miller 1974

Upload: ryanboehm

Post on 08-Apr-2018

216 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/7/2019 Miller 1974

    1/31

    The Altar of the Six Goddesses in Thessalian PheraiAuthor(s): Stephen G. MillerSource: California Studies in Classical Antiquity, Vol. 7 (1974), pp. 231-256Published by: University of California PressStable URL: http://www.jstor.org/stable/25010672 .

    Accessed: 26/02/2011 14:25

    Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .

    http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

    may use content in the JSTOR archive only for your personal, non-commercial use.

    Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .http://www.jstor.org/action/showPublisher?publisherCode=ucal. .

    Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

    page of such transmission.

    JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

    content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

    of scholarship. For more information about JSTOR, please contact [email protected].

    University of California Press is collaborating with JSTOR to digitize, preserve and extend access to

    California Studies in Classical Antiquity.

    http://www.jstor.org

    http://www.jstor.org/action/showPublisher?publisherCode=ucalhttp://www.jstor.org/stable/25010672?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/action/showPublisher?publisherCode=ucalhttp://www.jstor.org/action/showPublisher?publisherCode=ucalhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/25010672?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=ucal
  • 8/7/2019 Miller 1974

    2/31

    STEPHEN G. MILLER

    The Altar of the SixGoddesses inThessalianPherai

    The purpose of this study is to present two recently discovered ancient blocks from Pherai in Thessaly.1 Because these twoblocks are only part of amuch largermonument, and because they haveseveral self-contained anomalies, this presentation must consider thequestions of reconstruction and interpretation. Although speculativeand incapable of proof, the reconstruction and interpretation of themonument offered here will, I hope, be taken as serious and possibleexplanations of the peculiarities presented by the blocks and the areawhere theywere discovered.

    I. DESCRIPTIONBlock A, Volos Museum no. E 1270 (fig. 1, pl. 1:1)

    Height: 0.685 m.Length: 0.993 m.Thickness: 0.372 m.Letter height: 0.015-0.016 m. (lines 1-2),0.026 m. (line 3)

    1For permission to studyand publish thismonument Iwould thank the thenEphor of Antiquities of Thessaly, D. R. Theochares. For helpful comments and suggestionsduring thepreparation of thispaper Iwould thankVictorine vonGonzenbach, Ch. Habicht,M. Jameson, Stella G. Miller, R. Stroud, Maria Theochare, and Dorothy B. Thompson.Their assistance does not necessarily imply that they subscribe to all the hypotheses presented here. For financial support during the initial phases of study I would thank theAmerican Council ofLearned Societies (see infran. 7).

  • 8/7/2019 Miller 1974

    3/31

    ?'?~'::.:~?~~?'?~?;: ._,1, .

    ''m

    11, - ? ? ? - -I

    -~~~~~~~~~~~~~~~~~~~~~~~

    iS:~~~~~~~~~~~~~~~~~~~~~~~~~~~~?f:QS

    I Block1 A.

  • 8/7/2019 Miller 1974

    4/31

    Altar of the Six Goddesses 233Dark bluemedium crystalmarble with white veins forminga beige-brown patina with exposure to the weather. Complete exceptfor chips and bulldozer scars. Top left rear corner broken away, but

    joining fragment survives (pl. 1:2). Right end worked smooth with finetooth-chisel. Anathyrosis on left end, back, bottom, and top. Lattersurface has hook clamp cutting to left,hook clamp cutting to rear nearleft end, and Z clamp cutting to rear near right end. Front worked withfine tooth-chisel, taenia 0.060 m. high across top, series of five-and-onehalf low relief (0.003 m.) simple stelai below. Each stele has threeinscribed lines the first two of which have been erased but are stilllegible. The hand inscribing each line is different, although the firsttwo lines appear very similar in character (pl. 2: 1,2):['tl]]ak1c 4 71p7?P ['Ev]oSla Aqposrrl'rva AQeyls >'Icra[]a J-777TP 'Ev[oS] a AO'1(v[a] Apf>po 1]r @9,4[ts] --[fE]iuS +Apotn ASi Ova 'Evo&Sa JrqprS p 'Ica ->Block B (fig. 2, p. 1:3)

    Height: 0.685 m.Length: 0.978 m.Thickness: 0.253 m.Dark blue medium crystalmarble with white veins form

    ing a beige-brown patina with exposure to the weather. Completeexcept for chips. Right rear bottom corner broken away, but five smalljoining fragments survive. Left end and face worked smooth with finetooth-chisel. Anathyrosis on right end, back, bottom, and top; lattersurface has hook clamp cutting to right, hook clamp cutting to rearnear right end, and Z clamp cutting to rear near left end.

    II. PROVENIENCEThe Acropolis of Pherai2 is a steep-sloped hill situated atthe southwestern outskirts of the modern village of Velestino in Thes

    saly. Although overshadowed by Mt. Chalkodonion to the south,3theAcropolis isprominent and easily distinguished by its broad tabletop configuration. To the north the land at the foot of the Acropolissteps down gradually in a series of broad terraces toward the so-called

    2Cf. F. Stahlin, Das hellenische hessalien(Stuttgart 1924) 105, fig. 5,Hill 150.3Cf. Apollonios Rhodios 1.49-50 and Scholion ad loc.: opos vTrrepavw 'epwv 'TO

    XaAKcwSo'vov.

  • 8/7/2019 Miller 1974

    5/31

    :,?1? '?.'? ??. ?:?.?''' '?? ?:: ' ??'?;?:. : -, .-..-.. -? ?.. ?.?????.??. ..?????.....?? ? ???.?'.?'??..' '.?'.?. ?? '.:?'..:;?'?'. ".?'.?.'' '

    ?lr?T?;'"?s??r??I?2????J"1?2*i M":"...."`'??I?; ??-?i :?:??:.?:?2?--5Li?i -?r'?T..?.1;,.-?-:i-????;.?;?;?l:;F.. ;; ;., ,,?????????to??";??f???;111,,,Rc????.,r?:?-???."' ?;L3CI??`E:'r_.er8ir,?"s; ,I? =???"Ct''-':::?:????=`?''Jfl:'??' ,?tl?t.i??zf?i?i:'i `i??- ; L?'`51:?;?3;a 19':?:'????r???- ?:....:ii???,;-?rl??f?-?tr?f"a I'.'' .I1.:11.Ilil.?1.C-?ai???c ? -,??I. i?.`; IF????i ;??.: ,:,, .??:?:???.Il?Li' i::?-r, ?:,.?-C` r? r?;???? ?ja '`''Yj?iil?tLc?; ?= CI?_ ::i1?111?ll(r;?li.l(k%? ;I??-?`t:Ssr-??: ;I?"`-P"i;??*i?*--.-??.1?`:.(???.;I;:????,?; =??5?11*?-?:????1(?IIll,.L. i' . ??E.:?C1?; I??S?1:?ll.f2211n?It?i 1112 Jr;i?\ ? - T.,1?I?'"'?:dil?I:"?:"?:"4?-( ?-?';?t==?,:`?\,-'li`'?'"-?2?"?';1;5 :,,,.?..fl...

    ?.: ??? '' ?? O?"?`??;.'??`. .?.. ? ??='?- ' ?'?? ???.`????.?':* ,??=??'I,?.??' ?? ?..?.,r?? , ?-?.._=- ..?? . ?; ?'? : ?'??. .....-.

    2. BlocklS.

  • 8/7/2019 Miller 1974

    6/31

    Plate i Milleri , ' ' a" ' . .r,

  • 8/7/2019 Miller 1974

    7/31

    Miller Plate 2....... :i~...'." ,* I..... . '{.J! ~- ' ."" . ...I?,~ ,;',~, . .~ u. .....,..............?J,, , ,;~, .~.'~ ........:..... ,:,, , ,,~~;4.~~~~~~~~~~~~?

    Z..*P....."~

    r~~~~~i""* '"';" i

    ?LI'~~~~~~~~~~~4

    ~~C?~~!~~d~~i~~rt~? :. i

    ~~~~~~~~~~~~~~~~~~~~~1,W..

    ? .. ? i.. . ,.

    ~'~'',~,~..A.

    ?.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I..,etail ;of incrpto on*,,::, fourt'stel" fro righ end!Iof Block:::.

    i lSPI:;~:'?~ ~i ...:.-~, ~-::!. ,"...~~**r:;iT~~~~~""....1

    ?..: . ...... ..:1.,.. ,;",. ~:.~ ,' i.....?, : 3~~~~~~~.. .....~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.P -::..::: ?:~? ,...; ~~.i..~,~~~~~~~~~~~~~r~Llf~:~a -.- . .-.~ .,~ ~~* .:;.~-:,-~'~-': "lk?~,~ 4Ei ;! ' '~:'.~.-.w~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~,' :.:"'" ;if":

    7 ? ?, :,~ ......~~~~~'* ?" " .....

    ?r ,':,, ~.: ?r ~ .fl ~ :'...? sI.~~~~ ..,... .: . .... . . . . .~?* ? ii'ir~~~ ~ ~ ~~; *~t.?t, : i--:~"~. t*~~~~~~~~~~~~~~~~~~~~?i::if .' ,?;. : 'E~6 ,,'~,:? "- : ,' "I w ?~ ?~~~~~~~~~~~~~~~'~....~.?

    ~~~, ~ ~ ~ ~ ~?~~. ~.. ~:-:~

    :~~~~~ ~ ~ ~~~~~~~~~~~~~~.. ." ...'.....

    !~.....'-. ":'..r;~r' ... ......,,:,,% ~*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~! , '':i~,e:,

    ?rr ~~~~~ ~~~~.~.':: . ~, 'E.~..~.'1{::.3i~!'~'~'~_~ 11.~.~..'l .....

    .i~~~~~~~~~~~~~:......FY'~~b~~~:ly?? :i~~~~~ii ~ ,,.:: ;:~+" , ': , ~..,~:,'~;~ ~~~~~~~~~~~~~~~~~~~~.:.. ?i~,:,~.-.'~, !;.; . .

    ~ ~~ ~ ??. Dealocito n?ut t1 ~m~?tedolc .

  • 8/7/2019 Miller 1974

    8/31

    Plate 3 Miller

    V ~ C*3e .. .

    .... :'.'i:-;P 4T :1 .:...

    ~ _ ..,;,:..~ :'.~,.....-.*..'"'; ",?; :. ,

    Wb. X *, i ?, *-r +

    _.,,~.~ : ,'~- ... ' .; ........,O.O."b SS, %,, ; ;' 0- X,

    . Acropolis of Pherai from west*,, ;Mt. Pelion in background, arrow at Block B.;2w~t=.Bg i ' , T e ,,41--?. -?

    1. Acropolis1--f1i-hra fromwet t ein nbcgonaro tBokB

    :, . .. .:;?,..:":;;.'.:.... ...'...""..'..

    _ ;D.i ???::jG

    ,.X '.. .~s,,;;~,-?,~

    ?5: : ::1F

    M, ~ ~ ~ ~ ,

    2.ree1 m _a headseo

    _l~~~~~~~~ ~~~~ * I1

    2. Three mable female eads embeded in porosslab in Larssa Museum

    2~~~~~~~~~~~~~? .heemrbefeal easemede i orsslbinLaisaMsem

  • 8/7/2019 Miller 1974

    9/31

    Miller Plate 4

    1. Relief stele from Pherai dedicated toZeus Thaulios inHalmyros Museum.

    2. edication to Dionysos from Glaphyrai

    41i..: . usu

    .:.. ..... ....... . X

    .e....,':S

    4. Bomiskos dedicated to Artemis Enodiafrom Demetrias, in Halmyros Museum.

    ??;;-;:? ~ ~~~~~~.:.C'

    ':"

    ::::::~~~~~~~~~~"''",::".:~,**,.

    ..":: . . '~ ...2..

    i~~~~~~n~~~~~l?: ':" " "'7~:' .

    A..

    4.Bomiskos dedicated tooseidon from Pheraifrom Demetrias inoHalmyros Museum.

    .:.;;vso

    A~~~~~~~~aiz

    |;i_~~~~~~~~~~~~~~.. ~ ~ ~ ~ ~ r ?8":3.~~ ?edcaio t Psedo frmPeainaoo uem

  • 8/7/2019 Miller 1974

    10/31

    Altar of the Six Goddesses 235Temple of Zeus Thaulios4 while behind, to the south, theAcropolis isconnected by a low ridge to another height.5 On the east, the sidewherethe ancient city of Pherai lay, the slopes of theAcropolis aremore steep,as is also the case on thewest. On the latter side the slopes end in a drystreambed, the 'Makalorevma', which seems to have formed the lineof the citywalls in this area.6About halfway up thiswestern slope thereis a fairly narrow naturally formed ramp sloping upward toward thesaddle which connects theAcropolis with Hill 154 to the south. It wasduring the bulldozing of this "ramp" in order to create a broader,flatter, and hence more arable area, thatBlock A was found.

    When firstmoved to themuseum at Volos, Block A wasmissing its upper rear left corner. The missing fragment was found inthe area of the bulldozed ramp of theAcropolis by the author inAugustof 1972. At the same time, Block B was found just below the ramp andcan be discerned in Plate 3:1 as the white dot to the left of center.7The discovery of two blocks of the same monument in the same arearaises the question of whether the blocks had been in situ before beingunearthed by the bulldozer. Another possibility is that themonumentonce rested on the top of theAcropolis near itswestern edge and thatthe blocks fell down onto the ramp and came to rest close to one another. In either case, since the foundations of the altar are presumablystill in place, excavations at the site could determine the original provenience of the altar.

    III. DATESince the blocks were not found in situ, the only evidencefor dating is that contained on the blocks themselves. The hook clampson the top surfaces are most characteristic of theHellenistic period, and

    ought not to be earlier than the fourth century B.C.8The Z clamp (our4Y. Bequignon, RecherchesarcheologiquesPheresdeThessalie (Paris 1937) 29ff.For the question of the identification of the remains seeE. Kirsten, RE Suppl.7 (1940) 998999.5Hill 154. The connecting saddle is the site of the Church of the Panaghia, a

    modern reconstruction of which is presently cutting through a previously well preservedsection of theancient citywalls.6Bequignon, op.cit. 15.7Both Block B and the joining fragment of Block A were taken to the museum inVolos inAugust of 1972. For provision of the funds necessary for workmen andtransportation of the block, aswell as personalmaintenance inThessaly Iwould thank theAmerican Council ofLearned Societies.8R. Martin, Manuel d'architecturerequeI (Paris 1965) 273-279.

  • 8/7/2019 Miller 1974

    11/31

    236 Stephen G. Millerversion is actually a double F) ought to be earlier than the hookclamps and is usually found in late Archaic or early Classicalcontexts. There are, however, examples known from the fourth century,themost significant of which is in the Daochos dedication at Delphi.9The letter forms of the inscriptionsmay give a somewhatcloser date. Although there are differences among the three lines (e.g.the sigmas of the last line are more elongated than those of the first),the similarities are more pronounced and argue that the dates of thethree lines are relatively close. The slightly splayed tips of the letters ofall three lines should place themwell into the fourth century, and thereisno primafacie reason fordating the lettersof the last line later than theearly third century B.C.10On the other hand, theremight appear to be a problemwith the orthography of the inscribed names which consistently uses etarather than the typicalAeolic-Thessalian alpha. This use of r for amustbe an example of the intrusion of the Attic-Ionic koine which becamevery widespread inThessaly, but not until the late second or early firstcentury B.C.11For example, the letters of Philip V to the ThessalianLeague from the late third century B.C.12were originally written inAttic but had to be presented in Larissa together with a Thessaliantranslation. This general tendency need not preclude, however, theearlier use of the koine inThessaly. Indeed, the closest parallels for theletter forms of our monument appear in an unpublished inscriptionfrom Pherai written in koine and probably dating to the middle of thefourth century B.C.13

    It is always tempting to associate monuments with knownhistorical events. If the real date of the altar is in themiddle of the fourthcentury, and the tenuous nature of the chronological evidence must beemphasized, then the damage and repairs to themonument (seebelow)

    9 Ibid.256.10For the closest parallel to the letter forms of the presentmonument seeinfra n. 13. 11Cf. C. D. Buck. The GreekDialects (Chicago 1955) 151, and ThumbScherer,Handbuch ergriechische-Dialekte12(Heidelberg 1959) 55.12IG IX2, 517; for the proper date seeHabicht. "Epigraphische ZeugnissezurGeschichte Thessaliens," AncientMacedonia (Thessalonike 1970) 273-278.13Volos Museum no. n 98. I would thank Ch. Habicht for calling myattention to the existence of this stone, forallowingme to use his squeeze of it forcomparativepurposes, and fordiscussing itsdate with me. Mr. Habicht argues convincingly that the textwhich is a proxeny decree of the Pheraians must, on historical grounds, date to the periodbefore theMacedonian domination ofThessaly.

  • 8/7/2019 Miller 1974

    12/31

    Altar of the Six Goddesses 237may have been a result of the capture of Pherai in 344 B.C. by Philipof Macedon.14 Another possible time for this damage is the siege ofCassander's garrison at Pherai by Demetrios Poliorketes in 302 B.C.,15but neither suggestion can be proven. It is perhaps better simply todate the monument to the late Classical or early Hellenistic period.16

    IV. RECONSTRUCTIONAs previously mentioned, Blocks A and B are twomembersof a largermonument. These two blocks belong together back to back asis shown by the similarity of heights and lengths, by the alignment ofthe clamp cuttings, and by the correspondence of theZ and hook clampsof the one block vis-a-vis those of the other. This joint establishes the

    depth of the monument as 0.635 m., and provides the additional information that the back and one end (probably both ends) of the monument were plain, while the front was covered with a series of lowrelief stelai each of which bore the name of a female divinity. Althoughwe lack the left end of themonument, its original length can be estimated. This is so because the successive lines of the inscription, the firstreading left to right, the second right to left, both begin and end withHestia and Themis which means that six stelai originally comprised acomplete unit. It will be shown below that the six goddesses listed onthese stelai comprised the Pheraian version of the female members ofthe Twelve Olympian Gods. Thus, there are two possible reconstructionsof theoriginal monument:

    1. There were six, and only six, deities worshipped whichmeans thatwe have to restore only the remaining half stele and a margin to the leftbalancing that to the right.This would be done bymeans14Dem. 8.59; cf.Dem. 7.32; 9.12; 19.260.Bequignon, op.cit. 19, argues thatPherai sufferedno damage during the take-overby Philip, but the long history of antipathybetween Pherai and Philip would not lead one to believe that Philip was particularly tenderwith the city; cf. H. D. Westlake, Thessaly in theFourthCenturyB.C. (London 1935) 192.B&quignon further attempts to assign traces of burning which he noted on the Acropolis(p. 15) and at the so-calledTemple of Zeus Thaulios (p. 30) to the later siegeof theAcropolisbyDemetrios (infra).Bequignon does not adduce any archaeological evidence forhis dating ofthe fires, and the two fires need not have been from the same cause, nor even at the same time.15Diod. Sic. 20.110.6.

    16However, the best time for the establishment of a monument such asours would seem to be during the Pheraian tyranny, either under Alexander (369-358 B.c.)or under the sons of Jason, the last of whom, Peitholaos, was ejected from Pherai by Philipfor the third and last time in 344 B.c. Although perhaps of no significance, it is interesting toremember that the Altar of the Twelve Gods in Athens was also established by a tyranny;cf. Thuc. 6.54.6-7.

  • 8/7/2019 Miller 1974

    13/31

    238 Stephen G. Miller

    of a thin slab, 0.12 meters thick,with the remaining half stele carved onone thin end, and having the full depth of themonument of 0.635 m.(fig. 3). The total monument length will have been ca. 1.113 m. Thetype of construction thus restored, with two relatively thick blocks set

    backo back and a thin slab masking their ends on one side, is unusual

    andmay not be original. This possibility issupported by a consideration

    3. Possible reconstruction of central course of monument.

    of the difference in the forms of the clamps used on the top surface ofthe twoblocks.We might thinkof themonument as originally consistingof the two blocks A and gthill haveere 1.113 m. long and bound together by a Z clamp at each end. The left end of the monument wassubsequently damaged, Blocks A and B smoothed off on this end, andthe end covered by the block restored asmeasuring 0.12 x 0.635 x0.685 m. The original Z clamp at the left (corresponding to that stillpreserved at the right) will have been cut away during the repairs andreplaced by a new hook clamp further to the right. The new slab willhave been bound to the ends of both Blocks A and B by a hook clampto each block.17

    17 If the monument is, as will be argued below, a part of a Dodekatheon,then this reconstruction means that there will have been two separate altars with six deitieson each. It will be seen below that the gods were in four groups of three in the Delian Dodekatheon, and therewas awidespread tendency to split the twelve into fourgroups of three, sixgroups of two, or two groups of six in both literature and art; cf.O. Weinreich, Roscher s.v.Zwolfg6tter 842.

  • 8/7/2019 Miller 1974

    14/31

    Altar of the Six Goddesses 2392. The second possibility is that the altar was dedicatedto all Twelve Gods with the sixmale deities to the left, the six female tothe right. In this case the altar will have had an original length of ca.2.16 m. Such a possibility, however, would cause an original verticaljoint to fall through one stele,18would fail to provide an explanationfor the difference in the forms of the clamps, and would take no accountof the fact that the three successive series of inscribed names form a

    complete unit as they stand.19 The first suggested reconstruction seemsmore probable.The anathyrosis on the bottoms of the blocks shows thatthey rested on another stone surface with which a close fitting joint wasformed. Since themonument is finished on the preserved three sides,wecan best imagine the lower element as a stepped krepidoma with a free

    standingmonument on top.Another element obviously belonged above Blocks A and Bas is shown not only by the anathyrosis, but also by the otherwise exposed clamp cuttings on their upper surfaces.The nature of thismissingelement cannot be restored with certainty and there are several possibilities-as, for example, a simple moulded flat slab, or a slab withbarriers (plain or decorated) at each end.20Another possibility is suggested by a consideration of the inscriptions on the front of the monument. In the original line the goddesses were named in the series:Hestia, Demeter, Enodia, Aphrodite, Athena, Themis. In the secondline, the series is retrograde, but almost the same: Hestia, Demeter,Enodia, Athena, Aphrodite, Themis.21 Finally, each name in line 2was turned around in line 3 so that the series reads from left to right,but is now reversed: Themis, Aphrodite, Athena, Enodia, Demeter,Hestia. What is the significance of these changes ?Itwould appear that the solution to this question is to be18With the firstpossibility discussed above, this vertical jointwill have beenthe result of later repairs,not a part of theoriginal construction.19A variation on this reconstruction would be to restore a plain centralsection and another group of six stelai further to the left, thusmaking one longmonument.Such a reconstructionwould require,however, the restoration of somany missing pieces that Irefrainfrom suggesting it.20 Cf. C. G. Yavis, Greek Altars (St. Louis 1949) passim, for the variouspossibilities. 21The only difference in the two series is the inversion of the sequenceAphrodite-Athena toAthena-Aphrodite. There may be a significance in this inversionwhichhas eluded me. It seems better to regard this change as an error on the part of the inscriber ofline 2.

  • 8/7/2019 Miller 1974

    15/31

    240 Stephen G. Millersought in one of two areas. The firstwould be in the field of religioussuperstition ormagic. If the explanation does lie here, it is not known tome. It cannot, however, be discounted as a possible explanation sinceThessalian women in general,22 and the priestess of Pheraian Enodia inparticular,23 were well known in antiquity for theirmagical powers.In this same religious context one other peculiarity of the altar should benoted. The first named goddess in line 1was Hestia and the last namedgoddess in line 3 was also Hestia. It may be purely coincidental, butwe should recall the proverb o' 'Earias aPXeaocat or E' 'EaEras apXEraOat.24As explained by the story of the birth of the children of Kronosand Rhea (their successive consumption by Kronos, and the stonesubstituted for Zeus which caused Kronos to regurgitate his children inreverse order), Hestia was both the first and the last of the children ofKronos and Rhea,25 and was worshipped first and last in sacrifices.26Since, however, the first line of names on our altar was probablyerased by the time the last line was inscribed, any connection betweenthe proverb and the changes in the inscribed names on the altar wouldseem to be fortuitous.

    A second possible answer to the question of themeaning ofthe name changes ismore prosaic and concerns the possible form of theupper member of themonument. If one considers the names to havebeen not merely a list of the divinities worshipped at this altar, but ashaving amore specific function as labels, it is possible to imagine thatthe upper member of the altar contained images of the goddesses andthat the inscriptionswere intended tohelp the viewer distinguish amongthe images. Such a type ofmonument isnot well-known and yet there isa certain body of evidence which suggests that an altar with a series ofstatues or statuettes of gods on top was fairly common in antiquity.The evidence has to do with various processions which included imagesof deities, and specifically with images of the Twelve Gods who wereworshipped as a group. A few of thepertinent sources are:

    22See, forexample, Aristophanes, Nubes 749-752; Ov. Met. 7.74-99, et al.;Apul. 2.1; Lucian, TheAss, passim.23Polyainos 8.43.24 PI. Euthphr. 3A; Kra. 401 B, D; Hesychios, s.v. &a' 'Earlra &pXo'Pevos; cf.Aristophanes, Vespae864.25Hes. Theog. 453-497.26 Cornutus c28 (p. 53, Teubner): MvOEverat 8e (sc. 'Eacrra) 7rprrq rE Kal&

    eaxanT7 yveveaOa rO eTi rearL vaAvEVraffa7' auTs yLvo4eva Ka& i aSr^ avviacrlaOa, KaOOKaV rTaL OVaucrsT lt 'EAves arro rp'wrrs Te aVcTrs PXOVRoKat els cacrXyaTvavorTv KaTcirauov. Cf.Cicero, De natura deorum 2.67; Homeric Hymn 29.4-6.

  • 8/7/2019 Miller 1974

    16/31

    Altar of the Six Goddesses 2411. At Aigai inMacedonia in 336/5 B.C.Philip is said tohave paraded the images (E'3Zs)a) of the Twelve Gods and to haveadded a thirteenth figure of himself which was included in the procession.272. At Rome there was an annual procession from the

    Capitoline through the Forum to the Circus Maximus which was saidto have been based on similar Greek processions and which includedimages (EdKo'VE)f the Twelve Gods carried on men's shoulders.283. At Magnesia on the Maeander the stephanephorosof the city was each year to carry the images of the Twelve Gods(odava 7Tdvrovv 'rv 83csE&Ka E6v) in a procession.29

    Our concern here is not with the processions but ratherwith the place of residence of the images of theTwelve Gods during therest of the year and with the fact that they could be removed andcarried about. Although the sources cited above do not indicate wherethe images were housed, it is logical to assume that they were kept alltogether in one place inasmuch as theywere worshipped as a group andparaded as a group. Such must have been the case atDelos, for example,where the attested twelve statues30were almost certainly housed in theDodekatheon.31 This shrine has been identified32 and the placementof the twelve statues was probably on a series of bases in front of thetemple in the temenos.33An orthostate of one of these bases survives34and the face of this orthostate is inscribed with the three names [AG10]vs AJto "Hpas.The Delian monument will have looked, then, something like themonument in a wall painting fromHerculaneum.35The purpose of the preceding discussion has been to show(1) that therewere in antiquity sanctuaries of the Twelve Gods whichcontained statues or statuettes of these deities; (2) that at least in some

    27Diod. Sic. 16.92.5.28Dion. Hal. 7.72.13; cf. 7.71.3.290. Kern. Die Inschriften von Magnesia am Maeander (Berlin 1900) 83, no.98.41-42.

    30 IG XI2 158.65: Tr&SoSEKa aytAcTaa; cf. IG XI2 203.45 and 287.72: ra&USEKa. 31Insc.Del. 401.12, 403.37, 406.36, et al.32R. Vallois, "Topographie Delienne II," BCH 53 (1929) 225-249.33 P. Bruneau, Les Cultes de Delos (Paris 1970) 439.34 E. Will, Ddlos XXII, Le Dodekatheon (Paris 1955) 157-158.35H. G. Beyen, Die pompejanische Wanddekoration II (The Hague 1960) fig.139.

  • 8/7/2019 Miller 1974

    17/31

    242 Stephen G. Millercases these figures could be detached and carried in processions; and(3) that all twelve figures did not have to be set into a single base. Thelast point shows the possibility that themonument from Pherai

    was oneof two bases in a dodekatheon with the sixmale deities on one base andthe six female deities on another.36The firstpoint permits the suggestedreconstruction of the upper member of the monument as consisting ofstatuettes of the goddesses, and the second point would permit thosefigures to have been removable. But if the images on the Thessalianaltar were full-figure statuettes, the use of labelswould not have beenabsolutely necessary.37 There is, in our Thessalian context, anotherpossibility. In the museum of Larissa there is a curious object thepurpose of which has never been explained (pl. 3:2).38 Found atDilophos (Chalitsi) south of Larissa and about 20 miles northwest ofPherai, this object consists of a large poros slab (about 0.20 x 0.50 x1.00 m.) with three circular holes cut through it. Into these holes areset the marble heads of three females with very similar features. Theheads are held in place by a collar of lead around their long shaft-likenecks. Ifwe restore some such upper member on our altar, an explanation for the reversal of the inscriptions on the face of themonumentbecomes possible. A conjectural history of the monument might havebeen something like this:

    1. The altar is constructed as a free-standing monumenton top of a stepped krepidoma with an upper member consisting of asingle slab pierced by six holes intowhich were set the heads of the six

    36Cf. supra n. 17.37 In the same area where the blocks of our monument came to light therewas discovered the fragmentary torsoof a marble statue of Athena. Although under life-size,this statue is still too large to allow its associationwith themonument. The publication of thetorsobyG. Bakalakes isforthcoming inThessalika.38Larissa Museum no. 948; cf. ZeAT7ov 16 (1960) 183-184. The piece wasdiscovered by Mrs. M. Theochare whom I would thank for permission to cite it here, but hasnot yet been fullypublished.Although thisThessalian three-headed slabmight seem extraordinary, one

    might remember the female terracotta heads of the second half of the fifth century B.C.discovered at Brauron. These heads, like those of the Dilophos slab, have elongated neckswhich obviously served tohold them inplace; cf.BCH 87 (1963) 707, and ERGON 1962, 3032. An effect quite similar to that of the Dilophos slab is achieved by a series of miniaturebronze heads recently discovered in Etruria which were set in stone bases. In this case, however, a smaller tenon projects down from the base of the neck to serve the same function asthe elongated necks of the Dilophos and Brauron heads; cf. M. A. Del Chiaro and A. Talocchini,AJA 77 (1973) pl. 60.5.

  • 8/7/2019 Miller 1974

    18/31

    Altar of the Six Goddesses 243goddesses. The front of the monument is carved with six low reliefstelai, one below each goddess, and inscribed with the name of the deityportrayed above.2. The altar isdamaged and in the subsequent repairs theleft end is trimmed and replaced by a thin vertical slab onto which iscarved the left half of the first relief stele (see above and fig. 3). Theupper slab is set back on top of the repaired monument, but turnedaround 180? (by mistake?) so that the backs of the heads of the goddesses are toward the front, and the stelai are reinscribed in retrogradefollowing the original sequence. Thus Hestia, originally the first on theleft facing the front of themonument, isnow the last on the right facingthe back of themonument.

    3. The "mistake" is corrected by turning each headaround individually (after an annual procession of the Twelve Gods?)so that it faces the front of the altar, and the stelai receive their finalreinscription in the form of turning the individual names around so thatthey read from left to right, but in the reverse of the original sequence.Both the head and the name of Hestia, for example, originally first onthe left facing front, then last on the right facing back, is finally last onthe right facing front (fig.4).

    Although the suggested reconstruction of the altar and itshistory ishighly conjectural and cannot be proven, itwould explain thepeculiarities on the front of themonument. But if the reconstruction iscorrect, the form of the monument isunique. A possible parallel mightbe a rockcut relief of theRoman period in France,39 but the differencein time and place is too great to allow even a hypothetical connection.Similar to the suggested reconstruction of our monument in spirit, butnot exactly equivalent in form, are threemonuments found in the Bayof Naples and dedicated to the Syrian god Dusares.40 All three areshaped like normal altars. Two have rectangular cuttings for threeobjects on top,while the third largermonument has cuttings for seven

    39A. Blanc. "Nouveaux Bas-reliefs des Deesses-meres et du Dieu au Mailletchez lesTricastins," Gallia 25 (1967) 70, fig. 3. It isequally possible that in the relief citedwehave to do not with three heads set upon an altar, but rather with a compression of the spacedividing two zones so that the relief should be considered in two zones with three busts in theupper and an altar in the lower, a much more common composition for this type of monument. Note also the circular Roman altar in the Louvre with the heads of the Twelve Godsapplied around the rim; cf.Weinreich, op.cit. (supran. 17) 826, fig.8.40V. Tran Tam Tinh, Le Culte desDivinites orientales nCampanie(Leiden1972) 127-131, 144-146, figs.65-67.

  • 8/7/2019 Miller 1974

    19/31

    1.o

    ?T~-_"*.~~~~~~~~~~~~~~~~~~~~~~~'r.. ZZ VC~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~CCC

    INS,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    4.~, Pb r nt t t1g4. Possible reconstructionof the altar of the six go

  • 8/7/2019 Miller 1974

    20/31

    Altar of the Six Goddesses 245objects. These objects were small simple stelaiwith rounded tops, as isshown by the discovery of four of them near the larger altar. The lowerpart of each stele is provided with a tenon which fits neatly into thesocket on topof the altar.If none of thesemonuments provides a precise parallel forthe suggested reconstruction of the Pheraian monument, they at leastshow that curious monuments do exist elsewhere in the ancient world,and that the reconstruction suggested above ought to be considered asa possibility, however remote.41 It should be remembered, moreover,that the suggested reconstruction may be rejected but the suggestedhistory retained by reconstructing full-figure statuettes on the upperelement.

    V. THE FORM OF THE ALTAREven without the hypothetical reconstruction of the upperelement, there are peculiarities which can be best explained in a localPheraian context. For example, the idea of individual stelai carved inrelief on a largermonument is not easily paralleled, and yet there isafragment of another monument from Pherai which must have beensimilar to our monument in this respect. This is a fragment of marblefrom the upper left corner of amonument upon whose face was carveda stele in relief (pl. 4: 1).42The inscription on the face of the stele is adedication to Zeus Thaulios.43 Unfortunately, we cannot know if therewere more than one stele on the face of themonument, and the form ofthe stele ismuch more developed (e.g., pediment, acroteria, etc.) than

    that of the stelai on the Pheraian altar.41Although I have sought toexplain the curious nature of the inscriptiononthe face of themonument in termsof theupper element, other explanations are possible.Onewhich has been suggested is that lines 1and 2 are contemporary on Block A and represent alater copy of an earlier inscription as indicated by the use of retrograde in line 2. These linesare then replaced by the new "edition" of line 3 which does away with the use of "oldfashioned" retrograde in line 2. The reproduction of lines 1 and 2 on themonument is to beattributed to a religious conservatismwhich is, however, then disregarded by the erasure ofthose linesand the inscriptionof line 3.42Now in theHalmyros Museum, no. 27; cf. Bequignon, op. cit.94, no. 71,and N. Giannopoulos, 'Apx. 'E+. 1913, 218, no. 1.Perhaps themost obvious parallels, although in a funerary context, to the form of our altarwith individual stelai simulated on alargermonument are those bearing the fifth-centuryAthenian casualty lists; cf. D. Bradeen,Hesperia33 (1964) 23-29, andHesperia36 (1967) 324-328.43The meaning of the epithet is not agreed upon; see F. Hiller von Gaertringen, Hermes46 (1911) 154-156; F. Solmsen, Hermes46 (1911) 286-291; V. Costanzi,Athenaeum 914, 49-51.

  • 8/7/2019 Miller 1974

    21/31

    246 Stephen G. MillerThese stelai are, in fact, one of themore intriguing aspectsof themonument. Simple pointed stelai, although common as gravestones throughout Greece, are extremely rare in other contexts. I havebeen able to find only two other instances of the use of this shape inconnection with the gods.1. Volos Museum no. E 432, fromGlaphyrai (pl. 4:2)44

    Height: 0.955 m.Width: 0.235 m.Thickness: 0.060 m.The stone has suffered greatly since its discovery and only

    tracesof a few strokes still survive in the last three lines. I reproduce thetext of the Corpus.'EpwkwvvI;&VLEVOSdLOWVevos Atovv

    awruVSeaVO7 KEKal nappeVLXOsKayCtK[a]va8pos

    2. Volos Museum no. E 1019, from Pherai (pl. 4:3)45Height 0.54 m.Width at top: 0.152 m.Width at bottom: 0.19 m.Thickness (maximum): 0.081 m.NvSaczos7Trpaldovvos 17Hrel

    One might rather expect EvSapcosas the name in line 1,but the reading of the first letter as nu is secure. The Thessalian form Trepisattested at Pherai,46 as is themore usual 7Trep.4744 IG IX2, 411. This is the stone which Stahlin, op. cit. 61, n. 16, dates to theArchaic period and attributes toBoibe. I can findno reason forStahlin's change of the thirdcentury date and theGlaphyrai provenience published byKern.45Unpublished; discoveredMarch, 1940.46Bequignon, op. cit. 96, no. 78.47 Ibid. 91, no. 62.

  • 8/7/2019 Miller 1974

    22/31

    Altar of the Six Goddesses 247These two stelai are obviously aniconic monuments andthe discovery of them in the same region as our altar suggests thatPherai was the center of a well developed and even formalized aniconism.Aniconism was, of course, well known throughout Greecein the historical period, and there are many ancient references toaniconic images. Zeus,48Hera,49 Aphrodite,50 Artemis,51 and Apollo52were all worshipped in aniconic form together with other deities andheroes in various parts of the ancient world. These aniconic monuments are sometimes portrayed in other art forms as, for example, on an

    Apulian red-figured bell crater in the Cleveland Museum, where aplain stele bears the inscription APPOAITH and must be an aniconicimage of the goddess.53 Even more pertinent to the present discussionare the regionswhere a flourishing aniconism isattested as, for example,at Pharai inAchaea where some 30 squared stones in the agora were

    worshipped with the name of a god given to each.54More notable yet isTegea inArcadia where Pausanias noted a square image of Zeus andcontinued with his own editorial comment: "It seems tome that theArcadians particularly like this form."55This statement has been welldocumented by the discovery of nearly 30 such "statues" inArcadia,mostly fromTegea.56 These rETpdycywayaC ya-ra fall into two categories.One is the familiar human-headed Herm, the other a plain squareshaft crowned by a pyramid which is set off from the shaft by a scotia

    48Paus. 2.9.6 (Sicyon).49Clem. Al. Protr.40P (Samos), and Strom.418P (Argos).50Tac. Hist. 2.3 (Paphos on Cyprus).51Paus. 2.9.6 (Sicyon).52 Especially under the epithet Agyieus; cf. Harp. s.v. Ayvt&s ... . .Ayyu SEaTt KIwV EtSV V A'4ywv, Ov lariaat rpoTrv Ovpcv ... tSiovs 8 etvacl xaaov avtrovs Aiovos.Note also Paus. 1.44.2. Perhaps themost obvious ancient representationofApollo in this formis on a terracottaplaque fromRome; cf. Fasti Archaeologici 2 (1967) fig. 48.53 Cleveland Museum no. 24.534; CVA Cleveland, fasc. 1 (USA 15) pl. 43.1.Note also the stelai ofNike on bell craters now inBonn, museum no. 79, andMadrid, museum

    no. 11081; cf.A. Cambitoglou and A. D. Trendall, ApulianRed-FiguredVasesof thePlain Style(Rutland and Tokyo 1961) pl. 20.96 and 98. A stele of Zeus appears on a vase atRuvo; cf.L.R. Farnell. TheCultsof theGreekStatesI (Oxford 1896) pl. Ia.54 Paus. 7.22.1-5.55 Paus. 8.48.6: neirroL'-oal 8S Kal Alos TeXAlov3fculos Kat ayaXAia Terpdaywvov

    7Teplaacs yap 8'j Ti rZ aoXjLaT' ToV-rthoaIvovrat pLot XalpeLV ol 'APXdEae.Note also the presence ofApollo Agyieus (n. 52 supra) atTegea; Paus. 8.53.1 and 6.56 Cf. K. A. Rhomaios, ApX. 'E+. 1911, 149-159.

  • 8/7/2019 Miller 1974

    23/31

    248 Stephen G. Millerand a toros.57 The variations between these "pyramidal pillars" aresmall, and it is clear that a formalized aniconism was active inArcadiain general and in Tegea in particular. One of these monuments is ofespecial interest in that it consists of three such pillars carved from asingle block and inscribed NviuL v.58 Like our Thessalian altar, theArcadian dedication is amultiple aniconic monument. The "pyramidalpillars" of Tegea and the simple pointed stelai of Pherai are local, butwell-defined and formalized, variations of a strong aniconic influence.

    Although this interpretation accounts for the stelai on theface of the Thessalian altar, it does not take into consideration the(hypothetical) upper member with its female heads. If the reconstruction is valid, then the monument as a whole is not purely aniconic.Rather, itwould represent a stage halfway between aniconism and iconism. Such an intermediate stage iswell attested elsewhere, particularlywith reference toDionysos. One can, in fact, trace the development of

    portrayals of Dionysos from a simple aniconic monument59 to a complete anthropomorphization. Particularly instructive for the intermediate stage in this evolutionary process are the many instances invase painting where a pillar, or a column, or even a tree trunk, has amask of Dionysos affixed to it and wrapped in clothing below.60 Theaniconic image is transformed into a quasi-anthropomorphic statue bythe addition of a human head. That this practice survived well intothe classical period, and probably even later, is shown not only by theportrayal of the custom on red-figured vases, but also by the survival ofthemasks themselves both in terracotta and inmarble.61

    It was probably from some such rather temporary combination that the more formal Herm was developed.62 This is not to57 IG V2 59-66 (Tegea) and 280, 290 (Mantinea), and eight uninscribedshaftsmentioned byRhomaios, op.cit. 149.58 IGV2 65.59Clem. Al. Strom.418P; cf.Paus. 9.12.4.60See M. P. Nilsson, Geschichtedergriechischen eligion I3 (Munich 1967)207-208, 572, forbibliography.61Cf.W. Wrede, "DerMaskengott," Ath.Mitt. 53 (1928) 66-95.62H. Goldman, "The Origin of the Greek Herm," AJA 46 (1942) 67-68.Miss Goldman's additional contention, that theHerm was originally and specificallyDionysiac is not necessary to our argument; cf.Nilsson, op. cit. 207, n. 4. Miss Goldman herself

    admits (p. 63) that the mask was not unique to Dionysos, and what is important for us is theexistence of the aniconic-iconic intermediate stagewhich is best documented with regard toDionysos and Hermes. The use of the quasi-anthropomorphic herm was almost certainlyonce more universal for, as Miss Goldman points out (p. 61): "the custom of weaving gar

  • 8/7/2019 Miller 1974

    24/31

    Altar of the Six Goddesses 249deny the generally accepted derivation of the herm from the waysidepiles of stones, or ppa,which served as sign posts,63 but it is toposit anintermediate stage of anthropomorphic evolution. The manifestationof this stage is the attachment of a human head to an aniconic body.Although inconclusive, it is tantalizing to note that the earliest attestedHerm was attributed toHipparchos son of Peisistratos.64One of theseHerms is known, and it shows traces of anathyrosis on its top surfacewhich indicates that the head was attached to the shaft as a separatepiece.65 In other words, theHipparchan Herm may retain the physicalmanifestations of the first attachment of iconic head to aniconic body.

    The straight shaft with phallus, stubby shoulder pieces,and bearded human head is best known to us as a representation ofHermes, but there is evidence that this form was once used formanyother divinities (cf. notes 48-53). That this form did not remain in usefor most deities was due to the impulses toward full anthropomorphization; but in an area, such as Arcadia, where aniconism wasparticularly strong, this impulse would be resisted.We have alreadyseen that aniconism remained strong inArcadia, but iconism also intervened. One possible result of a conflict between the traditional aniconicform and the new iconic formmight well be a compromise which tookthe form which we know as a Herm. In fact, Tegea once again providesevidence that such a compromise could and did take place given theproper conditions of a strong aniconic tradition in a fairly isolated area.That aniconic tradition is represented by the "pyramidal pillars"mentioned above, but Tegea also has a series of human-headed shafts,or Herms, which would not be particularly surprising were it not forthe fact that these are all "statues " of gods other thanHermes. Aphrodite,66 Artemis,67 andAgathos Theos68 are all represented by individualherms. Even more interesting is the fragmentary multiple herm nowbroken at the top, bottom, and right side, but clearly once topped byhuman heads.69 In its present state we have only three shafts althoughments for statues could only have arisen at a time when the full anthropomorphic statue hadnot yet evolved." 63 See E. B. Harrison, Agora XI (Princeton 1965) 113 and n. 48.64 [Plato] Hipparchos228D.65J. Kirchner and S. Dow. " Inschriften vom attischen Lande," Ath. Mitt. 62(1937) 3. 66 IG V2 69 and 70.

    67 IG V2 68.68 IG V2 67.69 IG V2 73.

  • 8/7/2019 Miller 1974

    25/31

    250 Stephen G. Millertheremay have been more originally. Each shaft is inscribed with aname identifying the image: Zeus, Poseidon, and Demeter. Another

    multiple herm of six contiguous shafts is sadly defaced and its inscription badly scarred, but itwas obviously amonument of this same type.70If one now looks again at the suggested restoration of thePheraian altar (fig. 4), itwill be seen, I believe, that its spirit and motivation-its content in an art historical sense-is precisely the same asthe Arcadian monuments, even if its form is not precisely the same.Certainly there can be little doubt that Thessaly, likeArcadia, had astrong aniconic tradition, and if this tradition, confronted by the impulses of anthropomorphization, could produce monuments such ashave been seen from Arcadia, then it is possible that Thessaly couldhave produced, under a similar iconic impulse, a monument such asthatwhich has been restored.One would then have to do with an intermediate stage in the evolution of anthropomorhization where humanheads are set on top of the traditional forms of aniconic images.

    VI. THE Six GODDESSESWe now turn to the deities represented on the altar. Asmentioned before, I believe that these six goddesses represent the distaffside of the Twelve Gods inThessaly, or at least in Pherai. Four of the

    goddesses are definitely part of the canonical Twelve and require nospecial pleading for their places here. Hestia, Demeter, Athena, and

    Aphrodite are all attested in Thessaly and are all charter members oftheTwelve Olympians. Normally, one should expect Artemis and Heraas the remaining two goddesses, but nothing is normal about thismonument, and Enodia and Themis require explanation.At thispoint, it should be noted that theTwelve canonicalmembers were not immutable. Although the great bulk of evidenceconfirms the membership as belonging principally to Zeus, Apollo,Ares, Hermes, Hephaistos, Poseidon, Hera, Demeter, Athena, Aphrodite, Artemis, and Hestia, some of these gods and goddesses were occasionally replaced by other deities. Dionysos frequently usurps the seatof Hestia, and local deities were sometimes enrolled as replacements

    70 IG V2 72. For other such multiple heads upon plain bases see AA 1941,652 and fig. 130 (Locri), and E. Esperandieu, RecueilgeneraldesBas-Reliefs, Statues et Bustesde laGauleRomaine IV (Paris 1911) 3426,where five joined female heads from theMuseum ofChatillon are shown.Although broken below, these headsmay well have belonged on a baselike that shown ibid. no. 3411.1.

  • 8/7/2019 Miller 1974

    26/31

    Altar of the Six Goddesses 251for less popular divinities. Such appears to have been the case, forexample, with the original Twelve Gods worshipped at the shrine intheAthenian Agora,71 and at Olympia the deified Alpheios riverwaslisted among the local version of the Twelve.72The case for Enodia as amember of the Twelve Gods atPherai is simple and straightforward. She replaces Artemis.73 Enodia,although well-known throughout Thessaly,74 was particularly veneratedby the people of Pherai. She appears on their coins,75 and is frequentlycalled simply Pheraia, or the Pheraian goddess, in literary sources.She is a combination of Artemis and Hekate, with a dash of Alkestisthrown in.76Daughter of Zeus, Enodia's maternal grandfather wasAiolos, a satisfactory genealogy in Aiolic Thessaly, and her generalimportance in the areamay well have been connected to the prevalenceof witchcraft.77 Her iconography iswell established, not only by thecoins of Pherai, but also by fragmentary dedicatory reliefs fromMacedonia, a composite of which shows Enodia seated side-saddle on horseback and bearing a single torch.78 In other representations, as on thecoins cited above, she carries a torch in each hand.

    In addition toMacedonia, Enodia is attested epigraphically outside Thessaly at Epidauros,79 Oreos in Euboea,80 and as faraway as Issa inDalmatia.81 Pausanias, who calls her Artemis Pheraia,

    71H. A. Thompson and R. E. Wycherley, Agora XIV (Princeton 1972) 131132.72 Pind. 01. 10.48-49.73 P. A. Clement, "A Note on the Thessalian Cult of Enodia," Hesperia 8

    (1939) 200. 74IG IX2 575-578, andMnemosyne 3 (1970) 251 (Larissa); 'ApX. E+. 1911,127, no. 61 (Gonnos); IG IX2 358 (Demetrias); JeAnov 1926, 52, no. 4 = Hesperia8 (1939)200 (Phalanna). 75E.g., E. Babelon, Traite desMonnaiesGrecques t RomainesIV (Paris 1932)pl. 294.8. 76Hesychios s.v. epatca;Lykoph. 5.1176-1180 and scholion ad loc.;Polyainos 8.43. 77L. R. Farnell, The Cultsof theGreekStates II (Oxford 1896) 504-505; cf.supran. 22-23. 78 L. Robert, Hellenica 11-12 (1960) 588-595. Robert has collected severalrepresentations of Enodia to which we might add the red-figured kylix inVienna, museumno. 204; CVA,Wien, KunsthistorischesMuseum, fasc. 1,pl. 30.4; and the gem stone of a goldfinger ring;A. Furtwangler, Die antikenGemmen (Leipzig and Berlin 1900) pl. 25.21.79 IG IV 1191, 1192, and 1542.80 IG XII9 1193.

    81CIG 1837; cf. J. and L. Robert, REG 66 (1953) 147, no. 118. It is interesting to note that Jason sacrificed to Artemis in a temple on an island in this same area;Ap. Rhod., Argon.4.470.

  • 8/7/2019 Miller 1974

    27/31

    252 Stephen G. Millertells us of her cult inAthens, Sicyon, and Argos.82 Both theArgives andthe Sicyonians claimed that their xoanaof the goddess had been broughtfrom Pherai. Itmight be noted in passing that these xoanawere ofwoodand, on the basis ofwhat has been seen of aniconism inThessaly, itmaybe conjectured that they were less than fully anthropomorphized. Bethat as itmay, Enodia was clearly a goddess in her own right, existinginmany places side by side with Artemis, but frequently conflated andidentified with her as, for example, on a small altar or bomiskos fromDemetrias (pl. 4:4).83We do not know why Enodia was so especially revered atPherai, but her name provides a clue. Pherai is situated at a cross-roadsand controls the passage to Pagasai-Demetrias, the only viable harboravailable to the otherwise land-locked Thessalians. Control of thisaccess route was crucial to Pherai,84 and Enodia, the goddess of theroad, would have been a natural benefactor to the Pheraians. If onefurther considers the previously mentioned importance of Enodia toThessalian magic, it isnot difficult tounderstand why shewould replaceArtemis among theTwelve Gods at Pherai.

    The appearance of Themis on the altar ismore difficult toexplain, and the argument here must also consider Hera, the goddesswhose place Themis takes. If themonument is, as suggested, the femalepart of a sanctuary of the Twelve Gods, thenHera must be missing. Atfirst thismay seem incredible, but with closer examination of theThessalian context it is not so unlikely, for there is evidence that Hera wasnot welcome inThessaly. For example, in the hundreds of dedicationsknown from Thessaly, Hera ismentioned only twice.Her first appearance is in an inscription of the first century after Christ which mentionsthe priestess of Livia Hera Sebaste.85 This is obviously a reference to acult of the deified Livia where she is equated with Juno, which is translated intoGreek asHera as a matter of course. The other epigraphicalmention of Hera in Thessaly86 couples her with Zeus in the canonicalfashion, but also belongs to the Roman period. Furthermore, Hera isnot represented on any Thessalian coins, nor is there any literary evidence for a cult of Hera inThessaly.8782Paus. 2.10.7 and 2.23.5.83SEG III 485; Halmyros Museum no. 210.84 Cf. Westlake, op. cit. (supra n. 14) 10-11.85 IG IX2 333.

    86 'ApX. 'Ep. 1931, 177, no. 13.87Without being in a position to understand the significance, scholars have

  • 8/7/2019 Miller 1974

    28/31

    Altar of the Six Goddesses 253The documentation for the presence of Themis in Thes

    saly, while not overwhelming, is impressivewhen compared to that ofHera. Themis isattested epigraphically four times in the period rangingfrom the sixth to the third centuries B.C.88Moreover, one of the Thessalianmonths, Themistios,89 was named for her, a fact which obviouslygives her a high rank on the religious calendar. In this context itmay benoted that Demetrias, the early Hellenistic settlement of DemetriosPoliorketes on the bay of Volos, named its twelve months after theTwelve Gods, apparently along Platonic guidelines.90 Of the twelvemonth names atDemetrias, ten have been recovered and they representten of the canonical Twelve Gods.91 Itmay be fortuitous, but one of thetwomissing Olympians isHera. Future epigraphic discoveries may, ofcourse, reveal Hera among these month names, but if the local tradition inThessaly was adequately strong, Themis may rather be found asone of the twelve-month deities atDemetrias.

    One other body of evidence lies in the personal names ofThessalians. There are attested some 14 examples of names which arederived from Themis, such as Pasithemis, Themistion, Themison,Themistokles, and, especially popular, Themistogenes.92 Most of theseareHellenistic, the latest areAugustan in date. There are 11 examplesof Hera-derived names in Thessaly, such as Herais, Herophilos, andHerodotos.93 Of these, two are late Roman (third-fourth centuriesA.D.) and three are slaves.This leaves a proportion of Themis names toHera names of more than 2: 1.The proportion of Themis toHera namesinAttica is, for example, about 1: 1.94So far we have adduced only arguments ex silentio, butpreviously noted the sparcity of evidence for a cult of Hera in Thessaly; Farnell, op. cit.(supran. 53): " [The cult ofHera does not] appear tohave had such vogue inThessaly andalong the northern shores as it had inBoeotia, Euboea, Attica, Sicyon, Corinth and the Peloponnese." S. Eitrem, RE s.v.Hera 370: "Thessalien. Die Spuren eines Kultus sind sehrdurftig." P. Philippson, ThessalischeMythologie (Zurich 1944) 69: "Hera hat in Thessalienkaum einen bezeugten Kult." M. P. Nilsson, op. cit. (supran. 60): "Der Kult der Hera....[ist] sehr sparlich inThessalien, Attika, Phokis undAchaia."88IG IX2 1236 (Phalanna); IpaKrLKd1908, 171 (Thebes); RevPhil 35(1911) 300-301, no. 49 (Magnesia); SGDI 1557 (Mondaia).89 E.g. IG IX2 274 and 277.90P1.Leg. 745B-E, 828B-C, 848D.91F. Stahlin, E.Mayer, A. Heidner, Pagasai undDemetrias (Berlin 1934) 186.92E.g. IG IX2 104; 109a; 219; 415; 530-532; 539; 568; 1232.93E.g. IG IX2 18; 526; 538; 590; 824.94This proportion is based on the names which appear inKirchner, PA,

    which are of sufficient quantity to yield an accurate general ratio.

  • 8/7/2019 Miller 1974

    29/31

    254 Stephen G. Millerthese arguments are highly suggestive, if not conclusive. They areadequate, I would submit, to formulate the hypothesis that Hera wasunpopular in Thessaly, and that Themis was her local equivalent. Ifone accepts this as a working hypothesis, certain literary passages become explicable. Most interesting among these is the peculiar situationof Hera in theArgonautica. In this story,Hera is cast as the protectressof theArgonauts and as the patron saint ofJason, but there are anomalieswhich have escaped the notice, so far as I can discover, of commentators. In the first two books of the epic Hera appears only six times;she plays an active role in the story only upon two of these occasions,and in one of them she is actually responsible for an obstacle in thepath of theArgonauts.95 It isAthena, not Hera, who is theprotectress inthese two books. Athena lays the plans for the Argo and aids in itsconstruction; Athena presentsJason with his wonderful robe, theequivalent of the Shield of Achilles in the Iliad; and Athena pushes apart theSymplegades and helps the Argo pass through.96Hera's active role,which completely supplants that of Athena, begins only in Book 3together with the entrance of Medea. The Argonauts have made thepassage to Colchis, and are within sight of the Golden Fleece beforeHera steps into the action. Her reason for doing so, her avowed purposein supporting Jason, is to see that Medea returns to Thessaly in order tobring vengence upon Pelias, Jason's half-uncle, who does not pay herhonors. This motif is stated several times; Pelias does not worshipHera, and her interest inJason isbased solely upon a self-serving desirefor revenge on Pelias.97This same theme is taken up by [Apollodoros]98 whorelates that Pelias and his brother Neleus, finding that their motherTyro was being mistreated by her step-mother Sidero, chased Sidero,who took refuge in a temenos ofHera. Pelias then slew Sidero upon thealtar in this temenos and subsequently refused to pay honor toHera.This sounds very much like an aetiological tale designed to explain the

    95Ap. Rhod., Argon. 1.997; cf. 2.216, 2.865. At 2.216 it is stated that theArgo isunder the special protection ofHera. This isobviously a foreshadowing sinceHera hasdone nothing to thispoint to deserve the title ofProtectress of theArgo.96Ibid. 1.110; 1.721-768; 2.598-599.97Ibid. 3.64-65; 74-75; 1134-1136; 4.241-243. Although the motif of therevenge of Hera on Pelias is not exploited until the second half of the epic, the stage is setearly, for at 1.14 Jason enters to find Pelias sacrificing to Poseidon and all the gods exceptPelasgian Hera. 98 [Apollodoros], Bibl. 1.9.8.; cf. 1.9.16.

  • 8/7/2019 Miller 1974

    30/31

    Altar of the Six Goddesses 255lack of popularity of Hera in Thessaly aswell as her antipathy towardPelias. While it is not very likely that the reason for this antipathy wasthat given by [Apollodoros], the ancestry of Pelias vis-a-vis that ofJasonmay provide a clue. Jason and his ancestors traced their lineage directlyback toAiolos, the eponymous hero of theAiolic Thessalians. There isno reason to suppose thatHera was unpopular with theAiolians, andthere is evidence of a cult of Hera Aioleia on Lesbos.99 Pelias, on theother hand, although a great-grandson of Aiolos on his mother's side,was sired by Poseidon in the form of the Enipeos river, the large centralThessalian tributary of the Penios river.100We may suggest that Peliasrepresents an indigenous culture in which Themis played the role ofHera. This culture may have been represented in historical times bythe Penestai, a group of lower-class serfs not unlike the Helots of Laconia; the pure Aiolians, such as Jason, would have formed the overlord baronial class.101The indigenous culture may have had sufficientinfluence toprevent the replacement ofThemis byHera.There is, moreover, good evidence that Hera was notoriginally the wife of Zeus, and that she replaced Themis as his consortin the rest of Greece. Hesiod says that Themis was an earlier wife ofZeus,102 and Pindar calls Themis the original wife of Zeus.103 If true,and if Themis remained enthroned next to Zeus in the ThessalianPantheon, then Thessaly represents, forwhatever reasons, a holdoverfrom earlier times. Themis is a member of the Thessalian version of theTwelve Gods and a legitimate holder of her position on the altar of thesix goddesses at Pherai.

    VII. CONCLUSIONIf the hypothesis is accepted that Themis does hold the

    position in Thessaly more usually held by Hera in other parts ofGreece, then the interpretation of our monument as a part of a sanctuaryof theTwelve Gods isconsiderably strengthened. That thisdodekatheonhad an element of aniconism can also be regarded as reasonably well99Ch. Picard, "La Triade Zeus-Hera-Dionysos dans l'OrientHellenique,"

    BCH 70 (1946) 456.100 [Apollodoros] loc. cit.101F.Miltner, RE s.v.Penesten; cf.Westlake, op. cit. (supran. 14) 27-28.102Hes., Theog. 901.103 Pindar, frag. 30 (Snell) 5: &pXalov aMoXov Jdo'. If correct, the theorythat Hera was unable to supplant Themis as the wife of Zeus in Thessaly would support theopinions of A. B. Cook, Zeus III (Cambridge 1940) 1065: "the case forHera as essentiallyand ab origine the bride of Zeus is neither proven nor probable."

  • 8/7/2019 Miller 1974

    31/31

    256 Stephen G. Millerestablished, but the questions of the precise form and, to a lesserdegree,the chronology of themonument must be left open. The solutions tothese questions which have been offered in this paper are tentative andcan be proven or disproven only by future excavations at the site of thealtar of the sixgoddesses on theAcropolis of Pherai.

    University of CaliforniaBerkeleyPostscript: On p. 250 an intermediate stage in anthropomorphicdevelopment was discussed. I have recently found reference to anapparent equivalent intermediate stage of the evolution inRome. Themanifestation was, however, the opposite of that posited on p. 250withthe aniconic head of theMagna Mater set into an anthropomorphicbody; cf.Am. Adv.Nat. 7.49.