'ocvw, l5,;;-r~v -rou ~occrtMw,;; r::!xr::-roc~tv.'Automenes reigned one year, then the Bacchiadai of Herakleid descent, more than zoo in number, took the rulership and all together managed the public affairs of the city; they chose one of themselves each year as prytanis, who fulfilled the office of the king'. Apparently therefore the rule of seniority within the generation was combined with some sort of selection. as in practice such a rule must always be, to avoid the appointment of the totally incompetent or impure. The difference is that when successions are frequent the exceptions to, or limits on, the general rule are more obvious, and therefore we may suppose survived in the tradition,
THE KINGS
OF CORINTH
The comparison with the H.alikarnassian genealogy raises a question which the Corinthian document does not answer explicitly: ,whether there was a limitation of inheritance to the seventh or some other degree. But the genealogy itself answers this question in the affirmative: Bocxxtc; (the fifth king) ·. . . yr::v6µr::vo,;; emcpocvfo't'OC't'OI,; 't'WV7tfl0 OCU't'OU. Oto XOCL cruve~'Yj't'OOI,;µr::-ra::'t'OCU't'OC ~OCO'LA.r::OmXV't'OCI,; . oux ~'t'L 'HpocxA.dooc,;;, &A.A< Bocxxlooc,;; 1tpocrocyopr::6r::cr0oct. 'Bacchis achieved greater fame than his predecessors. As a result those who reigned after him no longer had the name of Herakleids, but of Bacchiads'. In Halikarnassian terms, once Bacchis was king from among the great-grandsons of Aletes, the other great-grandsons (third cousins to Bacchis) fell out of the group with rights of succession, which was now confined to those .within seven degrees of agnatic kinship to Bacchis himself. ·Now if the rule of succession is that (within 7 degrees of kin) the eldest survivor succeeds, it must often have happened that the first representative of a new generation was not the eldest member of that generation (who was already dead), but a brother or cousin: in the generation of the great-grandsons therefore the succession would often pass to a cadet. This may be illustrated by the following diagram, in which the persons are numbered in the order in which they were born: X, the founder
I
I
L (2) .
(1)
I
I
I (5)
(3)
I
(8)
t
I
I
(12)
I
(9)
I
I
I (4) I
(7)
t
I
I
(13)
I (6) I
(10)
I
I {II)
Where (7)'s surviving son was older than (8)'s, all the descendants of (I) drop out of the succession as third cousins (of the eighth degree); in the next generation all the descendants of either (4) or (6), and so forth. · ' Thus when Bacchis succeeded, all the descendants of Ixion's brothers would drop out of the succession, whether Ixion was the eldest son of Aletes or not: and this Corinthian king-list is properly not a list of kings · (of whom there would be several to each generation), but a list of generations, a legal document giving the title to lawful kingship as it passed to each surviving senior member of the next generation. We are now in a position to interpret the traditions of usurpation and assassination in the last generations of the monarchy: 'Aptcr-roµ~l>'Y),;; ...
42
FOUR SIMPLE
THE KINGS
GENEALOGIES
ae -re:)..s;u-r~crou;; &1tfi.ms;vulov Te:AeO"TIJV m:x:Wcx: 't'~V~AL,dcx:v, 00 XCX:'t' 0e:'fov. 'cousin' means the issue of siblings; the siblings' sons are adelphidoi: as cousin is the reciprocal of cousin, so is adelphidous to uncle.
(c) scholiast to Homer, Iliad 9. 460 in Cramer AP III r73: &ve:q;wtol &:ae:).,_qi&v ulo£, Wt;'AxLAAe:ui; xoctAlocc;,&:ae:AqiLao r:ae ol "t"filV &:ae:).,_qi&v 1toc'r:ae:c; we;o 'Opfo't"'Y)t;1tpoc;Me:veAOCOV. . cousins are siblings' sons, like Achilles and Asia~; adelphidoi are the siblings' children, as Orestes is to Menelaos.
(d) Ammonius, P· 54: E:~OCVEqJLOL xoct &ve:q;rntaLocqiepe:L ... e:ta1 ae ol µev &:ve:q;rnt [ "t"6lV]&:ae:).,_qi&v 1toc'r:ae:i; ..• E;~OCVEqJLOL ae ol ["t"6>V] &:ve:q;L&v micioe:i;. . exanepsioi and anepsioi are different .... cousins (anepsioi) are the children of siblings . . . ·. exanepsioi are the children of cousins. .
By comparing these passages we can see that Aristophanes, on introducing the term anepsiadous (fem. anepsiade) compared the word to the similar formation in adelphidous (fem. ade!,,phide),both being tokonymics, from adelphos (sibling) and anepsios (cousin). Then he described the relationship, as that of Ego's cousin's son (and no doubt gave the reciprocal, and a heroic example like those in the third testimonium): he would then add examples of the use in the literature (no doubt including Solon, as Pollux does for the 'cousirihood', see below). The section on the · ~nepsiadous would come between those on the a'-nepsios (cousin) and on the exanepsios (second cousin). · · The grammar of the descriptive phrases should be noted: ol 1toc'r:ae:c -rwv&:a,.e:).,_qifuv, o"t"ou&:ve:q;LOu ul6i;, ol "t"WV &:ae:Aqi&iv 1tociosc; mean Ego's siblings' \ children, Ego's cousin's son, and Ego's siblings' children respectively; while ot &:ae:).,_qi&v ulo£ means the sons of those who are siblings to one .another. Probably some such-very necessary-convention was esfablishby Aristophanes, whose exactitude is great. It is not followed consist'.)\\~ntly in these testiinonia: the grammarian is less careful than the scholiast, ;while Ammonius can safely neglect it. I have marked the passages critically, but more for the convenience of the reader than in any belief, that copyists only are responsible for the lack of consistency.
,:ed
124
, CLASSICAL ATHENIAN
THE FAMILY UNIT
FAMILY LAW
The most detailed passage on the anepsiadous comes from Book III of . the Thesaurus written by ,Pollux in the second century AD, some four· centuries later than Aristophanes of Byzantium. This work is extant only in an abridgement, and seems to have differed widely from A,ristophanes in the treatment of general terms. So far as can be seen, Aristophanes would take yevoi:;and cruyye:ve:(cx in the pure Greeksenses of descent and kinship respectively, but Pollux uses them (III 5) as equivalents to the Roman c~gnatio and affinitas. As Greek kinship terms were preponderantly bilateral, and Roman in stricter usage unilateral, part of Pollux's treatment of the anepsiadous is probably due to the fact that although he was · writing in Greek, he was thinking in accordance with Roman practice. His account is: (e) Pollilx, Onomastikon III 28 f (Bethe): mxALv't"o(vuv&oe:A