American Journal of Ancient History [7.2] 9781463206758, 1463206755

Citation preview

American Journal of Ancient History

American Journal of Ancient History

7.2

The American Journal of Ancient History is a peer-reviewed academic journal covering ancient history and classical studies. It was established in 1976 and edited by Ernst Badian until 2001. It is continued by the American Journal of Ancient History: New Series, edited by T. Corey Brennan.

American Journal of Ancient History

Volume 7.2 Edited by

Ernst Badian

gp 2017

Gorgias Press LLC, 954 River Road, Piscataway, NJ, 08854, USA www.gorgiaspress.com Copyright © 2017 by Gorgias Press LLC Originally published in 1982 All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of Gorgias Press LLC. ‫ܐ‬

1

2017

ISBN 978-1-4632-0675-8

Printed in the United States of America

TABLE

OF CONTENTS

EdwardChamplin:The Suburbiumof Rome ........................................

97

RobertA. Moysey:Isokrates'On the Peace:RhetoricalExercise or Political Advice?

..............................................................

118

N.G.L. HammondandM.B. Hatzopoulos: The Via Egnatia in Western Macedonia I .........................................................

ArthurKeaveney:SullaAugur .......................................................

128

150

M.H. Hansen:Demographic Reflectionson the Numberof Athenian Citizens 451-309

•c

.............................................................

172

TABLE

OF CONTENTS

EdwardChamplin:The Suburbiumof Rome ........................................

97

RobertA. Moysey:Isokrates'On the Peace:RhetoricalExercise or Political Advice?

..............................................................

118

N.G.L. HammondandM.B. Hatzopoulos: The Via Egnatia in Western Macedonia I .........................................................

ArthurKeaveney:SullaAugur .......................................................

128

150

M.H. Hansen:Demographic Reflectionson the Numberof Athenian Citizens 451-309

•c

.............................................................

172

THE

SUBURBIUM

OF ROME

The classicstudyof the historicaldevelopment of a modemsuburbhas observedthat it "is clearly lessof a geographicalexpressionthan it is a

stateof mind".•Ancientterminology rendersthisparticularlyvivid for Rome:thenounsuburblure is practically non-existent in literaryLatin.2 What we do find are the occasional abstraction, suburbanitas, and the

commonsubstantiveuse of the adjectivereferring above all to landed property,as in suburbana(villa understood)and suburbanum(praedium) or, much less often, suburbanus(ager). That is to say, the "suburb"of Rome is not an areabut a conceptwith a very narrowand privatesignificance.The concept,the attitudeof mind, is of coursethatof a literateand very urban elite, and at first sight it has little to do with the economic bondsbetweencity andterritory,or with the variousreligiousassociations of the land adjacentto the city. The purposeof this paperis limited, then, to a surveyof what this conceptmeantto the elite of Rome, rather than

to ananalysis of itshistorical reality.Thetwoneednotcoincide.3 First, the problemof definition.On the face of it, a suburbis simply somethingadjacentto the urbs;however, beyondthis it is clearly quite dependenton the city, it performssomeof the city's functions,andit may be seen(and was seenin antiquity)as in someway a part of the city. On the otherhand, the suburbequallyclearly, at someindeterminateplace, borderson or runsinto true country,it sharesmostof the characteristics of the countryside,and to someRomansit was merelya specialkind of countryside,russuburbanurn. It was alwaysat Romea placeof ambiguity, a borderregion, and its physicallimits (to beginwith the mostobvious) defy precisedefinition.In the caseof Rome, thereare too manypossible boundaries for the city itself, someof themcapableof shiftingovertime: the sacredboundaryof the pomerium;the Servianor the Aurelianwalls; theouterbordersof thefourteenregions;a veryobscurecustoms boundary; the first milestoneon eachroad;the built-uparea,aedificiacontinentia-eachboresometerminalsignificance.No wonderthenthatin theAugustan peaceDionysiusof Halicarnassus couldobservewith awe thatno oneknew how largethe endlesscity was, or how far it extended,or whenit stopped

beingcity/At theouterextreme, theonlysureboundary between suburbium and country,that is the juridically significantone-hundredth mile97

¸ 1985 by E. Badian.All rightsreserved.

98

EDWARD

CHAMPLIN

stone, obviouslyincludeda vast area which no one seriouslyconsidered to be suburban?

Where geographyfails, attitudecan help. A very crudedefinitioncan be reachedby simply chartingon a map the use of the word "suburban" by ancient authors(irrespectiveof date), for the towns and territories directlyor indirectlycalledsuburbanto Romeform a regularandcohesive region:SaxaRubra,Fidenae,Nomentum,Ficulea,Tibur, Gabii, Praeneste, Tusculum, Bovillae, Aricia, Velitrae, Lanuvium, and Lavinium; with the

AgerPomptinus andAntiumto the South. 6 (Directreference to anyof thesetownsas suburbanin themselvesis very rare. Far morecommonis the citation of someone'ssuburbanpropertyat or near such-and-such a town, a distinction which is of course in itself a useful indication of

attitude.)Practicallyspeaking,the areathusdefinedis simplythat of the modemRomanCampagna,the greatundulatingplainboundedon oneside by the Tyrrheniansea and on the otherby a semi-circleof pre-Apennine ranges(Monti della Tolfa, Monti Sabatini,Monti Sabini,Monti Lepini). The crestsof theserangesprovideeffectiveouterlimits to the suburbium and its choicer areas, althoughriver valleys, the barrier posed by the volcanicclumpof the Alban Hills, and the coastalplain extendingas far as Antium, make for an irregularand uncertainboundary.Despitethese irregularities,an appropriateborderis implicit in the geographerStrabo's descriptionof Tusculumaslying on a ridge"adornedby the plantingsand villas encirclingit andparticularly by thosethat extendbelow the city in the general direction of Rome"; the same, he adds, holds true for the foothills of the Alban Mount. 7

The crucialdeterminantof the regionis of coursetime: how quickly couldonegetto andfromthecity?"Urbisvicinaiuvantfacilesque recessus." ThusColumellarecommended suburbanpropertyin general,andPliny his maritime villa in particular,for the easeof gettingthere after a busy day in the city; vicinitasurbis was the first requirementin the searchfor a suburban propertyfor Suetonius; it wasfelt worthyof remarkthatAugustus

tooknotonedaybuttwoforthejourneyto Praeneste or Tibur.8In short, a comfortableday's journey providesa rough limit, with the natureof terrainand roadsmakingfor greatfluidity: very broadlyspeaking,a 20to 25-mile radius from Rome. 9

A roughinner boundarycan alsobe established. Hortus in Latin is a garden;"horti"however,in theareaof Rome, signifiesnotmerelygardens but an estatenearthe city. Here one thinksimmediatelyof the greenbelt closelyhedgingthe urbancentre,the famouspleasure-gardens of Caesar, Clodia, Lucullus,Sallust,Maecenas,andmanylesserfigures.It hasbeen suggested that"horti"impliesproximityto thecity, andin fact mostknown

THE SUBURBIUM

OF ROME

99

horti in the region of Rome lie immediatelynext to the city, that is, on the right bank of the Tiber, in the CampusMartius, on the Pincian,on the

Esquiline,andon the Aventine. •ø One couldof coursecite the smaller funeral gardensby the side of the great roadsand byroads,or stray uses of the word elsewhere, as in Cicero's reference (Ad Att. 9.9.4) to some hortuli at (suburban)Lanuvium, but to a Roman the word would first raise an imageof the urbangreenbelt. Now the words"horti" and"suburbanurn"

canandnaturallydo overlap:thesegardens weresub-urban. • However, thetwo ideasareoftenclearlydistinguished in themindsof ancientauthors, and mostsignificantlywe have severalimportantpassages which showa

clearmentalsequence suburb•horti•ity (orthereverse). •2Otherconnotationsmay have attachedto "horti", suchas relative smallness,perhaps absenceof a villa, lack of seriousagriculture,or a greateremphasison pleasureand relaxation:the evidenceis unclearand there may have been no consensus amongtheRomansto differentiatehortifromsuburban property in general. Yet it would have been generallyagreedthat the horti

werestrictlysuburban but closerto thecity thanthe truesuburbanurn. •3 In short, they were the inner ring of the suburbium. Within theseroughgeographical limits, the"suburb"remainsverymuch an idea, more a matter of shared attitudes than of location. At the outset

we mustabandon conceptions of modemsuburban li'fe.Rapidpublictransit and tract housingare productsof the nineteenthand twentiethcenturies, along with the notion that large numbersof people--and not just the rich•ould live in oneplaceandeachday work in anothersomedistance away. With the expansionof the city modemsuburbsareto a muchgreater extentthanbeforemerelydecentralized partsof thecity. For thisphenomenon the Victorianscoinedthe sinisterand for ancientpurposesinvalid word "suburbia",allegedlythepluralof suburbium,anda word whichhas

beensaidto suggest a pictureof an "expressionless half-urban steppe". 14 At the otherextreme,and equallyfar removedfrom the Romanconcept of the suburb,is thepictureof themediaevalsuburb,the settlement outside the town walls of peoplewho stoodfigurativelyas well as literally on the marginof society,the homeof inns, amusements, industries,foreigners, riots and disease.The closestparallels to Rome are imagesthat fall chronologicallybetweenthe two extremes,in the early moderuperiod, particularlyRome from the sixteenthcenturyonward,with its customof villeggiaturaon the very sitesof the ancientsuburb,and Londonof the later seventeenth to early nineteenthcenturies,as it slowly expandedto the northand the west. As with thesein general,so it will appearwith classical Rome: the idea of suburbanlife is above all an elitist one, and

suburbanitas restsfirmly on a foundation of wealthandleisure. •5

100

EDWARD

CHAMPLIN

The averageRomanprobablyheld a fairly consistent view of life in the suburb,or at leastif he had no personalknowledgehis literaturecertainly presentedhim with a consistentimage, from the secondcentury•3c, when suburbanlife as suchbeginsto emerge, up to the fifth centuryAD, when it is violently disrupted,thoughby no meansdestroyed.Three separate ideas can be traced through the literatureof these six centuries,ideas expressedagain and again in the samethree words. First, and the most commonlymentioned,salubritas.Onewentto one'ssuburbanurn to restore one's health or (even more often), when one was healthy, to avoid the dangersof the city. Regularlyof courseone retreatedfor the summer,the aestivus secessus. For coolness' sake, the seashore and the hillside were

obviouslypreferredbeforeall, but anywherewoulddo:the suburbancoun-

trysidewassimplymorehealthythanthecrowded andunsanitary city.16 Second,the suburboffered otium, a place to think, to read, to converse, to write and to relax from the pressuresof the city; while, on the other hand,it mightafforda comfortablealternativeto thedutiesanddeprivations of true country life, such as the complaintsof tenantsor the lack of

cultivated neighbors. •?Thesuburb offeredotiumbutnotnecessarily quies, and literary societycouldflourishthere. As Symmachusphrasedit nicely,

"I amin thecountry butI amnotrusticating". •8Andthird,thesuburb had amoenitas,beautythat delights,elegancewith scenery.The amoenitasof a suburbanvilla is a subjectmarkedlycongenialto epistolographers, from Ciceroto Gregorythe Great, the classicdescription,detailedand loving,

beingprovidedby the younger Plinyfor hisLaurenfine villa.•9Indeed, one letterwriter couldnot resistdescribingthe Christianparadisein terms of the terrestrialsuburbas "amoenissimum praedium,et urbanisconvenustatum et rusticis consitum". 2ø

To statetheobvious,whatweavesthesethreestrandstogether,andwhat distinguishes suburbanamoenitas,otium and salubritasfrom any other variety, is simpleproximityto the city: the qualitiesof the suburb,close yet separate,are definedandusuallyenhancedby thoseof the city. Two of thesequalitiesstandout. First, the suburbaboveall signifiesprivacy in a time when even the well-off in Rome might lack opportunitiesfor salubritas,otium and amoenitas.City houseswere relatively small and crowdedtogether,eventhe upperclassesmight only rent quarters(some-

timesmodest)withinthe confines of the city.2• The residence evenof Augustuswas deemedtoo small. Accordingto Suetonius,wheneverthe emperorwishedto be aloneanduninterrupted he wasforcedto retireeither to an attic in his houseor to the suburbanpropertiesof oneor otherof his freedmen,while for longerperiodshe would often make for one of the

(suburban) townsnearest to thecity,Lanuvium, Praeneste orTibur.22The

THE SUBURBIUM

OF ROME

101

suburbis depictedrepeatedlyas the place of privacy, offering private retreat,refuge,mourning,celebration,exile, evenimprisonment: in a word, it filled a very urbanneed. As Martial put it, "You ask me, Linus, what profit does my land at Nomentumreturn to me.9 This it returnsto me,

Linus,thatI don'tseeyou."23A secondimplication of the suburban combination of salubritas, otium and amoenitas, and one connectedwith

the senseof privacy, is a senseof impermanence and detachment(as will emerge).Part of the peculiarcharmof vicinitaslay in the easewith which one could escapeto the suburbfrom the city and in the easeof getting

backintothe city andcity life? Mostof the upper-class inhabitants of the suburb were not natives of the towns in whose territories their villas

lay. They lived in Rome;theycamefrom, dominated,andretiredto regions elsewhere;from the local scenewe may look for detachment. This conceptionof the suburbis clearlyreflectedin its physicaldevelopmentby the Romans.At theheightof theempiretheCampagnawasfilled with villasandtheirgardensandparks,theseinterspersed with thetemples and grovesof variouscults, and it was coveredby a fine net of roadsand byroads,the bordersof which (particularlynearthe city andtowns)were lined with cemeteries,tombsand funerarygardens.This much is clear

fromarchaeology, andthepictureis supported by ancientliterature?In his descriptionof the battleof Cynoscephalae (in Thessaly),Livy embroidersthe accountof Polybiusto explain, vividly but anachronistically, the problemsencountered by the opposingarmies:"a greatimpedimentto the actionon both sides",he comments,"was a battlegroundcoveredwith densetreesandgardens,asin suburbandistricts,andwith itsroadsconfined and in somecasesquite blockedby walls." He is simplydrawingon the

Romeof Augustus.26 Ononeextreme view,thesuburb wasa greatenclosed park, enclosedby hook or by crook. Livy speaksof roadsblockedeven by walls, anda commonlegalproblemapparentlyaroseovertheencroachment by privateownersonto suchhallowedplacesas the grovesof cults and funerary plots. One writer on surveyinggrew quite upsetover the plight of thoseareasreservedin the suburbfor the burial of paupersand the punishmentof criminals, "for from theseplaces, becausethey are suburban,private owners are accustomedto usurp pieceswithout any

reverence for religionandto join themto theirhorti."27 Thustheprivate park threatened evenpublicandsacredproperty. As for privatesmallholders, Straborecordsthevirtualdisappearance by the time of Augustusof severalancientLatin towns near Rome, which hadbecomein hisday eithervillagesor theestatesof privatecitizens,and the declinein numbers of thefreecitizenryof theCampagna seemsto lie behinda commonliteraryperception of Latiumas rus vacuum?To be

102

EDWARD

CHAMPLIN

sure, many an independent municipality like Praeneste or Antium flourishes,andmany a village or hamletcoalesces alongthe greatconsular

roads, 29butthecontrast in literature between thepopulous daysof yore and the calm of the enclosedCampagnais a theme so pervasivethat we mustaccepta countrysideperceived(at least)as largely in the handsof powerful privati, who entrustedthe cultivationof their estatesto tenant farmersandhiredlabourersor slaves.Thuswe havethepardonableexaggerationsof Cicero, claimingin a famouspassagethat the land of Praeneste is held by a few, or of Seneca,quotingas the measureof a truly rich man that he ownsa suburbanestatethe size of which would arouseenvy even in the wilds of Apulia, or of Martial, imaginingthe great landlordsof Tibur and Praenestelookingdown with amusement uponhis friend'sfew acres(!) on the Janiculum.Yet on the AppianWay, no furtherthanthree miles from Rome and at the heightof the AntonineAge, the estateof the Athenian consuland millionaire, HerodesAtticus, encompassed its own

lati fundi, fields,vineyards, olives,orchards andpastures. 3øOutsideof literaturesuch huge suburbanestatesare an exception,but it must be emphasizedthat within the Roman empire the suburbiumof Rome is unmatchedfor the density of its villae rusticae. Thus a survey still in progresshas identifiedthe remainsof no fewer than 133 villas in the territory of Tibur (with another34 in the upper and middle Anio valley) and 153 in that of Praeneste alone. 3•

It shouldfollow that suburbanland was not cheap, indeedthat it was as expensiveas any in the empire.Unfortunatelythe survivingfiguresfor land valuesare so few and so disparatein date, location,and natureof propertyconcerned,that they tell us little beyondthe obvious,that it was expensiverainfact thereis only one ancientpassagegiving both the size

andthevalueof a suburban property. 32Naturallydistinctions weredrawn, someareaswere more desirablethanothers.One couldbe on the margin: Catullus' friends,for instance,would say that his suburbanavilla was at Tibur, but otherslessfriendlymightcall it Sabine.A Veii simplydid not have the cachetof a Tibur: summingup the evidence,a modem survey of the remains in the area concludedthat the owners there in the early principatewereof freedmanstock,"well-to-domiddleclassbusyacquiring propertyin what must have been at that time one of the relatively few areasnearRome whereland was still easilyavailable",and indeedalmost

no senators areattested in thearea.33Nevertheless, thelogicof vicinitas urbisdemandsgenerallyhighlandprices.As nowhereelsein the empire, Rome attractedwealth and men of wealth, and the political elite of that empirerequiredmorethana plainandperhapsrentedroof overtheirheads when its memberstook up residenceat the capital. In a florid passage

THE

SUBURBIUM

OF ROME

103

directedto just sucha readership,the agronomistColumella extolled the virtues of suburban land over all others: "Now since civil ambition often

callsmanyof us to Romeand havingcalledus moreoftenthannot keeps us there, it follows that suburbanpropertyis the most advantageous,so that a man busiedwith daily affairscan easilyget away after the business of the foram." Significantly, when Trajan insistedearly in the second centurythatcandidates for senatorialofficehaveonethirdof theirproperty in Italian land, Italian land pricesshotup--especially thoseof suburban

properties.34 A manprepared to buy a villa at Tusculumor Tiburfor the sakeof salubritasand aestivussecessus did not stopto calculatewhen his

investment Whatthen would of the pay suburban foritself: property sosaid market? themillionaire CiceroSeneca, offerssome whonoteworknew.35 thy anecdotes.His own belovedTusculanum he hadboughtaround68 BC from the later notoriousL. Vettius, who had acquiredit from the dictator Sulla, who in turn acquiredit from Q. LutatiusCatulus(cos. 102), that

is, fourownersin notmuchmorethantwentyyears. 36Similarly,he discussesanotherTusculanvilla, thatownedby Caesar'sagentandconfidant, Cornelius

Balbus: in addition to Balbus it had known at least four other

ownersin aboutfifty years, two of them Romanconsuls,one of obscure

background, andonea freedman. 37Andwhenin 45 Cicerowassearching for land near Rome on which to build a shrineto his daughterTullia, he consideredno fewer than nine horti, all of them owned by senators,all

eitheron the marketor mmouredto be available. 38Cicero,from whom so much of our evidence is derived, lived in a time of turmoil, it is true,

but there is no reasonto assumesignificantlyless turnoverin later and quieterperiods.Two generalpropositions shouldbe allowed.First, assuming a relatively low rate of successionwithin the ruling elite over the generations,and assumingthat nowhereelse in the empirewas theresuch a concentration of landedpropertyownedby the elite, the rate of turnover

in thesuburbshouldbe thehighestanywhere. 39Second,thereis a sharp distinctionto be drawnbetweenurbanand suburban propertieson oneside and rural propertieson the other. The suburbof Rome was simply not home for many. Deep and true sentimentis reservedfor the home town and the ancestralacres,witnessCiceroon Arpinumor Pliny on Comum,

andtherearegoodmodemparallels forthis.4øWhenTrajanforcedcandidatesto investin Italian land and suburbanpricesshotup, Pliny observed that buyersactivelysolicitingproperty,and offeringhigherpricesfor it, had the effect of bringingevenmore land ontothe market.This doesnot arguefor great sentimentality,family tradition,or local ties.

Suburban landisthenhighlydesirable, expensive, andeasilytransferred, asfar astheeliteis concerned. Its naturewill inevitablyaffectthatof the

104

EDWARD CHAMPLIN

societyon it. First and most striking is the fact that, on the evidence available,thesuburban landownerplayedhardlyanylocalrole. The various townsand citiesof the Campagna,someof which, particularlyTibur and Praenestewith their greattemples,were both rich and populous,played

littlepartin hislife.4• Thereare,to be sure,important exceptions. First, roughlyspeaking,the lower one'srank in the upperstratumof society, themoreonedevelopedor inheritedmunicipalcontacts: wherevilla society doesoverlapwith town life in patronageor administrationit tendsto do so in the personsof the upwardlymobile, transplanted from Rome or

elsewhere, or risingamongthenatives. 42Second,thelocalmagnates who manageto rise into the governingelite at Rome maintain,as one would expect,their ties with the patria. Thus, in the late Republicand early Principate,the Fonteiiof Tusculum,the Rubellii of Tibur, or the Acilii andEgrilii of Ostiacontinueto supplypatrons,benefactors andevenlocal magistrates. Yet, thatsaid,theimpression remainsoneof unusualindifference.Despitetheir obviouspresencethere,the suburbof Romeis onearea where(beforethe late Empire, at least)Romanaristocrats seldomassu•'rne

thecustomary functions of localgrandees. 43Thisrelativesilence fitswell with the generalliterary record, that is, Cicero, Seneca,Pliny and Symmachus, who tell us so much about suburban life in their letters, have

almostnothingto say aboutthe life of the townswithin whoseterritories their villas lay. And excludingthosewhosefamilies actuallycamefrom thosetowns,thereseemsto beonlyoneinstanceof theRomaneliteactually possessing a domusin a suburbantown, as distinctfrom a villa in its

territory. 44One'stownhousewasof course in Romeif onehadone,and this very idea, that a propertyin the territoryof (say) Tibur could be suburbanto Rome, ratherdiminishesthe significanceof Tibur itself. An explanation is againto be soughtin thesenseof impermanence anddetachment.In hishomeprovinceor regiontheRomangentlemanwasenmeshed in a fine net of obligationsto friendsand clients,while if he were to cut any figure at Rome the demandsof friendshipand patronagewould be evenmoreconstricting. The suburbwashoweverneithercity norcountry, its ethoswas avowedlyone of leisureand privacy, hencea conspicuous lack of local ties.

Secondly,wherethe suburbanlandowners had little to do with local townsas such,they had a greatdeal to do with eachother. The suburb wastheplaceto enjoythepleasures of truefriendship,a friendship notto be confusedwith urbanarnicitia.At the beginningof Cicero'sDe Oratore (1.24) we meet the great oratorof his age, L. Crassus,relaxingin his Tusculanurn duringthe holidaysin the late summerof 91 Be; therehe is joinedby his belovedfather-in-law,thejurist Q. MuciusScaevola,and his greatfriendand rival M. Antonius,and by othersenators and names

THE SUBURBIUM

OF ROME

105

of greatweightat that time. In the courseof two daysthey talk aboutthe stateof politicsand the stateof oratory, they dine and amble aboutthe property,andthey arejoinedby yet moredistinguished friendsandneighbouts. Fictitiousthoughthe encounteris, the settingmust respondto a reality recognizableby and agreeableto Cicero's audience:he certainly recordssimilar real sojournsin his correspondence, and we can follow

dozens of suchvisitsamong thegenteel latter-day colonizers ofTusculum. 45 In this regard, the private amoenitasof one's suburbanretreat, the art gallery, the library, the garden,has a particularsocial face as well; that is, they are meant not solely for one's own pleasure,they are meant to arousethe pleasureandadmirationof a choicecircleof visitorsandneighbours.Symmachusgivesthe classicdefinitionof this society:"iucundum otium cum familiaribus

nostris in suburbano."

And that the suburb was

the place for friendly societyis confirmeddelightfullyby the younger Pliny, who pretendedto prefer his Tuscanvilla to a suburbanestateat Tibur, PraenesteandTusculumpreciselybecauseit offeredgreaterotium,

thatis,because people werenotconstantly dropping byfromthenextvilla.46 A third distinctiveelementin suburbansociallife is the presence,from the age of Augustuson, of the Imperial court. At sometime or other in the first three centuries,sometimesover the entire period, we know of Imperial residencesmaintainednear Tibur, Praeneste,Tusculum, Alba, Aricia, Velitrae, Lanuvium and Antium, and over in Etruria at Saxa Rubra, Forum Clodii, Lorium, Alsium and Centumcellae. Closer to town there are the great mansionsof the first few miles of the roads out of Rome,

andseveralof themostimportanthorti, includingthoseof Lucullus,Sallust,

and Maecenas. 47This pervasivesuburban presence, comparable to the modem so-called"courtsuburb"and naturalto the largestfortunein the empire, must have affectedthe surroundingsocietyand economy,most notablyin thoseareasbelovedover the yearsby certainemperors,Domi-

tian'sAlba or Hadrian'sTibur.48One instance is suggestive, Hadrian's greatestablishment at Tibur. Initially an attractionfor that emperorwas doubtlessthat the regionalreadygloriedin a colonyof Spanisharistocrats like himself beforehis accession;indeedhe may have been one of them, and his imperial presencethere later presumablyattractedothers. Not surprisingly,in andshortlyafterHadrian's reignno fewerthaneightRoman magnates,all of them consulsand at least four of them Spaniards,held either a local magistracyor a priesthoodat the great templeof Hercules

Victor.49Thisstartling burstof municipal patriotism, unprecedented and largely unparalleled,would have considerablerepercussions in the life of the town: it was doubtlesstouchedoff by the Imperial villa at the foot of the hill.

Finally,andnotunconnected withthis,theremaybe a relationship

106

EDWARD

CHAMPLIN

betweensuburbanlife and socialmobility, in effect mobilityupwardinto the elite. One thinks of the paradigmaticmerchantor professionalman who makeshis fortune, sinksmuch of it into a countryplace, perhaps continuesto commuteto theofficeor to overseehisbusiness,andeventually retiresto the life of a countrygentleman:a modem phenomenonclosely reflectingCicero'sclassicstatement:"Mercatura... si satiataquaestuvel contentapotius, ut saepeex alto in portum, ex ipso portu se in agros

possessionesque contulit.... -5o Or compareMartial'sedifyingtale (12.72) of the formerlylandlesslawyerwho retiresfrom the city andhis small-timepractice,purchasinga tiny estatehiddenout amongthe tombs, and who is now in dangerof starvingbecausesuburbanpropertydoesnot

producemuchincome? But for thosewho couldaffordit a suburban property,lying convenientlynearto boththecapitalcity andthe seashore, wasideal,andinscriptions amplyconfirmtheconnection betweengeographical and social mobility. Moreover, we know about the previousowners of Cicero'sand Balbus' Tusculanvillas preciselybecausecertainneighbourshad allegedlyresentedsuchupstartsbuyingup propertieswhich had belongedto Romanaristocrats.Cicero'stelling reply was to showthat he, the equestrian-born lawyer from Arpinum, and Balbus, the Spanishmillionaire, had been precededin their titles by a mere countrygentleman

fromPicenum andby a freedman. 52Onemightnotein addition thatif the risingnativeof a city like Romeor Ostia-•or evena boatmanon the Tiber (Martial 10.85)--tended to buy land near the placeshe knew, the suburb of the largestcity in theempiremusthaveaffordedoneof the mostheavily travelledroutesof upwardmobility. What then did suburbansocietythusdefineddo on or with its property and, more particularly,what pursuitsseem speciallyto flourish in the suburbanclimate or to acquirea distinctivelysuburbanflavour? While at his Tusculanvilla one day, Cicerofoundthat he neededto use certainbooks,includingsomecommentarieson Aristotle, which he knew to be in the library of the youngLucullus(son of the famousgeneral),so he went over to consult them, as was his custom. At Lucullus' villa he

discoveredthe youngerCato surroundedby the works of Stoic authors. Cato informed him that he too was in residence at his Tusculan villa, since

Rome was presentlyin the grip of the games, and he was preparinga courseof study for his ward Lucullus. He and Cicero then fell into a

learnedconversation brokenoff onlyby nightfall,andthatallegedconversation is recordedin the De Finibus (3.7-10, 4.80). Similarly, Cicero's Topica(1.1-6) are developedfrom a discussion with a jurist friend in his library at Tusculum,and the De Divinationeand aptly namedTusculanae Disputationesarise from leisurelystrollswith friendsthroughthe villa's

THE SUBURBIUM

OF ROME

107

grounds.In a word, the suburbwas held to be the ideal settingfor otium

litteratum. 53 Where Cicerospecifiesa settingfor his rhetoricaland philosophicaldiscourses,eight are placed in or take their start from a suburbanencounter,while two more are lessclearbut arguablysuburban, one set at his pleasurevilla at Cumae on the Bay of Naples, one at his

ancestral Arpinum--andnonein the city of Rome.54Not to labourthe point, thesescenesare fictitious,and Ciceropassedmuchof the time of Caesar'sdictatorshipat Tusculum,so that it was much on his mind, but his choiceof settingmustsurelyhave beenappreciatedby his audience. For otiumlitteratum,for the leisureto work at whatwasreally important, that is composition,one did not retire to the territory of one's distant homeland, be it Venusia, Sulmo or Verona: that was rustication. Nor did

one stay in Rome, if one could avoid it: there was too much negotium there. Ratherthe selectfew, the creatorsandmenof culture,soughtsociety that was both leisured and cultivated, and this could really be found in

onlytwoplacesin Italy, in thesuburb of RomeandontheBayof Naples. 5s Here we must rememberthe very social natureof culture at Rome. The act of compositionwas intimatelyboundup with discussionand reading aloud--for this, suburbanvilla life, both relaxed and stimulating, was ideally suited. Two elementsof this life are of considerableinterest;they are mentionedhere briefly as subjectsworth further investigation.First, the "high culture"of the Roman world was obviouslya completelyurban culture, its literature the product of city life and city men. The country and its pursuitsare seen inevitably through urban spectacles.Take for instancethe powerful and pervasiverural idyll, conveyedby, but not confinedto, the pastoral:the countrysideas a place of happy innocence and true friendship, as a temporaryand leisurely retreat, as a pleasant garden(the locus amoenus),and particularlyas the anti-city, where the

vicesof urbanlife are tacitlyor explicitlya topicof discourse?The pastoralideal is neithertruly rural nor truly urban,but an intimatebinding

of thetwo:in short,perhaps, suburban. 57 Secondly,urbancultureinvadedthe adjacentcountrysidephysically,in theform of librariesandartgalleries,thepropersettingfor otiumlitteratum. The suburbwas a greatshowplacefor the connoisseur: the real alternative to the displayof great art in public at Rome was its displayin private in the suburb, at one's magnificentvilla and near one's choicestbooks. If you wantedto seeart treasures,Ciceroclaimed,you hadto go to Tusculum,

andpeoplewentto Lucullus'villasjustto seehispicturegalleries. 58An immensequantity of the art now housedin Roman museumsactually derives from the ancient suburb.59

After culture,a secondsuburbanpastimeof the Romanelite was, not

108

EDWARD

CHAMPLIN

surprisingly, horticulture,whichonemightassumewouldhavea broader appeal.Amusinganecdotes abound,suggesting aninterest oftenamounting to passionandreminiscent of the actionsof the greatpiscinarii.Thusthe oratorHortensiusonceurgedCiceroto changeplaceswith him whenthey werepleadinga casetogether:he hadto get out to his Tusculanvilla to water with wine a plane-treehe had plantedthere. Similarly,a century later, the oratorPassienus Crispusdisplayeda quitebizarrepassionfor a certainbeech-treein a groveof Diana at Tusculum,whichhe was accustomed to kiss and embrace, to lie under, and to water with wine. Seneca

indulgedhis costlylove for viticultureat Nomentum,L. Vitelliuswon famefor importingthe fig andthe pistachioto Alba, A. Gabiniuslooted Cicero'sTusculanvilla notonly of theexile'sart andfurniturebut of the treesaswell.60Particularly tellingisMartial'ssketch of a parvenu imitating the essentialsof aristocraticsuburbanstyle: where the great nobleman

Torquatus(cos.AD 94) had a mansionat the fourthmilestone,Otacilius boughta smallfield; whereTorquatushadmarblebaths,Otaciliushad a bath-tub;andwhereTorquatus laid out a groveof laurel,Otaciliusplanted onehundred chestnuts. 6•The suburbgaveurbanmanhisgarden,andthe

concept is intimately boundupwithhorticulture. 62 The significantelementhereis that suburbanhorticultureis the pastime of menof power.This hasone strikingresult.To readimpressions ancient and modem of the Campagnais to form a pictureof one greatparkland. Gardens,orchards,vineyardsand parks, not to mentionfishponds,game preservesand aviaries,requirea greatdeal of water. It follows that water and water-rightsplay a conspicuous role in suburbanlife. The ancient sourcesemphasizehow well-wateredthe suburbwas, andseveralpropertied

authors showa keeneyefor watersupply. 63In factwaterwasabused. In the early secondcentury BC, Cato the Elder won great unpopularityfor

simplycuttingoff thosewho divertedpublicwaterunlawfullyinto their housesand gardens.In the first centurythe sameproblemwas treatedin

a speech De Aquisby Cicero'swild youngfriend,M. CaeliusRufus. 64 Over a hundredyearslater again, the elder Pliny complainedthat two of the great aqueducts,Aqua Marcia and Aqua Virgo, were quite lost to the publicbecauseof the ambitionand avaricewhich divertedthemto the use of villas and suburbanproperties,and in the next generationFrontinus confirms that public aqueductswere actually brought to a standstillby

privatecitizens justto watertheirhorti.65As themagistrate in chargeof theurbanwatersupplyandauthorof a monographon the subject,Frontinus shouldhave known. He furtherreportsnot only dealingwith suchwrongs daily, but constantlyrepairingthe damageinflictedon the aqueductsby the irnpotentiapossessorurn, the unrestrainedproprietorswho built houses

THE SUBURBIUM

OF ROME

109

and plantedtrees too close to the aqueducts,laid roadsover them, and

refusedaccess for repairs. 66Suchcomplaints tallywellwithotherimpressionsof partsof the Campagnafalling into the handsof the few, and of private landownersencroachingupon public land. For the Romanelite, the cultivationof one's gardenwas not an act of retirement. It follows that public life was in someways attractedfrom the city into the suburb, whatever the private ideal may have been. Again, this is difficult to document.The role of the suburbcould of coursebe quite formal, asthe areawherea governorreluctantto leavethe centreof things might linger before taking up his province(witnessM. Lepiduswhen caughtup in the Ides of March), or whereonereturningfrom his province wouldstop,sometimes for monthson end, waitingor hopingfor a triumph. In the Imperial period the suburbas a thresholdtakeson a much clearer ceremonialaspect,as the area where a formal entranceinto the capital

commences, andwheretheexitends. 67Equallyimportant, thesuburb with its combinationof privacy and proximity to the city is ideal for private transactionsand even intrigue. The prime if rather extremeexampleis offered by Pompeythe Great in the last decadeof the Republic, whether he was operatingfrom one of his horti next to Rome or from his beloved

Albanvilla.68In hissuburban properties he couldquietlytaketheadvice of Cato or intriguewith Crassus,there he could distributelarge bribesto the voters,therea cronycouldengineera massiveprivateloan to the king

of Egypt? Pompey's casewasof course unique,butit is easyto imagine other politiciansand businessmen using suburbanprivacy for the same purposes.

The emperorswere Pompey's successors in this as in so much else, Augustuslikewise retiring to the suburb. In good times and in bad, moreover,muchof the normalbusiness of theempirewasconducted from

theregionaround thecity.TM Intrigues andconspiracies should alsofollow thecourt,butsuchthingsaregenerally hiddenfromusundertheEmpire. 71 One practicedoes however standout, first observablein Tiberius' long and notoriousperegrinatiosuburbanaaroundRome, pointingthe way to mutualdistrustbetweenruler and ruled. The suburbbecomesparticularly importantin time of trouble, a placeout of immediatedanger,from which one couldeithercrushoppositionor take flight. In 68, Nero retreatedfirst to his Servilian horti to plan his escape,and he found refuge eventually in a freedman'ssuburbanumto the north of the city: he went no further, and he was long searchedfor beforehe was discovered.Commodushad the foresightto be in the suburbwhenthe mobroseagainsthis henchman Cleander, and after the conspiracyof Maternushe retired from Rome, either to his suburbanvillas or to more distantestates;similarly, after the

110

EDWARD

CHAMPLIN

fall of Plautianus,SeptimiusSeveruspassedhis dayseither in the suburb or in Campania,and Elagabaluswaited hopefullyin his horti for newsof

thedeathof Severus Alexander. TM Fortheemperors, suburban privacyhad the extra value of safety. The suburbis thusnot only ambiguous,it is a paradox.In theirpursuit of suburbanhealth, suburbanbeauty, and suburbanlearnedleisure, the rulersof Romecannothelpimportingintothesuburbtheirurbanaesthetics, their urbanavocationsand, in the end, their urbanaffairs. Legislatorsand jurists accordinglyseemalmostto acceptthat the suburbwas in fact part

of thecity.73Theirpeersmightbe loathto admitit. Thereare,to besure, otherhistoricalsuburbshardlytouchedon here:the suburbof thetenacious smallholderand the tenantfarmer revealedby surveyand excavation;the suburbof town life in the independentcentresof the Campagna,some flourishing,someghost-ridden;above all the suburbof the dead, ringing the city and stragglingout alongthe roadsandbyroadsnearby,the tombs and cemeteries,pagan and then Christian, that drew a steadystreamof

thepiousandtheholiday-makers, andgreatcrowds onthegreatfestivals. TM But therewas only one suburbium,the rusticretreatof the urbanelite. No otherregion was quite like it, but then no otherregionlay adjacentto the heart of the empire. In the late Republic, the electionof a native son to high office at Rome would causelittle stir amonghis fellow citizensat

Tusculum: Tusculumwasfilled to burstingwith ex-consuls. 75But when a little hill town on the Samnite border seventyor eighty miles away produceda Roman magistrate,that was a great event for the town and evenits neighbours.Why shouldAtina know noneof the nobilityof Rome so familiar to Tusculum?The answerwas simple,and simplysummedup

by Cicero:it wasnontamsuburbana, notsocloseto Rome.76 Edward Champlin

PrincetonUniversity NOTES

1. H.J. Dyos, Victorian suburb:A studyof the growthof Camberwell(1961) 25, and chapter1 in general. 2. Apparentlythe only classicaluse of the word suburbiumis that of Cicero, Phil. 12.24. OtherwiseI have foundit only at Schol.ad luvenalem4.7.

3. This paperis meantto be no morethanan introduction to a subjectfor which there is a vast amountof information,much of it undigested.Two late Romanbureaucrats compileda work, now lost, De !ocissuburbanisvel diversis itineribuspergentiumin suasregiones(exiguousfragmentscollectedin F. Blume eta!., Die Schriftender riimischenFeldmesserI (1848) 347 f.). The only modem historicaltreatmentshave beenthoseof the archaeologist G. Lugli, "II suburbio

THE SUBURBIUM

OF ROME

111

di Roma", BCAR 51 (1923) 3-52=his Studi minori di topografiaantica (1965) 368-383; and of L. Quilici, "La Campagnaromanacomesuburbiodi Roma", PP 29 (1974) 410-438, and "La villa nel suburbioromano:problemidi studioe di inquadrimento storico-topografico", ArchClass31 (1979) 309-317. Of theimmense amountof archaeological and topographical informationthereare severalusefulif

out-of-date surveys: T. Ashby,TheRomanCampagna in classical times 2 (1970), and his seriesof articleson "The classicaltopographyof the RomanCampagna" in PBSR from 1 (1902) 125-285 to 5 (1910) 213-432; and G. Tomassetti, La Campagnaromana antica, medioevalee moderna, 4 vols. (1910-1926; revised and expandededitionin 7 volumesby L. Chiumentiand F. Bilancio (1979)) and more importantlyhis huge paper "Della Campagnaromana nel medio evo" in almostevery volume of Archivio della SocieuiRomana di Storia Patria from 2 (1879) 1-35 to 30 (1907) 333-388. Much of this informationwill presumablybe replacedby the volumesof Forma Italiae as they appear:see e.g., the splendid

volumeof L. Quilici, Collatia (Forma Italiae I 10, Rome n.d.), and particularly the "Letturastoricadel territorio"at 27-55. For a conciseintroductionto the major survivingremains, see F. Coarelli, Dintorni di Roma (1981). The period surveyedin this papercoversroughlythe secondcenturyBc to the sixth AD, with necessarilyheavy concentration on the yearsbetweenCicero and Tacitus. I must emphasizeagainthat this is a broadintroductionto the subject:I havethereforelargelyignoredvariationswithintheregionandchanges overtime. 4. Dion.

Hal. 4.13.3-4.

I am indebted for this reference to N. Purcell.

5. Yet Symmachus(Epp. 3.13.2) couldpretendthat Spoletowas suburbanto Rome. The 100th milestone,from at leastthe later secondcenturyAD, markedthe limit of the urban prefect'sjurisdictionand of that of lesserofficers, the urbana dioecesis,and had been usedearlier for (e.g.) the boundaryexcludingdediticii from Rome. See J. Partsch,"Der hundertsteMeilenstein", in Beitr/igezur alten GeschichteundGeographie.FestschriftfiJrHeinrichKiepert(1898) 3-19, discussing Rome's boundariesin general;R. Thomsen,The Italic regionsfrom Augustusto the Lombard invasions(1947) 153-163; A. yon Gerkan, "Grenzen und Gr6ssen der vierzehnRegionenRoms," BJ 149 (1949) 5-65. 6. I selectone referencefor each:Ovid, Fasti 6.57 if. assuresTibur, Praeneste, Bovillae, Aricia, Lanuvium and Lavinium; Martial 4.64.14 f., Saxa Rubra and Fidenae;Pliny, NH 14.50, Nomentum;Cicero, Att. 7.3.6 + 12.34.1, Ficulea; Liber Magonis (F. Blume et al., Die Schriftender r•imischenFeldmesserI 349),

Gabii; Cicero,De Oratore 1.98, Tusculum;Suetonius, Aug. 6.1, Velitrae;Martial 5.1.3-4, Antium; Pliny, NH 26.19, Ager Pomptinus.Note the virtual absenceof Etruria in the literature:beyondthe immediatevicinity of Rome it may have been thoughtlessdesirable(seebelow), yet the greatroadsinto Etruriaarejust as lined with gardens,villas and imperialpalacesas thoseon the otherside of the Tiber (including,for that matter, the Via Ostiensis),and oblique argumentscan show that the ancientsconsideredthem suburban.E.g., EdwardChamplin,Fronto and AntonineRome (1980) 22 f.; or the villa of the Septimiiat Veii (HA Severus4.5 as emendedby Hammond,with Statius,Silvae4.5.55), the futureemperor'sonly propertyin Italy.

112

EDWARD

CHAMPLIN

7. Strabo5.3.12 (Loebtranslation,my italics). Seethe mapsin Ashby,Quilici, etc. (n. 3 above).Generallyspeaking,the closerto Rome the denserthe habitation: cf. T.W. Potter, The changinglandscapeof southernEtruria (1979) 120 if. 8. Martial 6.43.9 (urbis vicina etc.); Columella 1.1.19; Pliny, Epp. 2.17.2 and 1.24; Suetonius,Aug. 82.1.

9. FIRA2 1.68.5,104ff.: senatusconsultum of 4 BEempowers themagistrate empanellingcertainsenatorialjuries to selectfrom thoseliving in Rome or within 20 miles of the city. 10. For the notion that proximity to Rome distinguisheshorti see D.R. ShackletonBailey, Cicero'slettersto AtticusIII (1968) 307, cf. Phoenix30 (1976)

209.Onthepurpose, nature andlocation, P. Grimal,Lesjardins romains 2(1969). 11. E.g., Statius,Silvae 4.4.7 (suburbanihorti), cf. CIL II 4332=Inschr. v. Tarraco 368 (hortos coherentessive suburbanurn).

12. Thus, Cicero, Att. 8.2.3 "non in suburbanis,non in hortis, non in ipsa urbe";Cicero,Phil. 8.9 "aedisoptimas,hortos,Tusculana,Albana";Tacitus,Ann. 14.53.6 "hortos... suburbana. .. agrorumspatiis. .. lato faenore"(q.v.); Ulpian, Dig. 49.4.1.9 "quodad domumeius non veneritquodquein hortosnon accesserit, et ulteriusquodad villam suburbanam;" andaboveall Nepos,Att. 14.3 "nullos habuit hortos, nullam suburbanamaut maritimamsumptuosamvillam, nequein Italia, praeterArretinumet Nomentanum,rusticumpraedium"(a passage worthfurtherthought). 13. One connotation of thisproximityis certainlya heightenedsenseof refuge, a greaterartificialityfor thosewho couldaffordit, with the emphasis on pleasure andlittle hintof moreseriousoccupations. Comparetheoutskirtsof eighteenth-century London, where private parks nearestthe city tendedmore to stablesand greenhouses than to livestockand crops, farms or timber. "They are ornamental gardens,suburbanplaygrounds for men-about-town. Their charmsare rustic,not rural" (H.C. Prince, in J.T. Coppockand H.C. Prince, Greater London(1964) 343). The samedistinctionshouldapply between(say) Lucullus' horti and his Tusculanum.

14. Dyos, op. cit. (n. 1) 21.

15. I havefoundthefollowingparticularly illuminating: Dyos,op. cit. 19-50; F.M.L. Thompson,Hampstead.Buildingof a borough,1650-1964(1974) 1-131; L. Stoneand J.C.F. Stone, "Countryhousesand their ownersin Hertfordshire, 1540-1879",in W.O. Aydelotteet al., Thedimensions of quantitative researchin history(1972) 56-123;L. Stone,"The residential development in the WestEnd

of London in theseventeenth century," in B.C. Malament (ed.),AftertheReformation. Essaysin honorof J.H. Hexter (1980)167-212;D.R. Coffin, The villa in the

lifeofRenaissance Rome(1979)7-60;J. Delumeau, Viedconomique et viesociale de Romedansla secondemoitiddu xvie siOcle(1959).

16. E.g., Seneca, Epp.104.1:"inNomentanum meumfugi•-quidputas? urbem? immofebremet quidemsubrepentem", cf. "protinus mutatam valetudinem sensi"

andBen.4.12.3(n. 35,below); orCatullus 44onhisvillasuburbana: "malamque pectoreexpuli tussim". Salubritas of the suburbis acclaimedfrom at least Cicero's

Cato(Cicero,Rep.1) to Symmachus (Epp.2.22.1,cf. 3.50, 9.83). 17. E.g., Juvenal 3.190if., contrasting thequietof Praeneste, Volsinii,Gabii

THE SUBURBIUM

OF ROME

and Tibur with Rome; Martial 6.43, "Nomentani...

113

otia ruris", cf. 12.57; Strabo

5.3.5, Antium asthe retreatof Rome'srulers.ScipioandLaeliusrelaxedat Laurenturn (Cicero, De Orat. 2.22), while Symmachusenjoyedits otium five centuries later(Epp. 4.44). Cf. J. Summerson on GeorgianLondon(1946) 255: "The essence of suburbanlife is the townsman'sdeliberatepursuitof healthand relaxation." 18. Epp. 3.82. 19. Epp. 2.17. Amoenitasis notedalso from at leastthe day of Sulla (Cicero, Rosc. Am. 133) to that of Cassiodorus(Var. 4.51.2). 20. Ps.-Jerome,Epp. 6, Ad amicumaegrotum2 (PL 30.83B). I owe this reference to Peter Brown.

21. B.W. Frier, Landlord and tenant in Imperial Rome (1980) 39-47. Note Suetonius,Tib. 35.2: a senatorwho waited in his horti until 1 July, when rentsin Rome went down. Particularlysuggestive(and not notedby Frier) are the many over-confidentRepublicanswho, beforePharsalus,sentto Rome to rent houses appropriatefor praetorsand consuls(Plutarch,Caesar42.2): that is to say, they presumablydid not own appropriatepropertyin the city. 22. Aug. 72.2. 23. 2.38.

24. Implicitelsewhere, citedexplicitlyasapleasure atTerence,Eun.971-973. 25. Cf. the works cited in n. 3 above.

26. Livy 33.6.7, withthecomment ofJ. Briscoe adloc.A verysimilarsituation in the suburbof Rome at Tacitus, Hist. 3.82.

27. AgenniusUrbicus,in F. Blumeet al., Die Schrifienderrrmischen Feldmesser I 88 and 86.

28. Strabo5.3.2; Lucan7.391; Cicero,Planc.23; Jerome,Epp. 127.8:all standard references. Seeaboveall P.A. Brunt,Italian manpower (1971)345-350 for discussion. Forthearchaeologist's viewof "a persistent tendency towards the absorption of smallholdings intolargerestates", especially nearRome,seeG.D.B. Jones,PBSRn.s. 18 (1963) 146 if., andcompareA. Kahaneet al., PBSRn.s. 23 (1968) 151 ff. See the informativesurveyby L. Quilici of settlements in the countryside,at PP 29 (1974) 425-430.

30. Respectively, Cicero,Leg. Agr. 2.78; Seneca,Epp. 87.7; Martial4.64.31 if. (cf. 1.85: slaves,flocks,produce);IGRR I 194.1.50 if., 2.9 if., 20, 23 f. Suburbanestatesof 1000 iugeraat Cicero,Att. 13.31.4 andVarro, RR 2.3.10. In

lateantiquity, theImperialpalacead duaslaurosontheviaLabicana purportedly covereda vastarea:LP 1.183 D, cf. JDAI (R)72 (1957) 45. CompareCoffin (n. 15 above)369:thegrowthof Renaissance villeggiatura hasa similareffect,larger estatesdriving out smaller.

31. The materialfor Latiumis to be published by M.-A. Tomei,to whomI am indebtedfor showingme somepreliminaryresults.

32. Pliny,NH 14.48.At NH 14.50,Plinyis uniquein claimingthatpricesin suburbanareas were low (nota vilitas), but the text is troublesomeand the sense

doesnotseemto fit thecontext.Contrast Columella, recommending suburban land to hisreadersaboveall otherlandasbeingeasyto oversee,andreadersof Martial will recallthat onedid not expecta profit from one'ssuburbanum. 33. Catullus44; A. Kahaneet al. (n. 28 above) 153-157.

114

EDWARD

CHAMPLIN

34. Columella 1.1.19 (Loeb translation);Pliny, Epp. 6.19.1 35. Seneca,Ben. 4.12.3: the point of the remark is not that an invalid caresfor nothingmore than health, not even money, but rather that suburbanvillas were boughtfor their beneficialqualities,not for seriousagriculture,henceonepaidno regard to profit. Tibur had preciselythe same attractionsas Hampsteadwould have, with the sameresults:"The slopesgaveadvantages of drainage,watersupply, andfreshair whichthewealthiersortwerepreparedto payfor, thusdrivingproperty valuesbeyondthereachof thelowerranks"(F. M. L. Thompson(n. 15 above)73). 36. Pliny, NH 22.12; Cicero,Att. 4.5.2. 37.

Pro Balbo 56.

38. D.R. ShackletonBailey, Cicero's letters to Atticus V (1966) 404-413. 39. For low succession rate amongthe senatorialorder,at least,seeK. Hopkins and G. Burton, in K. Hopkins,Death and renewal (1983) 31-200. 40. Cicero, Leg. 2.3-5; Pliny, Epp. 7.11.5 and 2.15.2, explicitly excluding ancestralestatesfrom thosefor sale to a friend. For rapid turnoverin antiquity: E. Rawson,"The Ciceronianaristocracyand its properties",in M.I. Finley (ed.), Studiesin Roman property (1976) 85-102• with importantmodificationsby S. Treggiari, "Sentimentand property:some Roman attitudes",in A. Parel and T. Flanagan(ed.), Theoriesof property (1979) 53-85. The phenomenonof clinging to family land in the countrywhile easilyacquiringand discardingpropertynear the city is observablein earlymodernLondon:L. Stone,"Residentialdevelopment" (n. 15 above) 195. For sale,resaleand exchangein Renaissance Rome, seeCoffin (n. 15 above)59: "The desireto seekthe mostbeneficiallocationsfor villeggiatura createda confusinggameof •musicalchairs' . . . ." 41. On the centralimportanceof the shrinesof Latium: G. Bodei Giglioni, "Pecuniafanatica.L'incidenzaeconomicadeitempiLaziali", RS189(1977) 33-76. 42. Note in this respectthe importantpaperby N. Purcell, "The apparitores: a studyin socialmobility", PBSR 51 (1983) 125-173, esp. 161 if. 43. This is very difficult to quantify without a full analysisof the evidence. Take for example,however,the convenientlistsprovidedby W. Eck, "Die Pr•isenz senatorischer Familienin den Staidten desImperiumRomanumbis zum sp•iten3. Jahrhundert",in W. Eck et al., Studienzur antikenSozialgeschichte. Festschrift Friedrich Vittinghoff(1980) 283-322. List Ia gives examplesof senatorsholding municipaloffice in Italy, IIa examplesof non-magisterial activity. To take each

itemwithoutfurtherdiscussion, thereare34 instances of suchpresence in suburban towns;of these17 itemsare instantlyrecognizable as concerning nativegentry,8 more can be excludedas a very specialcase(n. 49 below), and 3 are to all intents anonymous.Given the extraordinarilyhigh densityof senatorsin the region,their local showingis particularlypoor. This will haveto be developedelsewhere.(My impressionis that local activityincreasedunderthe later Empire:perhapsa result of the Romanaristocracy'slosingits Mediterraneanrole.) 44. Cicero,Att. 9.9.4, 13.47a.1 (Antiurn):soldby Ciceroto Lepidus. 45. For example, from Cicero's lettersto Atticus in the summerof 45: "commodum discesseras heri cum Trebatiusvenit, paulo post Curtius, hic salutandi causa, sed mansit invitatus. Trebatium nobiscum h•bemus. hodie mane Dolabella.

THE

SUBURBIUM

OF ROME

115

multussermoad multumdiem"(13.9); "de Varroneloquebamur: lupusin fabula. venitenimad me.... paulopostC. CapitocumT. Carrinate"(13.33a);"fuit apud me Lamia post discessumtuum..." (13.45). 46. Symmachus, Epp. 2.57.1; Pliny 5.6.45 ("nemoaccersitorex proximo").If Cicerois historicallyfaithfulatRep. 1.14, 17, 18, suchgatherings occurredalready in the mid-secondcenturyBC. 47. Recordedby O. Hirschfeld,"Der Gmndbesitzder r6mischenKaiserin den erstendrei Jahrhunderten",K!io 2 (1902) 55-70=his Kleine Schrifien(1913) 528544. There is no completemodemcollectionof the archaeological record. 48. J.H. D'Arms, Commerceand social standingin ancientRome (1981) 95 f., emphasizes the effectof the presenceor absenceof the imperialcourton social and economiclife aroundthe Bay of Naples, drawinga parallelwith the rise and declineof the ch•teauxof the Loirevalley. For comparison thereis alsothe court suburbof "ImperialKensington",a villagewith not only greatsuburban mansions and parksbut a royal palaceactingas a magnet,both to rural aristocrats needing a place near town and to Londoncitizenson the rise: D.A. Reeder, "A theatreof suburbs.Some patternsof developmentin West London, 1801-1911", in H.J. Dyos (ed.), The studyof urban history(1968) 253 if. 49. lnscr. It. 4.1.98,109, 113, 115, 126, 127,128,129 (etal.). On theSpanish colony,R. Syme, Tacitus(1958) 602. 50. De Officiis 1.150. 51. Note alsothe articleby N. Purcellon the apparitores(n. 42). 52. Referencesin nn. 36 and 37. Similar resentmentat Plutarch,Pompeius40.8 andCicero,Rosc.Am. 133, bothagainstfreedmen. 53. On the conceptof otium, literaryandotherwise:J.-M. Andr6, L'otium dans la vie moraleet intellectuelleromainedesoriginesd l' gpoqueaugustdenne (1966). At Epp. 1.24, the youngerPliny, looking out for a suburbanfarm for Suetonius, listsas oneof a scholar'sneedsenoughlandfor enjoymentbut not distraction. 54. Suburban:Topica, De Oratore, Brutus, TusculanaeDisputationes,De Divinatione,De Republica,De Finibus(thelasttwobooks),De Amicitia;andpossibly De Natura

Deorum

and De Senectute.

55. On the latter:J.H. D'Arms, Romanson the Bay of Naples (1970). 56. R. Poggioli, The oatenflute. Essayson pastoral poetry and the pastoral ideal (1975) 1-41; T.G. Rosenmeyer,The greencabinet.Theocritusand theEuropeanpastorallyric (1969). Note D.M. Halperin,Beforepastoral:Theocritusand theancienttraditionof bucolicpoetry(1983) 61-64, quotingL. Marx, Themachine in the garden(1964) 22: "Ideal pasturehastwo vulnerableborders:one separates it from Rome, the otherfrom the encroaching marshland.... Living in an oasis of rural pleasure,he (Tityms) enjoysthe bestof both worlds--the sophisticated orderof art andthe simplespontaneity of nature." 57. Most of Rome's major poets, it might be noted, ownedpropertyin the suburb,for whichtheyexhibita warm, if urbane,fondness.Of theleadinglights: Terence,Catullus,Horace,Tibullus, Ovid, Persius,Statius,Martial, possiblyJuvenal. We know of no properties belongingto Plautusor Lucretius;Vergil had a farm near Nola.

116

EDWARD

CHAMPLIN

58. Cicero,Verr. 2.4.126; Varro,RR 1.2.10; cf. Pliny,NH 36.23, 115, etal. 59. A statementnecessarilyvague, but true: the provenances of the art work accumulatedoverthe centuriesin the existingcollectionsin Romehaveneverbeen collectedandanalyzedto form a pictureof ancienttaste.Impressions canbe formed from R. Lanciani, Storia degli scavidi Roma e notizieintornole collezioniromane di antichitd,4 vols. (1902-1912),coveringthe years1000-1605;and H. Helbig,

Fiihrerdurchdie6ffentlichen Sammlungen klassischer Altertiimer in Rom4, ed. H. Speier,4 vols. (1963-1972). Brief remarksat L. Quilici, PP 29 (1974) 433; C.C. Vermeule, Greek sculptureand Romantaste(1977) 47 f., 65-72. 60. Macrobius,Sat. 3.13.3; Pliny, NH 16.242; Seneca,Epp. 104; Columella 3.33; Pliny, NH 14.51 and 15.83; Cicero,De Domo 62. 61.

Martial

10.79.

62. The suburbwasof courserenownedfor bothstapleproduceanddelicacies: e.g. CIL VI 33840 (colonushortorumolitoriorumqui suntvia Ostiensi)andVarro, RR 1.16.3. (profitsfrom suburbanviolet and rosegardens),respectively. 63. Strabo5.3.11; Pliny,NH 31.42, etc. NoteCicero,Art. 5.12.3: "eosisanimo quo solesessede aqua." 64. Plutarch,Cato 19.1; Frontinus,Aq. 2.76. 65. Pliny, NH 31.42; Frontinus2.75, cf. Tacitus,Ann. 15.43. 66.

2.126.

67. Another subjectworth investigation,with considerableevidence.Early examplesat Dio 56.1.1, 56.31.2. Cf. Coffin (n. 15 above)149if., for Renaissance Rome.

68. In the end, he sat with his troopsin the suburband thuswon the senate's loyalty, saysDio (41.2.1 and 3.3), even summoningit to meet there, and (in Caesar'saccusation,BC 1.85.8) he ran the statefrom the gatesof the city. 69. Plutarch, Cato Minor 48.1; Cicero, Art. 4.11.1; Plutarch,Pompeius44.4; Cicero, Rab. Post. 6; Plutarch, Cicero 31.2 f. (an extreme instanceof seeking favour).

70. Much evidenceis collectedby F. Millar, The emperorin theRomanworm (1977) 22-27.

71. Dio 53.19, but note a coupleof casesof tamperingwith suburbantroops: Suetonius, Galba 12.3; Dio 79.4.6. 72. Suetonius,Tib. 72.1; Tacitus,Ann. 3.47.5; Dio 58.21.1 and24.1; Suetonius,

Nero47, Dio 63.27.3, 72.13.4; Herodian1.11.5, 3.13.1 ;HA Elag. 13.5 and 14.2.

73. FIRA2 1.68. (4 Bc, seen. 9 above);Dig. 32.41.6 (Scaevoladecidesthat

property at Gadesincluded a suburbana adiacens possessio); Dig. 33.9.4.4(Paul decideswhether"athome"includesintracontinentia orjustintramurum).Compare Cicero, Planc. 22, contrastingrusticvirtue with artificiumsimulationisvel suburbanurn vel etiam urbanurn.

74. J.M.C. Toynbee,Death and burial in theRomanworm (1971), for a convenientintroduction.For the Christianpresence,L. Reekmans,"L'implantation monumentale chr6tienne dans la zone suburbaine de Rome du iv eau ix e si•cle",

RAC 44 (1968) 173-207;and for its urbansignificance,P. Brown, The cult of the saints.Its rise andfunction in Latin Christianity(1981) 42-49.

THE

SUBURBIUM

OF ROME

117

75. Cicero, Pro Plancio 19. Cicero'spoint is thatTusculumhasmoreconsular familiesthan all the otherrnunicipiaput together,as he saysat the beginningof the passage,while Atina of coursehasnone:nativesonsrise becauseof proximity to the capitalin the former. Yet he is also sayingthatthe inhabitants of Tusculum are blas• quia referturnest rnunicipiurn consularibus, whichcan only meanthat they seeconsulars of all kindsdaily, that is the suburbani,for the simplereason that in 54 Bc therewere no knownconsularsof Tusculanorigin. 76. My thanksfor helpfulcommentfrom N. Purcelland, again,for the patient and thoughtfulcriticismof ProfessorBadian.

ISOKRATES'

ON THE PEACE: POLITICAL

RHETORICAL ADVICE?

EXERCISE

OR

Isokrates remains anelusivefigurein thehistoryof fourth-century Athens. • By professionhe was an educator,but today we are more interestedin Isokratesthe masterrhetoricianand political philosopher.In assessing theseaspects of Isokrates'worksrecentscholarship hastendedto emphasize the rhetoricalelements.Severalarticlesand monographs have beenproducedin reactionto the nineteenth-and early twentieth-centurytendency

to overemphasize thehistorical and/or political significance of thecorpus. 2 In reopeningthis debateI want to stressat the outsetthat I do not doubt Isokrates'devotionto rhetoricaltechniquesnor do I meanto denigratethe value of recentresearchinto the effectswhich thosetechniqueshad both on Isokrates'style and the contentof his works. My goal is to prevent thoseeffortsfromobscuring thehistoricalsignificance of Isokrates'political pamphlets. The lengthof Isokrates'careerand the varyinghistoricalcircumstances in which he wrote his political speechesmake him vulnerableto charges of inconsistency.Seeminginconsistencies have led somecritics to doubt the sincerityof the political views he expressesin his major works. The trendin Isokrateanscholarshiptowardsreevaluationof whole speecheson the basisof apparentcontradictions in thepoliticaladviceexpressed needs to be examinedcarefully in the light of what we know about the real purposeof each individual oration. This study is an attemptto clarify Isokrates'intentionsin writing his orationOn the Peace. A proponentof the trend to view Isokratesas almost exclusivelya rhetorician,Phillip Harding, has arguedthat we have extantin the works of Isokratesa pair of epideicticorations--aview whichhadnotpreviously

beenexpounded. 3 HardingpositsthattheArchidamos andOn thePeace were written as an antilogicalunit sometimeafter 355--the first speech arguingthe casefor war using a specificcontextin history,the second using a different historicalsituationto make the oppositeargumentfor peace.It is envisionedthatIsokrateswrotethesetwo set-pieces at the same time as a rhetorical tour de force for the edification of his studentsand for the greaterglory of his school.This attractivetheoryis howeverbrought into questionby Isokrates'own commentsin the Antidosiswhichtestify to the historicalauthenticityof On the Peace. 118

ISOKRATES'

ON THE PEACE

119

The argumentthatOn thePeacewasnotwrittenascontemporary political advice4 to the Atheniansat the end of the Social War, but was rather written as a rhetoricalexerciseto arguethe case for peace, is basedon two points. These are, (1) inconsistencies in advice, and (2) patterned oppositionof rhetoricalpointsin On thePeaceandArchidamos.As to the firstpoint,Hardingasserts thattherearemanycontradictions in thepolitical advicepropoundedin Isokrates'writings.In particular,he arguesthat the whole of Archidamos"runs counterto views that Isokratespropounds

elsewhere". 5 Hardingsupports Baynes'viewthattheArchidamos ispurely a "show-piece",againstthe opinionsof Mathieu, BringmannandCargill,

all of whomconsider it a significant politicaloration. 6 Secondly,Harding notesthat the ideasof Archidamosseemparticularly at odds with the views Isokratesexpressesin On the Peace. He then proceedsto point out sevenspecificoppositionsof rhetoricalargumentin passagesfrom the two works, concludingthat sincethe two speechesare "so closelyopposedthey mustbe takentogether"and that they "represent an exampleof the sophistic-rhetorical deviceof antilogiai,two speeches

arguing twosidesof thesametheme. "7 Fromthispointof view, thesespeeches weremeantto illustraterhetorical argumentsfor oppositepolicies. If we considerOn the Peace serious advice, we shouldconsiderArchidamosthe same, and converselyif Archidamosis epideictic,thenOn thePeaceshouldalsobeconsidered epideictic. This leadsto the conclusionthat if the two were antilogicalthey must have been "conceived as a unit", (147) and since On the Peace has the later of the two historicalsettings,they were probablynot written until after 355/4.

Hardingspeculates thatthosewho "believein thetopicalityof Isokrates' speecheswill resistthis conclusionand make objectionson the basisof the historical referencesin the Archidamos" (148). In the first half of his predictionhe was correct,but I basemy objectionnot on the historicity of referencesin the Archidamoswhich, as he says,could easilyhave been includedby Isokratesat a later dateto lend an air of authenticity.Rather, onemustcite Isokrates'own testimonyto the authenticityof On thePeace. This testimonyis presentedin Isokrates'elaboratepersonalapologia, the Antidosis(62 if.), and, to my mind, offers proof that On the Peace was

writtenas actualpoliticaladviceto the Athenianstowardsthe end of the Social War.

Beforeproceedingto the chief proofthat On thePeace was an authentic political pamphlet, one must examine the specificpoints of rhetorical oppositionbetween that speechand the Archidamosas enumeratedby Harding. There can be no doubtthat the themesof the two orationsare in

120

ROBERT

A. MOYSEY

part antithetical.Thereforeit is not surprisingthat the specificarguments Isokratesmakesin the two differenthistoricalcontextsare virtually oppositestoo. This simplefactneitherprovesnordisprovesthehistoricalauthenticity of the speeches. Harding admitsthat one of the parallelargumentsin the two speeches, the exhortationto live up to one's ancestors,was a standardtheme"commonly employedby other fourth-centuryorators"(144) and found frequently in Isokrates' various works. There is no difference in Isokrates' useof thisthemein the two worksin question.Isokratesurgesthe Spartans to imitate their ancestors'military prowessin Archidarnos16-24, but tells the Athenians(in On the Peace 37- 44) to avoid the evils of fifth-century imperialismand emulate insteadthe foundersof the Delian League. This "inconsistency"is necessitatedby the argumentsbeing put forward in oppositehistoricalcontexts.The theory that the two speechesmight be antilogicalepideicticworksis neitherprovednor disprovedby this sortof circularargument. The same statementcan be made with regard to Isokrates'comments aboutempireand confederacy.On the Peace is not the only Isokratean orationin whicharch• is denouncedandsyrnmachiaof free andautonomous

alliesis applauded. 8In thecontext ofArchidarnos Isokrates naturally urges Spartato recoverher archg in the Peloponnesus. He doesnot however urgetheSpartans to reacquire theirpost-Peloponnesian War empirebeyond the Peloponnese.Isokrates himself differentiatesbetween Sparta's hegemonyand her attemptto win an arch• of the sea (On the Peace 100-103).9 HardingusesArchidarnos 24-25 to arguethatIsokrates'attitudetowards Thebes is inconsistentin the two works (145-146). This view cannot be

credited.Isokrateswasneverwell disposedtowardsThebesandwasespe-

ciallyopposed to Thebanimperialism. •0Isokrates musthaverealized that liberationof Sparta'ssubjectstateswasnot the Thebans'true objective. They wantedto hamstringSpartaandassumeunquestioned hegemonyof Greece.This wastheunderlyingconsequence of theTheban-dictated Peace of 366 which Archidamosin his speechurgesthe Spartansto resist.One mustdistinguish betweenIsokrates'expressed idealisticobjectives of freedom andautonomyfor Greekstatesandhispracticalrealizationthatthese idealswere not soonto be realizedby stateslike Messenia. Threefurtherpointscanbemade.First,in thecontextof theArchidarnos IsokratescouldhardlyhaveArchidamos proposethesurrender of Messenia even if Isokrateshimself thoughtMesseniashouldbe freed. Second,I doubt that Isokratesdid applaudthis action, sinceMesseniawas only

"liberated"by the Thebansto be a thomin the sideof the Spartans and

ISOKRATES'

ON THE PEACE

121

to furtherThebanhegemony whichIsokrates neverfavored. •t Third,there was a differencein the caseof the Messeniansin thatthey hadbeensubjects of the Spartansfor the entire classicalperiod. Archidamoshimself, in Isokrates' oration, claims that Messenia had been held for more than 400

years(Archidamos26 f.). The next contrastingthemeis attitudetowardstyranny.Harding (146) pointsout that Dionysiosof Syracusecomesin for indirectpraise, or at least credit for success, in Archidamos 44, while in On the Peace 91,

111-113 and 142 tyrannicalempiresare denounced.Isokrates'point in mentioningDionysiosis to praisehis tenacityagainstforeigninvaders, not to praise his archg. Note that just before (in 6.42), Isokrateshas Archidamospraisethe Atheniansfor defendingthemselves againstinsolent invaderswhile observingthat they lost their good reputationwhen they attemptedto dictateto otherGreekstates.Isokratesthencontinues:"Athens, however,is not the only instanceby which one might showhow greatare the advantages of daringto resistone'senemies.There is alsothe caseof the tyrant Dionysius. . ." (6.44, transl. G. Norlin). Thus, Isokrateswas not making a point of praisingDionysios'tyrannicalempire and there is no oppositionto his denunciationof suchempiresin On the Peace. The next oppositionof themeis no moreconvincing.Hardingfinds it "striking. . . thatthe Atheniansareheldup as examplesto the Spartans and the Spartansto the Athenians"(146). He refersat this point to Archidamos 42- 43 (cf. 73), whereIsokrates hasthespeaker usetheexample of Athens' recoveryafter the evacuationof Attika at the time of Xerxes' invasionto urgethe Spartansnotto be discouraged by recentsetbacks and to carryon the fight, even if it is necessary to evacuatewomen,children and the elderly from Sparta.Likewise in On the Peace 95 Isokratesuses the exampleof Sparta'sdeclineafterher acquisition of a navalempireto urgehis fellow Atheniansto giveup theirdreamsof reacquiring a maritime arch& Surely there is no premeditatedor calculatedcontrasthere. The exampleschosenare not only obvious,but amongthe chief examplesin Greekhistoryfor thepointsIsokrates is makingin thetwoopposite contexts. The choiceof examplesfrom the follies or gloriesof the otherleading Greek state'shistory is certainlyno rarity. Thucydidesand Xenophon reportthat Athenianactionswere discussed in Spartanassemblies as precedentsor examplesandreferences to Spartandeedsare not infrequently

citedin Athenian decision-making councils. 12 The suggestion that "mercenaries are also treateddifferentlyin these two speeches" can be disputed.In On the Peace 79 (not 73, as Harding says),Isokrates,decryingtheevilsof theAthenianEmpire,castsaspersions at the sortof menhiredto row the triremes,but his mainpointis thatthe

122

ROBERT

A. MOYSEY

Atheniansof that age mistreatedtheir fellow Hellenes. The use of hired rowersof depravedcharacteris only a secondarypoint. In Archidamos 76, Isokrates'purposeis notto glorifymercenaries. He is makingthepoint that if the Spartansunitein the resolveto reconquerMessene,no Peloponnesianstatecouldsuccessfully resistthe determinedSpartanarmy "which, in its freedomfrom ordinarycaresand in havingno otherduty but that of war, would resemblea mercenaryforce, but in point of nativevalourand of disciplinedhabitswould be like no army that couldbe leviedin all the world..." (transl.Norlin). Isokratesis not praisingmercenaries, he only draws a parallel betweenthe freedomfrom other concernswhich mercenariesshowand the undividedSpartanresolvewhich he urges.In fact, Isokratesimplies criticismof mercenaryforcesin this passagewhen he speaks of nativevaloranddiscipline(whichmercenaries traditionallylack). Finally, thereis no formaloppositionof thesetopicsin any recognizable schematic designor pattern,of thesortonemightexpectin two consciously plannedrhetoricalset-pieces.If sucha patternedstructureof opposing views existedin the two works one might be able to make a casefor consideringthem antilogiai despitethe lack-lusterquality of the oppositions, but none exists in the works under considerationand other evidence

makesthis hypothesisunlikely. Let us turn then to the evidencewhich seemsto prove the purposeof On the Peace. Before quotingthe passagein question,it is necessaryto justify the use of the Antidosisto prove the historicityof On the Peace. Isokrates wrote the Antidosis as a defense of his career in 354/3 BC.•3 He

was82 yearsold whenhe waschallengedto undertakea trierarchyin place of the plaintiff or exchangepropertywith him. Isokratestells us in the openinglinesthat the speechis neitherlike thoseproducedfor actualuse in the law-courtsnor an epideicticoration (my emphasis).Since it was an unusualpiece, Isokratesexplainsits purpose.He had been surprised by the numberof people who displayedantagonismtowardshim during his recenttrial concerningthe trierarchy.This attitudewas engenderedby the attacks of critics. He asserts that he decided to cast the defense of his

careerin the form of a speechfor a law-courtrather than attempta selfeulogywhichmightarouseevengreateranimosity(Antidosis1-8). It seems unlikely that Isokrateswould deny at the outsetthat he was writing an epideicticoration if that were in fact what he was doing. There was no needto deny the true natureof the work. Surelythis denialwas not meant to be ironic or to be part of the rhetoricalillusion. The denialitself would break the illusion of reality. To my knowledge,no one has arguedthat

theAntidosis wasanepideictic work.14It seems veryfar-fetched tosuppose Isokrateswrotean epideicticorationspecificallysaidnot to be epideictic

ISOKRATES'

ON THE PEACE

123

aboutan imaginarytrial in which he quotedfrom otherepideicticorations in his imaginarydefense. Since this work had as one of its goals the exonerationof Isokrates' civic reputationin the eyesof his fellow Athenians(cf. Antidosis55-56), it is mostunlikely that he would have attemptedto misrepresent simple historicalfactsaboutwhich any activeAtheniancitizen would haveknown thetruth.Thereforethereis no reasonto doubtthevalidityof thetestimony given in this orationconcerningthe purposeof On the Peace. Indeed, the fact that On the Peace is one of only three of his orationsfrom which Isokrateschoseto quoteverbatimin his defensemakeshis testimonyall the more noteworthy. In the AntidosisIsokratesmentionsspecificvirtueswhich, he thinks, makehisdiscourses especiallycommendable to hisfellow Athenians.Each of thesevirtuesis supportedby a quotationfrom one of his orations.The first pointhe makesis thathe hasarguedfor Athens'rightto berecognized as hegemonof Greece. To supportthis assertionIsokratesquotesfrom Panegyrikos51-99 where this point was arguedin somedetail on the basis of Athens' past glories. He makes it clear (Antidosis57 if.) that the Panegyrikoswas not merely a rhetoricalexercise,but was specifically aimedat arguingthe casefor Athenianhegemonyca. 380. Isokratesthen allowsthat someof his criticswould arguethat theseare pretty words, but that "those discoursesare better and more profitable which denounceour presentmistakesthan those which praise our past deeds,and thosewhich counselus what we oughtto do thanthosewhich recountancienthistory" (Antidosis62, transl. Norlin). To prove that he hasalsoprovidedexcellentadviceon themostspecificcontemporary issues he citeshis argumentthat Athensshouldmake a peaceagreementto end the Social War. I quote the passage(Antidosis63-65, transl. Norlin, adapted):

At the beginningof this oration I speakon the questionof making peacewith the Chians, the Rhodians,and the Byzantines;and, after I have shownthat it is to the advantageof Athensto end the war, I decryour dominionover the Hellenesand our sea-power,showing that it is no different, either in its conduct or in its results, from

tyranny.I recall alsothe evils whichthat powerhasbroughtupon Athens, upon the Lacedaemonians,and upon all the others. After havingdwelt uponthis subject,deploredthe misfortunesof Hellas, and urgedAthensnot to allow herselfto remainin her presentstate, finally, I summonher to a careerof justice, I condemnthe mistakes she is now making, and I counselher as to her futurepolicy.

124

ROBERT

A. MOYSEY

Isokrates thenasksthatOn thePeace25-56and132-145be readto prove his claims.

It might seemthat thesetwo examplesare contradictory. One argues that Athenshas a right to hegemonyover the rest of Greecebecauseof her gloriouspast,while the otherarguesthat Athens'SecondConfederacy was tyrannicaland oughtto be abandonedfor the sakeof peace.In fact howeverthesetwo arguments makesensein theirproperhistoricalcontexts andarein keepingwith whatIsokratesconsidered to be in thebestinterests of Athens in those two different

situations.

The probabilitythat the Antidosiswas written only a year or two after On the Peace also lendsweight to the significanceof Isokrates'decision to quote from On the Peace in his defense.It might have been difficult for the averageAthenianto rememberwhat Isokrateshad saidor written ca. 380 (some were only childrenthen), but many of his readersand listenerswould have known and rememberedthe argumentfor peacehe had madeonly a year or two before. Again, the Antidosiswould needto be epideicticif we are to denythe historicalpurposeof On the Peace. The explicitstatementin Antidosis62- 65 abouttheseriouscontemporary purposeIsokrateshadin writingOn thePeacemustoutweightheconjecture Harding makeson the basisof supposedinconsistencies and oppositions of rhetorical

themes in On the Peace and Archidamos.

Thus the consensus

of scholarlyopinionon the historicalsignificanceof On the Peaceshould be sustained.z5

Universityof Mississippi

RobertA. Moysey NOTES

1. This article is basedin part on a paperpresentedon December28, 1981, at the AnnualMeetingof the AmericanPhilologicalAssociationin SanFrancisco. I thankmy colleagueProfessorEdwin Dolin who readan earlierdraftand offered suggestions for improvement,and the editorand anonymous readersof AJAH for their criticismsand suggestions. 2. See for instanceN. Baynes,"Isocrates",in Byzantinestudiesand other essays(1955) 144-167; G. Kennedy, The art of persuasionin Greece (1963) 174-205;A. Lesky,A historyof Greekliterature,transl.J. Willis and C. de Heer (1966) 584-591;R. Seager,reviewof K. Bringmann(below, n. 4), CR N.S. 16 (1966) 405, whocallsIsokrates"a manof smalloriginalityof thoughtandnegligible politicalinfluence";G. Heilbrunn,"The composition of Isocrates'Helen", TAPA 107 (1977) 147-159; E. Rummel, "Isocrates'ideal of rhetoric:criteria of evaluation", CJ 75 (1979) 25-35; and especiallyP. Harding(below, n. 3). 3. PhillipHarding,"Thepurposeof Isokrates'Archidamos andOn thePeace",

ISOKRATES'

ON THE PEACE

125

CSCA 6 (1973) 137-149. It might be noted that GeorgesMathieu, Les iddes politiquesd'lsocrate(1966) 116-117hadalreadyquestioned whetherOn thePeace andArchidamosmightnot be only show-pieces written later thantheir historical settings.

4. I prefer to use the more neutralterm "advice"ratherthanthe prejudicial term"propaganda" whichHarding(138) andothersemployin thiscontext.There has been considerabledisagreementaboutthe date of On the Peace. Antidosis 62- 65 makes it clear that On the Peace was written while the Social War was still

in progress.Internalevidencein On the Peace 15 furtherindicatesthat the peace hadnot yet beensigned,but Athenswaslosingthe war. This suggests a dateafter the defeat at Embata. Jebb'sargumentthat the speechshouldbe dated in 355, whilethepeacenegotiations werepending,still seemsvalid(R.C. Jebb,TheAttic

orators(1893)II 2 182-183).KlausBringmann, Studienzu denpolitischen ldeen des lsokrates(1965) 59 note 2, also dates the speechjust before or after the conclusion of peacein 355/4. SeealsoD. Gillis, "The structureof arguments in Isocrates'De Pace", Philologus114 (1970) 195-196andG. Nodin (ed.), lsocrates (1929) II 5. P. Orsini, "La date du discourssur la paix d'Isocrate",Pallas 12 (1964) 9-18, arguedfor a date in autumn356 Bc. W. Judeichin Kleinasiatische Studien(1892) 290-291 note I and P. Cloch6 in lsocrate et son temps(1963) 76 and 109 date On the Peace shortlybeforethe battleof Embata.Mathieu (above, n. 3) 116 datesthe orationat the end of the war, and E. Meyer, Geschichte des Altertums(1910) V.494 thoughtit was publishedin 354 after Athensrenounced theConfederacy. A. Momiglianodatestheworkin autumn355/4andfully supports thehistoricityof thespeech(Per la storiadellapubblicistica sullaKotv• E•6}•vrI nel secoloIV a.C. =Annali R. Scuola Normale di Pisa II 5 (1936) 110). 5. Harding 143.

6. Harding 143-144;Baynes(above,n. 2) 160-161;Mathieu (above,n. 3) 105-07; Bringmann(above, n. 4) 56; J. Cargill, The SecondAthenianLeague (1981) 176 note 29 and 177 note 30. Cloch6 (above, n. 4) 60 also considersthe Archidamosto have had a seriouspoliticalpurposein 366, thoughhe doesnot believethe speechwas writtenfor Archidamosto deliverin Sparta. 7. Harding(above,n. 3) 147. It is perhapsworthmentioningherethatIsokrates specifically denounces useof sophistic techniques in hisorationAgainsttheSophists (writtenearly in his career,accordingto Antidosis193) and indirectlycriticizes his own teacherGorgias'Encomiumon Helen in his Helen 14 (cf. Jebb(above, n. 4) 96-103, who datesthe work ca. 370 Bc). 8. See for instanceAreopagitikos17, Panegyrikos20-28 and Panathenaikos 53-58, 67. For more detailed observations on the use of the terms archd and

hegemoniaseeJ. Buckler,"Allianceandhegemonyin fourth-century Greece:the caseof the Thebanhegemony",The AncientWorm 5 (1982) 79-81 and C.D.

Hamilton,"Isocrates, IG ii2 43, Greekpropaganda andimperialism", Traditio36 (1980) 83-109 and especially95-97. 9. Cf. W. Jaeger,Paideia,transl.G. Highet(1944)II171-83, whoalsoargued thatIsokratesdistinguished betweenhegemonyandarchd. 10. For Isokrates'consistentoppositionto ThebesseeP. Cloch6, "Isocrateet

126

ROBERT

A. MOYSEY

Thebes", RH 193 (1942-43) 277-296 and J. Buckler, The Theban hegemony, 371-362 B.C. (1980) 21, 275.

11. In thecontextof On thePeace58, whereIsokratesis arguingfor moderation, he finds it convenientto hurl a missile--albeit indirectly--at the Thebans' imperialisticaimsby pointingoutthattheymighthaveshownmoderationandwisdom by liberatingtheGreeksandpursuingpeace,but theychosenotto do so. Isokrates' pointin thatpassage is thatThebesandAthenswerebothmakingthe samemistake in their mistreatmentof their fellow Hellenes,but it suggests that IsokratesunderstoodclearlythatThebeswasno truechampionof Greekautonomy.His opposition to Thebesis consistent (seen. 10). Thereforeone canhardlydoubtthat he would have immediatelyrecognizedthe Theban"liberation"of Messeniafor what it was and would havepreferredto seeit undertraditionalSpartandominationratherthan subservient to Thebes.

12. From as early as the late sixth centuryat least Sparta and Athens were recognizedasthe leadingpowersof Greece(Hdt. 5.97). Thucydides(1.68-72) has theCorinthianscontrastSpartaandAthensduringthefirstdebateat Sparta.Likewise the Athenianspresentcite Spartandeedsin theirdefense(Thuc. 1.72-76). Perikles cited Sparta'spolicy in expellingalienswhenhe refusedto revokethe Megarian decree(Thuc. 1.144). Cf. Thuc. 4.17-18; 6.33, 82-83. Xenophonreportsnumerous other examples,see Hell. 2.2.20, 3.34; 6.3.13-14, 5.33 if., esp. 43 and 46; 7.1.4-11. On Isokrates' use of historicalexamplessee C.D. Hamilton, "Greek rhetoricandhistory:thecaseof Isocrates",in Arktouros:Hellenicstudiespresented to B. Knox, ed. G.W. Bowersocket al. (1979) 290-298. 13. For the date see Antidosis 9.

14. The obviousliterary parallelsbetweenthe Antidosisand Plato'sApology have often been noted, but literary parallelsdo not disprovethe historicityof Isokrates'defense.Isokratesis writing a defenseof his own career,not attempting to eulogizehis martyredteacher.Isokratesconsciously emulatesPlato'swordsin somepassages,but the two works differ in contentand purpose.The plodding, tediousAntidosis,overburdened with exemplaandrhetoric,is almostthe opposite of the eloquentsimplicityand poignancyof the Apology. 15. Bringmann(above,n. 4) 58-74;Cloch6(above,n. 4) 72, 75-77, 82,102-111; /d. (above, n. 10) 284; Jebb(above, n. 4) 182-194; Kennedy(above, n. 2) 192; Mathieu (above, n. 3) 118; D. Gillis, "The ethical basisof Isocrateanrhetoric", PP 24 (1969) 329, 335,347; id. (above, n. 4) 195-210; G. Heilbrunn, "Isocrates onrhetoricandpower",Hermes103(1975) 154, 173-175;H.L. Hudson-Williams, "A Greek humanist",G&R 9 (1940) 166-169;id., "Political speeches in Athens", CQ N.S. 1 (1951) 68-73; F. Jacoby,Atthis (1949) 74 and note 19; W. Jaeger (above, n. 9) III 132; M.L.W. Laistner, "The influence of Isocrates'political doctrineson somefourthcenturymenof affairs",CW 23 (1930) 129-131;Miinscher in RE s.v. "Isokrates"2205; J. de Romilly, "Eunoia in Isocratesor the political importanceof creatinggood will", JHS 78 (1958) 95, 97-98 and "Les modrrrs athrniensvers le milieu du IV ½Si•cle", REG 67 (1954) 328, 332, 340-342, 349; L. Van Hook, "Alcidamas versusIsocrates",CW 12 (1919) 89-91; C.B. Welles,

"Isocrates'view of history",Studiesin honorofH. Caplan,ed. L. Wallach(1966)

ISOKRATES'

ON THE PEACE

127

19, 21; and mostrecently,J. Cargill (above, n. 6) 176 note 29, and "Hegemony, not empire", Ancient World 5 (1982) 91-102. In generalterms the validity of Isokrates'worksas historicalsourcesis upheldby A. Fuks, "Isokratesand the social-economicsituationin Greece", Ancient Society3 (1972) 17- 44, and G.L. Cawkwell, "Noteson the failure of the SecondAthenianConfederacy",JHS 101 (1981)40-55. Alsoin agreement is T.T.B. Ryder's"Isocrates andAthenianpolicy" as summarized in PCA 56 (1959) 30-31.

THE

VIA

EGNATIA

IN WESTERN

MACEDONIA

Part I: The routes through Lyncus and Eordaea in Western

Macedonia*

This article stemsfrom travelswhich we made individuallyand sometimes togetherin 1981 and 1982. In describingwhat we saw, we have usedthe expression"we" for convenienceeven thoughonly one of us may have beenthere. The article containsthe views which we hold in common. The first two sectionsare concernedwith the pre-Roman phaseof the road from Lyncusto Eordaea,the third sectionwith the Via Egnatia from Heracleato Edessaand the fourth sectionwith evidence of ancient settlementsin the vicinity of this stretchof the Via Egnatia.

I. Brasidas'withdrawalfrom Lyncusto Eordaeain 423 Bc When Brasidaswithdrew from Lyncus, harassedand pursuedby the Illyrians, he had to passthroughthe watershedrangebetweenLyncus and Eordaea, which links Mt Vitsi (2126m) on the west and the outliers

(1192m and 1440m) of Mt Barnouson the east. There are today only two routes:the northernpassingfor severalmiles throughdifficult hilly countryto the north end of Lake Vegoritis (Ostrovo), and the southern which is longer, but easy, and mostlyover open ground.It is certain that Brasidastook the latter;for his troops,beingmainly hoplites,had to avoid difficult hilly ground, and Thucydideswrote of "the pass", which is evidentlythe singlepassbetweenthe plain of Lyncusandthe plain of Eordaea.It was, too, the only routesuitablefor the ox-drawn waggonswhich carded the baggageof the Macedonians,the allies of

Brasidas. For the gradients arenot steepandthetwo combes • which constitutethe passcarryno streamsand are firm underfoot. Approachingfrom the north, one entersthe combenear the village of Vevi and ascendsvery gently and steadilyto its highestpoint, some 800m above sea level. This is by Klidhi station, where the railway crossesthe road. From the entry to this point the combeis about 150m wide with fiattish fields of arable red soil, which has in part been washed down from the limestonehills on either side. Even allowing 128

o

0

• Stadion-stone '•' Milestone

==

Via Egnatia

Based onaFrench Staff Map

,3

.i

Map 1: The Passof Kirli Dirven

. .'

/•-•

•

132

HAMMOND

AND

HATZOPOULOS

for the effectsof erosion,one can be confidentthat this part of the pass has not changedsubstantiallysincethe fifth centuryBc. From Klidhi

stationonebeginsthe descentof the othercombe,at first gradualand curvingto one's right, still in a flat areaof the samewidth. The wide partnarrowsandfinally becomesa cleft, andonedescends quitesharply to reachthe foot of the passbetweentwo knolls, onecarryinga votive chapel to one's left and the other a memorial to the British dead of 1941 to one'sright. In the narrowpartthereis an accumulationof earth broughtdownby erosionof the hillside;therea stadion-stone wasfound 2m below the presentsurface.After passingthe memorial one enters the wide rolling plain of Eordaea.Plate 1 showsthe narrowpart of the pass,looking southwards. This passis knowntodayasthe Kirli Dirven Pass.Thucydidescalled it simply "the passof Lyncus" (Thuc. 4.83.2 T.q•o[•o•,qTqgA6¾xov) and describedit as "the pass which is narrow between two ridges" (Thuc. 4.127.2 •qv •o[3o•.qv/] •o•t [t•m•/• 6vo•v •.6r. potv o•v/I). He distinguishedone part of the passas "itself the difficult part of the route" (ct/•T6q6n z6 &•ooov zqg 6600, which Jowett translated"the most dangerouspart of the defile"). To anyonewalking throughthe pass, as we did, there is no doubt that Thucydidesmeant the narrow part which beginsas one leavesthe fields southof Klidhi station. We turn next to the "two ridges"betweenwhich the passruns. That on the west side is the higher, attaining 1009m at its summit;that on the east does not exceed 860m and 767m. There is, however, a more

importantdifference.The westernridgestartsat theentry(belowVevi), andits top runsparallelto the line of the pass;the easternridgestarting a little southof the entry, runs away from the line of the passbehind Klidhi village and thenreturnsto form the easternside of "the difficult part". If, then, one were to approachthe passfrom the north and find at the entry that the entirepasswasblocked,one'sbestplan wouldbe to ascendthe westernridge, follow its fairly level top and descend alongthe spurwhichcomesdownto the foot of the pass.On the other handthe easternridgeimposesa long detour.The westernridgeis well wooded;the easterncames sparsebushesof garigue(pournfiri). In 423 Bc Brasidas,the Spartancommander,was desertedby the Macedonianforces,which fled duringthe night towardsEordaea,the nearestpart of their homeland, and he and his Greek troops, mainly hoplitesbut also somelight-armed,were set uponby the Lyncestians and their Illyrian allies. Although greatly outnumbered,Brasidas' Greeksdroveoff their attackersrepeatedlyand were clearlycapableof doing so as long as they were in the plain of Lyncusand had room in whichto manoeuvre.Consequently, only onegroupcontinuedto harass

THE

VIA

EGNATIA

I

133

the Greeks. "The bulk of the enemyrushedaheadto kill any Macedonianson their way and to occupyin advancethe narrowpassbetween thetwo ridges;for theyknewthatBrasidashadno otherline of retreat," wroteThucydides(4.127.2), andcontinued(we translateliterally):"and as Brasidasapproaches the alreadydifficultpart itselfof the route,they

areencircling him in thehopeof cuttinghimoff.2 (128.1)Realising their plan Brasidasgave advanceordersto the three hundred(these beinga pickedforceunderhis own command).They wereto proceed towardsthatoneof the ridgeswhichhe thoughthe wouldmorereadily capture,and they were to proceedat the double, as fast as each man couldmanage,and not in formation.They shouldattemptto drive off the ridge those barbarianswho were already upon it (or "about to

attack"), 3beforethelargerencircling forceof barbarians couldengage themthere. (128.2) And they, makingtheircharge,overpoweredthose on the ridge, and the main body of the Greekstoo was now making itsway to theridgewith lessdifficulty;for thebarbarians werefrightened by the routingthere of their own men from the high point, and they did notanylongerpursuefurther,reckoningthattheenemywerealready at the borderand had escapedtheir grasp." When Brasidaswas approaching "the... difficultpart itself of the route", he was in the fields between Klidhi station and the narrow

mouthof the pass,andthe ridgehis men were to seizewas that shown

onPlate2, ontheeastern side.4 It is in facta spurprotruding fromthe main ridge and givesthe appearance of beingan isolatedheight;it is this which was called•6 Cebaco@or ("the high point"), onceoccupied by the enemy but now capturedby the dashof the 300. The rest of Brasidas'forcessoonjoined them--it takesonly ten minutesto walk up. The Greeks,beingnow themselvesabovetheir enemy,were able to proceedquicklyalongtheridge-top(•& ½•co@tx)andthento descend into the plain of Eordaea.The distancesare short. The brillianceof Brasidaslay in his speedof decisionandhisquicknessin action.Earlier in the year he hadcomeup to the pass(Thuc. 4.83.2) andlaterhe had gone throughit (Thuc. 4.124.2); he must now have had its featuresin

his mind'seye. He hadthegift whichAlexanderdisplayedin hisreturn to the Pinarus river near Issus.

We may test our conclusionsby consideringthe eventsof 199 Then, in anticipationof a RomanadvancethroughLyncus,Philip V "went aheadto seizethe narrows,sothathis enemywouldnotbe able to win the entrywhichis enclosedin a narrowdefile" (ad occupandas angustias,ne superarehostesartisfaucibus inclusurnaditurnpossent, praecessit(Livy 31.39.7)). He improvisedas fortificationsa rampart, a ditch, piles of stonesand an abattisof tree-trunks,the adjacentarea

134

HAMMOND

AND HATZOPOULOS

beingwooded;and "as he himselfthought,he madea roadby nature difficultimpassable with the additionof theseworksall downthepass" (ibid. 9: ut ipse rebatur, viam suapte natura difficilem obiectisper omnestransitusoperibusinexpugnabilem fecit). "The narrowdefile", so fortified, was clearlythat part of the passwhichbeginssouthof the fields by Klidhi station. Part of the Roman force (ibid. 14-15) formed a testudo,a massof menwith close-packed shields,anddelivereda frontalattack."Another group, makinga shortdetour,cameout on to the ridge of the hill and routedthe panickingMacedoniansfrom their guardandpicketposts." It was evidentlythe turningforceandnot the testudowhichcarriedthe pass;for in a speechattributedto Flamininusby Polybius(18.23.3; echoedby Livy 33.8.5) the Romanswere praisedfor "forcing their way openlyto commandingpositionsand so routingthe Macedonians who were holdingin advancethe high passinto Eordaea"(•&g •g u/IV 'Eo@bctlctv 6xe@[3oXdtg). "The short detour" (brevi circuitu) and the expression"openly" (4x xofi x@oqmvofig) am hardly compatiblewith one another,and it is likely that the detourwas madewith secrecy(the countrywas woodedthen) and that only the final attackwas open. The bestmethodfor escapingobservationwas to proceedfrom Vevi to just north of Klidhi

and from them to follow woods on the east side of the

watershed,until the troopswere ready to attackpicketson the ridge which Brasidashad followed. Sucha routepresentsno difficulties.It would take two to three hours.5

In 199 Be "most of the surroundingregion was wooded"(erant pleraque silvestriacirca (Livy 31.39.10)). Today this is true only of the westernridge; for the regionbetweenthe easternridge and Lake Petreshas only a few trees(theseare mainly fruit-trees),and there are obvioussignsof erosionin the valley-sidesof the torrent-bedswhich run down towardsthe Petresbasin.In antiquitythe narrowpart of the Kirli Dirven pass was deeper and narrowerthan today. Even so it carded a road which was usedby the fleeing Macedoniansand their ox-carts in 423 Be. Plate 2 shows the southern end of the Kifii Dirven pass.

"The heightsonce gained, Brasidasnow proceededmore securely, and the sameday arrivedat Arnisa, the first town in the dominionsof

Perdiccas. The soldiers,enragedat the desertio•n of theMacedonians, ventedtheir rage on all their yokesof oxen which they found on the road, and on any baggagewhich had tumbled off (as might easily happenin the panicof a nightretreat),by unyokingand cuttingdown the cattle and taking the baggagefor themselves"(Thuc. 4.128.3-4, transl.R. Crawley). It is evidentfrom thispassagethatBrasidas'men,

THE

VIA

EGNATIA

I

135

havingcircumvented the pass,returnedto the roadandfollowedit for a sufficientdistanceto overtakeox-teamswhichhadnotbeenintercepted by the Illyrians in the pass.For this to be so the ox-teamsmust have been alreadyeast of the river Amyntaswhen Brasidas'men overtook them, and it follows that Arnisa was well east of that river. The first

site with substantial ancientremainsthereis at Vegora(seepp. 142-3 below). Thus the Macedonianroad which Brasidasand the Macedonian

ox-teamswere using must have gone via Vegora. (The old name of the village was Novigrad. For Brasidas'route, seeMap 1.) II. The Macedonian

road and the identification

of Bokeria

A stadion-stone, datedby its letteringto the fourthor the third century •c, was found in situ at a depthof six feet below the presentsurface in the Kirli Dirven pass at a point half a kilometre southof Klidhi station. This establishedthe existencehere of a major Macedonian road. The stone was inscribed•¾ Box•@ia• ox(•btotkx(xx6v,that is,

"fromBokeriaeighteen anda halfkm".6Asthisnameclearlyunderlies the name"Begorritis"whichLivy 42.53.5 givesto the lake in northern Eordaea(we may compareLychnidusandLake Lychnitis),we should look for Bokeriain Eordaearather than in Lyncus. It has to be the chief town in the vicinity of the lake for its nameto be attachedto the lake, and it hasto be a majorplacebetweenthe Kirli Dirven passand the city of Edessafor it to be the staging-pointin the measuringof the road.

It is certainfrom our visit to Pharangi(see p. 143 below) that the surprisinglylarge cemeterytherewith potteryof this periodbelonged to the most importanttown of the region. Bokeria, then, shouldbe identifiedwith theancientsiteat Pharangi,asindeedhadbeensuggested

by Mrs Hasluck. 7 (ThetownwasthencalledKelemesh.) Thedistance of 18.5km betweenthe findspotin the Kidi Dirven passand Pharangi will not permitof a passage roundthe foot of the presentlake andthen eastof it to Pharangi.The Macedonianroadmusthavepassedthrough what is now a lake in orderto reachPharangi. The presentlevel of the lake is controlledby an artificial tunnel, constructed in 1952. When the settlersof thepresentvillagecamefrom Asia Minor in 1923, they foundthe level higherthanit is today;and then in 1940 it fell lower than it is today, and even then it revealed tree-stumps andvine-stumps,which showedthat the level at sometime before1923hadbeenlowerstill.MargaretHasluckin hertwoexhaustive

studies 8 on thelakehasadduced ampleandconclusive evidence that beforethenineteenth century,whena majorblockageof thelimestone

136

HAMMOND

AND

HATZOPOULOS

swalletsseemsto have occurred,both its depthand its size were considerablysmaller.In particulartheshallowersouthernendof thepresent

lakewasdry groundin antiquity. 9 The distance on thestadion-stone 1o

to Bokeriaconfirmsthatthiswassoin thefourthor thirdcenturyBc. The existenceof a Macedonianroad here is not surprising.There was certainlya road throughthe Demir Kapu to the late fifth-century

Bcsettlement at Manastir,knownonlybyexcavation, • andArchelaus wasbuildingroadsat thattime (Thuc. 2.100.2). Tracesof a Macedonian road which precededthe Via Egnatiain Albania havebeenfoundfor the stretch between Ad Dianam (southeastof Babi6) and In Candavia

(Quk•s).•2Threeroad-markers, inscribed 6@og '[qg66o•, testifyto a Macedonianroadin easternMacedoniain the fourthor the third century BC.13

III. The Via Egnatia from Heracleato Edessa

The entriesof the Itinerariesfor this stretchmay be convenientlytabulatedas follows, with P and D giving varianttraditionsand an asterisk

showing thatanentryis irredeemable: 14 It. Ant. 319.1-320.1

Tab. Peut.

It. Ant. 330.3-7

lt. Burd. 605-6(reversed)

Heraclea-Cellis XXXIIII P XXIIII D XXIII Cellis-Edessa XXVIII

XXXII

XXXIII

*XLV

XXXIII

H. M. G. C.

D XXVIIII

XIII mut. Melitonus XIIII mut. Grande XIIII mans. Cellis XVI mut. Ad Duodeci-

mum D. XII civ. Edissa

That"civitasHeraclea" wassituated atBukovo,2kmsouthof Monas-

tir (Bitola),wasconjectured by Heuzey andhasbeenconfirmed by

excavation. t5Thedistance ofXIII m.p.(19.2km)fromthereto"mutatio

Melitonus" brings ustothecrossing oftheriverKallinikos (Gusachia), wherewaterwasavailable.Thenamemaybe a Latinised formof the

genitive case M•Xixtovog, andif sotheancient name of theriverwas

"Meliton".Thecourse ofthenextstage, XIIII m.p.(20.7km)to"mutatio Grande", is indicated onlyapproximately by a milestone withthree

different fourth-century dedications, whichusedtoliein some ruinsin Sitaria (Rosna). •6Thedistance brings ustotheheadoftheKirliDirven

pass. ThattheRoman roaddidinfactgothrough thatpass hasbeen

confirmed bythediscovery recently of twootherearlyfourth-century milestones at thesouthern endof thepass. •7We tracedthediscoverer of thetwomilestones, avillagerofPetres calledVassilios Hatziconstan-

THE VIA

EGNATIA

I

137

tinou, and we visited the place of discovery:in a ditch below the abandonedpassenger shelteron the loop of the old road, which is east of the modernroad (this now runs straighton a high embankment). This placeis just afterthe modernroademergesfrom the passbetween the monumentto the British dead of 1941 and the votive chapel. (See Map 1.) Returningto "mutatioGrande",we shouldplaceit just southeast of Klidhi station(not ason Map 1) in the flat field, wherewe foundmany sherdsof coarsered pottery. There is plenty of water in the nearby village of Klidhi, or lower down the passon the eastsidewherethere is a spring,if one wantsto keep to the road. The head of the passis obviouslythe right place for a mutatio, becauseit was necessaryto restor changeone'shorsesafterthelongpull up from theplain, whether of Lyncusor of Eordaea.But it wasno placein whichto staythe night. For the wind whistlesthroughthe pass;andthethickwoodsin antiquity gave cover for lions as well as wolves and bears. Yet a mansiowas neededafter someXXVII m.p. from Heraclea.It wasthisconsideration which led Hammondto proposein Mac 49 that the secondentry of XIIII m.p. "mutatioGrande"to "mansioCellis" in the BordeauxItinerary shouldbe emendedto read IIII m.p. The total of 31 m.p. from Heracleato Cellis then becomescompatiblewith the varyingtotalsof 32, 33 and 34 m.p. of the otherItineraries,when we rememberdifferencesin reckoningpartsof a mile (as notedin H, Mac 21). The IIII m.p. (5.9km) from "mutatio Grande" to "mansioCellis" bring us to the findspotof anotherRoman milestone(Plate 3) with dedicationsrangingfrom the beginningof the third to the middle of the fourth century,which is itself 1 m.p. from the two milestoneswe havejust mentioned.It was found uprightand so in situ by a villager namedEvangelosChristopoulos in his field at the placeshownin Map 1. t8 It follows that "mansio Cellis" was near this milestone. An ancient

town is to be expectednear a "mansio", and at a mile's distancefrom the findspotwe have the main site in this region at Gradista(see p.

141below).•9Thereis a springnorthwest of Gradista, anda verystrong flow of water on the southwestside of Petresvillage. The latterwater served Roman baths (p. 141 below). Yet anotherRoman milestone with an early fourth-centuryinscriptionhad beenseenand copiedby

Pappadakis andEdsonin a house-wall atPetres. 2øItspresence indicated that the Via Egnatiapassedsomewherenearby.

For the nextstageof theroad,XVI m.p. (23.7km), "mansioCellis" to "mutatio Ad Duodecimum",three routeshave been suggested:by

Edsonvia Vegora (Novigrad)throughwhat is now a lake to a point

138

HAMMOND

AND HATZOPOULOS

west of Pharangi(Kelemesh)and thenceroundAkroterion(Sut Burun) to the northeastheadof the lake;by Hammondvia Ayios Pandele•mon (Pateli)throughwhatis now a laketo a pointbetweenAmissa(Ostrovo) and Akroterion;and by MacKay via Vegora (Novigrad) and Pyrgoi

(Katranitsa), keepingwellto theeastof thepresent lake.2• It was MacKay's suggestionwhich promptedus to travel alongthe east side of the lake. While we are gratefulfor his initiative, we do not accepthis suggested route, which is far in excessof the mileages recordednot only for the stageCellis to Ad Duodecimumbut alsofor the stageCellis to Edessa.One shouldnot emendbothIt. Ant. andIt.

Burd.to fit a suggested route.22Edson'sroutetooinvolvesa distance to Ad Duodecimumsomewhatin excessof the XVI m.p. of It. Burd. Hammond'srouteis of the right mileagebut it shouldbe discardedin view of new evidence.

On the secondvisit to Pharangithe villagerstold us that there was a tradition of a road and a bridge in the lake between Pharangiand

Vegora. 23The bridge,theysaid,hadcrossed the river Pentavrysos (Nalbankioi)and is visible today. One of the villagerstook us to a point southwestof Pharangito show us the foundationsof "a long wall". This provedto be the line of an ancientroad some3.30m wide, of which the foundationstoneshad beendisruptedby the watersof the lake and, morerecently,whenthe level of the lake was low, by cultivation. (See Plate 4.) The line is roughlyparallelto the modemroad to Maniaki, and it can be tracedintermittentlyfor a distanceof several hundredmetresuntil it disappears into the lake. The line pointsdirectly towhrdsVegora.We conclude,then, thatthe Via Egnatia,as well as the pre-Romanroad, went from Cellis to Pharangivia Vegora. At Pharangithe road no doubt took the natural gateway from which Pharangiderivesitsname,a wide-bottomed gorgethroughthelimestone range.

Somewherein the regionof Vegorathe discoveryof an early fourthcentury Roman milestonewas reportedby the engineerAstima who was engagedin buildingthe railway. His note of its proveniencewas

published by Giannopoulos in 1893.24It runsas follows:"Novigrad. En dehorsde la ligne." This descriptionis so vague that it is of no service Continuing to us.25 from the natural gate by Pharangi we followed a gently

risingroad-•the "old road" as comparedwith the newly maderoad from Edessato Ptolemais--throughdry limestonehollowsand across small plains. On the way we were shownby the Secretaryof Peraia (Kotsana)a placewherehe hadseenthe remainsof a pavedroadsome

THE VIA EGNATIA

I

139

50 or 60cm below the surface,whenhe had beenemployedin making the "old road" in 1960 or so. This place, called"Pegadhia"("Wells") is south-south-eastof Peraia, and one well is still in use. He said that

in this district there is a straightstrip of ground in which the wheat doesnot grow high at all, but is evidently above somepaving or the

like. This is certainlyan appropriate placefor the Via Egnatiato have been.The presenceof wateris importanttoo;for thisis a dry limestone region. The Secretarytook us to an area east of "Pegadhia",where many sherdswereprobablyrelicsof the Turkishvillagewhichexisted beforerefugeesfrom Asia Minor foundedPeraia.From "Pegadhia"the road climbsgentlyup to the headof the Ostrovo-Edessa pass,which is where the railway crossesthe road. Since it was obviousthat "mutatioAd Duodecimum"lay near the

headofthepass, 26wemadeenquiry atthevillageofDhrosia (Dhrouska) and learntwherepotteryand coinshad beenfound. On our third visit we walked from the head of the passeastwards,alongsideand north of the railway, and as we proceededalong a low limestonespur we noted the line of an old road with some large stonesmarking what appearedto be its sides,andwe founda smallamountof pottery.The spurendedin a narrow marshychannel,fed from one's left by water comingfrom the "Zo6dochosPege" (on 24 October 1981, towardsthe end of a long drought).Crossingover the narrowchannelwe traceda pieceof old road, supportedon the lower sideby largestones,which swingsto one's right and passesbetweena field and the foot of the southernend of the Gradistaof Dhrosia, a high limestoneridge with

whitecliffs.27Theplacewherepotteryandcoinshadbeenfoundwas to one's half-left, up, on crossingthe narrowchannel;it was a small area perhaps100m by 150m in a ploughedfield, the highestfield at the foot of the Gradista.We noticedsomepiecesof poor black glaze and somefine red ware, and a shepherddescribedcoinshe had found there, includinga silvertetradrachmof Philip II. There is little doubt that this is the site of "mutatio Ad Duodecimum", being the nearest

watering-point to theheadof thepass. 28It hadnodoubtoperated also for the Macedonian

road.

Turning now to the Itineraries,we startwith an assureddistance, XII m.p. from Edessato Ad Duodecimum;for not only is this given by the BordeauxItinerary, but it is alsoconfirmedby the name, Ad Duodecimum.We have calculatedthe distancealong our proposed route from the Gradistaof Dhrosia, being Ad Duodecimum,to the fields of Petres, where we have located "mansio Cellis", as 3 lkm or

XXI m.p. Thus the total from Edessato Cellis is XXXIII m.p. This

140

HAMMOND

AND

HATZOPOULOS

distance,XXXIII m.p., is givenby the AntonineItinerary 330. But in the BordeauxItinerary the subsequententry is XVI m.p.; and this shouldbe emendedto XXI, so that the two entriesthenaddup to the sametotal as that givenby the AntonineItinerary, XXXIII m.p. The Gradistaitself is a limestonesummitwith a fiat but narrowtop, some120m long, and sheercliffs facingwestwards.Plate 5 showsthe limestonespurwith tracesof an old road, and then on the skylinethe northernend of the Gradista. Other antiquitiesare describedbelow (pp. 143-5). We havenotwalkedoverthe stretchfrom Ad Duodecimumto Agras. It seemsprobablethat the Via lEgnatiafollowedthe samerouteas the Turkish road and the railway as far as point 562 (Plate 6). Between that point and Agrasthe railway was probablyon the line of the Via Egnatia as far as Agras railway station.From the railway stationto "Longos", the site of ancientEdessa,there is a Turkish road along which we have walked. Partsof the old pavingare still visible. This road runstowardsthe Edessarailway station.The Via Egnatiathen, it seems,turnedsouthwards to a pointsome200m eastof the circuit-wall

of Longos. 29It wasat thispointthattheremains of Romanbathswere uncoveredin 1981. The Via Egnatia then turned towardsRizari and proceedednorth of the presentmain road on the line of the Turkish road near the foothills.

An early fourth-centuryRoman milestone,marked A (meaningI m.p.), wasfoundto the eastof Longosat a pointaboutlkm northwest

of Rizari.3 It evidentlymarkedthe firstRomanmilefromlEdessa on the way to Thessalonica.A secondmilestonewith dedicationsof the

sameperiod, TM foundin a yardopposite theGymnasion andnowin the lEdessamuseum,may have markedthe first mile on the Via lEgnatia in the directionof Ad Duodecimum,thatis, to thenorthwestof Longos. For an overall interpretationof the evidenceconcerningthe route of the Via lEgnatiain Lyncus and lEordaea,see Map 2. It is interestingto comparethis part with other partsof the Via

lEgnatia westof Thessalonica. 32Whereascents and descents were gradual,the intervalsbetweenplacesto stay(mansiones andcivitates) andthe numberof placesfor changinghorses(mutationes) are almost uniform.ThusfromThessalonica to Pellais 30 m.p. with twomutationes on the way; from Pella to lEdessa30 m.p. with one rnutatio;from lEdessa to rnansioCellis 33 m.p. with onernutatio;andfrom Cellis to Heraclea31 m.p. with twomutationes. But in themuchmoredifficult countryfrom Heracleato rnansioHis Campis(east of Elbasan)the distancebetweenplacesto stay was more like 18 m.p. (e.g. from

THE

VIA

EGNATIA

I

141

Claudanonto Grandavia, and from Grandavia to His Campis), and there were many more mutationes.We may conclude,then, that in relatively easy parts of the well-pavedVia Egnatia wheeledtransport with changesof horseswas expectedto cover some 30 m.p. in the

course of a day.33Sofasta rateis in keeping withthefactthattheVia Egnatiawas primarilya via militaris, and the baggage-train was ex-

pectedto keepup witha fast-moving infantryunit.TM IV. Evidenceof ancientsettlements in the vicinity of the Via Egnatia

At the entry into the Kirli Dirven passfrom the north there is some evidenceof an ancientshrine.When the railway was beingbuilt, As-

tima35reported thediscovery of a stonewitha Greekinscription at km 172, which is just on the eastsideof the modemroad as it entersthe pass, lkm north of Klidhi station. The inscription mentioned a sanctuary,perhapsof Athena.On the northernoutskirtsof Klidhi village, also at a distanceof lkm from Klidhi station,anotherinscribed

stone alsomentions a sanctuary, presumably thesameone.Thisstone 36 is now built into the inner face of the north wall of a chapel tou Evangelismoutes Theotokou,which overliesa Late Roman building, perhapsa Christianbasilica.The stonecamefrom the remainsof the Late Romanbuilding.Otherevidenceof habitationwasfoundin ancient

tombson the edgeof Klidhi village.37 In the districtof Grammota,shortlybeforeone entersthe passfrom

the south,findsof piecesof statuary--including a Hellenisticstatueof Artemis,whichis nowin theFlorinaMuseum38-•of muchpotteryand many Roman coins, someof which we saw, were made in the fields betweenthe crossroads and the bridge over the Amyntasriver. The main sitein the regionis on the steep-sided hill calledGradista,which is 1.5km west-south-west of Petresvillage. Its flat top, about100m x 50m, is ringedroundthe edgewith the remainsof a small-stone wall, and there are tracesof walls, made either for defenceor for terracing, at lowerlevelsof thehillside.Findsof potteryshowthatit wasinhabited from the Early Iron Age continuouslyinto the Roman period, and we

pickedup a Hellenistic stamped amphora-handle. 39Objectsfromthe Gradismform the bulk of the collectionof local antiquitiesin the school-house of Petres.As one proceedsfrom the Gradistato Petres

villageonepasses onone'slefttheremains ofRoman baths,4ø discovered duringthe buildingof an annexto the school-house, and thesebaths were servedby a greatspateof waterwhichdescends now to a water-

142

HAMMOND

AND

HATZOPOULOS

mill. Fartheron one comesto the old churchof Ayios Nikolaos, which is built on a numberof ancientblocks.There were a numberof grave reliefs and inscribedblocksbuilt into the outer face of the churchwalls, but most of them were taken to Florina Museum (nos. 98, 99 and 103) or to SalonicaMuseum (nos. 189 and 190). A huge block of worked stone was found a few metres to the southeast of the church when

water-pipeswere being laid. The material which has come from Grammota, the Gradista, the

Roman bathsand the churchof Ayios Nikolaos is surprisinglyrich; indeedwithin the fourthmerisof Macedoniait is rivalledonly at Kozani and at Aiane-Kaisareia.The pottery rangesfrom an Early Iron Age sherdwith a hatchedtriangle in matt paint, throughfine Hellenistic greyware andMegarianwarebothred andblack, to coarsered Roman ware. Coins range from the Macedoniankingdom down to the first Byzantineemperors.There are piecesof life-size marblestatuesand a Roman statueof Artemis, also in the Florina Museum (no. 4). Many inscriptionscome from thesesites:fragmentsof funeraryepigramsin

verse, 41othersin prose(oneon a magnificent Hellenistic stele42), an actof manumission, 43a uniquededication to ZeusEleutherios andthe goddess Roma44anda fragmentary milestone whichwehavedescribed. To thesewe may now add anotheract of manumissionanda dedication to a local divinity. There is no doubtthat the main site in Eordaeawas at the Gradista,whichcanbe identifiedwith no otherancientcommunity

than"Cellis",theunique polisofthewholeregion, 45andthatinpeaceful timespart of the populationlived lower down near Petresand in the plain of Grammota. As we havearguedabove,"mansioCellis"of theBordeauxItinerary is to be locatedat Grammota.The name may be explainedas the

ablativecaseof "cellae",meaning"at the granaries",but it is very probablya Latin form of the Greek nameKe•.•.•gor Ke•.•.•ov.The latter

nameoccurson a Greek inscriptionfoundin the plain of the upper Devoll, that is, in an area which was known as westernEordaea, and

this establishes the presumption that therewas sucha place-namein

theoriginalEordaea. 46It is possible thatthenameis preserved in the modem name of the passK[•.•.• A[@•v (K•.•.• A•0•v[ in standard modemGreek), as it is pronounced locally. The villagersunderstand the nameas meaning"the passof the (monk's)cell". At Vegoratherearetwo areasof ancienthabitation.Oneapproaches the so-calledacropolisto the northof thevillageby ascending a steep cleft, oneithersideof whichthereis muchpotteryof theRomanperiod.

The sitewasfirstmentioned by Hasluck(456) andmorerecentlyde-

THE VIA EGNATIA

I

143

scribedby MacKay (205). The otherareais in theplainbetweenVegora and Lakkia, and the plain is known as Tsai?ia tes Ekklesias("Church meadows"). In 1979 a Roman farmhouse in this area was excavated

by the Ephorate,and sherdsand architecturalfragmentscan be found over a squarekilometrein the plain. We visitedboth sitesand alsothe village, wherewe saw in the churchyardof the now demolishedchurch tes Koimeseostes Theotokouan inscribedfunerary altar. Finds from both siteshave been reportedor publishedin local publicationsand in

journals .47Theseincluded eightinscriptions (onebeingonthemilestone mentionedabove).We saw in the village otherfindswhichhadnot yet been reported:an inscribedbase, a fragmentof an inscribedsteleand an uninscribedrelief with an eagle. We weretold of an inscribedmarble arm, and also of two inscribed blocks which had been built into the

side-entranceof the churchof St George at Lakkia and are no longer visible. The ancient site which we have identified with Bokeria is some 2km

north-eastof Pharangi(Kelemesh). It lies between a limestonebluff andtoday'slake-shoreand is known as "Paliokhori."We visitedit with Mrs Siganidou,Ephorof WesternMacedonia,andtheepimeletriaMiss Gounaropoulou,who gave us mostgeneroushelp. The gently sloping groundof the siteis coveredwith many piles of weatheredstones,both largeand small;it seemsprobablethat someare the remnantsof pieces

of wall some3m wide,butit is impossible to makeoutanyplanof defensivewalling. Betweenthe site and the village of Pharangithere is a surprisinglyextensivecemetery.The site has yielded a headless statueof Artemis(now in the Museumof Thessaloniki)andpotteryof the Early Iron Age, theClassicalperiod(includinga Red-Figurecrater, now in the museumof Kozani), the Hellenisticperiodand the Roman

period. 48Someterracotta water-pipes of a Romanshapewerefound by the villagerswhenthey were laying a water-conduitfrom Pyrgoito Pharangi.Thesecasualfinds would seemto indicatethat Bokeriawas more importantthan the site acrossthe lake at Ayios PandeleEmon, whichhasso far yieldedonly undistinguished potteryof the Hellenistic

andRomanperiods. 49 There are two placesnear Dhrosiawhichhave yieldedancientremains, the lower slopeof the Gradistaand the moundof Saratsina which is just southof the railway line. We have alreadydescribed potteryand otherobjectsfrom the formersite, and we were told by the agrophylakasthat a funeraryrelief of a man and a woman, with an inscriptionallegedlyin Latin, which is now lost, had been found

there.Thisis alsothefindspot of a bronzeoenochoE andof animportant

List of Plates

1.

The headof the Kirli Dirven pass,lookingsouth,with the hill ascendedby Brasidasand(ontheright)thenarrowsof thepass

2.

The southern endof theKirli Dirvenpass,seenfroma lowerslope of theridgefollowedby Brasidas

3.

TheRomanmilestonefoundin theplainsouthof Petres

.

Remains(in the immediateforeground)of the foundationsof an ancientroad southwestof Pharangi,lookingtowardsVegoraat the southern end of Lake Petres

o

Lookingsouthwestfrom the Gradistaof Dhrosiaby mansioAd Duodecirnurn to the headof the passbetweenEdessaand Lake Vegoritis:thearrowpointsto thenearendof an ancientroadwhich runsto theheadof thepass;therailwayline is to theleft of theancient road

6.

TheoldroadfromEdessato Agras

(Plates1, 3, 4, 6 fromphotographs by M.B. Hatzopoulos; Plates2 and 5 from photographs by MargaretHammond)

Plate 1

Plate 2

Plate 3

Plate 5

Plate 6

THE VIA

EGNATIA

I

145

inscribed sherddatingfromthe Hellenistic period. søAn interesting inscriptionis associated with the moundat Saratsina.It was seenorig-

inallyby Pappadakis 51at thatplaceandlater,in 1937,byEdson,who tookphotographs andmadesqueezes. Edsondescribed theexactlocation as "100-150 metersto the east of the 130 kilometer mark" (i.e., of the

railwayline).52 Theinscription inGreek,nowlost,wascutona funerary altarof a Macedoniantype andrecordedthedate^D 203/4. We conclude

thattherewasonlya smallsettlement in ancient timesat the•place which becamethe mutatioAd Duodecimumon the Via Egnatia.• Clare College, Cambridge

N.G.L.

National Hellenic Research Foundation

Hammond

M.B. Hatzopoulos

Athens NOTES *Abbreviations Adams

= J.P. Adams,"Polybius,PlinyandtheVia Egnatia"in W.L. AdamsandE.N. Borza(edd.), PhilipII, Alexanderthe Great and theMacedonianheritage(1982) 269-302

Demitsas

= M. Demitsas, •H Maxebovia •v k•OOLg q)Oœ¾yOp.•¾OLg xGtL p•w}p•elotg oco•op•votg (1896) = C.F. Edson, "The locationof Cellae andthe route of the Via Egnatiain WesternMacedonia",CP 46 ( 1951) 1- 16 = Note Books,IVth p•e@ig

Edson

Edson, NB

Giannopoulos = N.I. Giannopoulos, "Nouvelleset Correspondance", BCH17 (1893) 634-35 Hasluck

= M. Hasluck,"The archaeological historyof LakeOstrovoin West Macedonia",GJ 88 (1936) 451-56 Hammond, JRS = N.G.L. Hammond,"Thewestern partof theVia Egnatia", JRS64 (1974) 186-94 H,Mac HG, Mac

= N.G.L. Hammond,A historyofMacedoniaI (1972) = N.G.L. HammondandG.T. Griffith, A historyofMace-

MacKay Mordtmann

= P. MacKay, "Therouteof theVia EgnatiaaroundLake Ostrovo",AncientMacedoniaI ( 1971) 201- 10 = J.H. Mordtmann,"InschriftenausEdessa",MDAI(A) 18

Pappadakis

= N.G. Pappadakis, '"Ex xqg •AvcoMctxebovlctg", Athena

Picard

= Ch. Picard,"Lesrecherches arch6ologiques del'arm6e fran•aiseen Mac6doine1916-1919", BSA23 ( 1918-

donia II (1979)

(1893)415-19 25 (1913) 430-62

1919) 1-9

146

HAMMOND

AND

HATZOPOULOS

[Editorial note: The authors'transliterations of Greek have been kept as written. ]

1. A.W. Gomme, A historical commentaryon ThucydidesIII (1956) 617, toyedwith a preferencefor the northernroute, unwiselyas he had neverbeen there. See Edson4 with n. 29 and H, Mac 52 and 104 n. 1. Combesare dry depressions withoutany streams,often foundin the limestoneformationsof Epims and Macedoniaand muchusedfor roads. 2. They aimedto destroyhismenwith missilesandrocks,asthe Athenians had destroyedthe Spartanson Sphacteria(Thuc. 4.34). 3. Poppoemended•t6v•c•g to •:•6v•c•g. 4. This supersedes the accountin H, Mac 104 f., writtenbeforeHammond walked over the ground. 5. We walkedfrom Brasidas'ridge to Petresin one hour. 6. Mordtmann419, Demitsas393-94, no. 4, and amongmore recentdiscussionsHasluck454-55, Edson4, Ph. Petsas,"Box•@@tog xct•Boxe@•ct", AAA 4 (1971) 115-17, and H, Mac 51-52 and 56-57. 7. Hasluck 454.

8. M. Hasluck, "A historical sketch of the fluctuations of Lake Ostrovo

in West Macedonia",GJ 87 (1936) 338-47; ead., "The archaeologicalhistory of Lake Ostrovo in West Macedonia", GJ 88 (1936) 451-56; cf. E. Fels, "Die westmakedonischenSeen in Griechenland", Die Erde 6 (1954) 316-35.

9. For fluctuations,besidesthe articlescited in the previousnotes, see also H, Mac 52-53.

10. Whethervia Pateli or via Vegora, as Hasluck455 pointedout. 11. See H, Mac 174 and HG, Mac II 140 with references. 12. See Hammond, JRS 192. 13. Edson 11-12 and Adams 269-78.

14. A full discussionof the evidenceof the Itinerariesfor the Via Egnatia from the Adriatic coast to Thessalonica will be found in H, Mac 19-142 with

mapsand in Hammond,JRS 192-93. One of the earliestmodemdescriptions of the Kirli Dirven passand of the areaaroundthe lakesPettonandBegorritis can be found in H. Barth's Reise durch das Innere der europiiischenTiirkei (1864) 155-59.

15. L.A. Heuzey and L. Daumet, Mission archdologiquede Macddoine

(1876)300;Fanoula, Papazoglou, "H6rac16e et P61agonie" (in Serbian withan abstractin French) ZA 4 (1954) 308-45.

16. Pappadakis 437, no. 21; Edson4-7. This inscriptionhad alreadybeen publishedin a veryconfused andconfusing mannerby Demitsas268-69, nos. 243-44; cf. CIL III (Suppl.) 14206;Picard7; P. Collart, "Une r6fectionde la 'Via Egnatia' sousTrajan", BCH 59 (1935) 402. 17. The two milestoneshave beentransportedto the Florina Museumand they will be included,alongwith all the new epigraphicmaterialmentioned in this article, in the new Corpusof the inscriptionsof the Greek fourthmeris whichDr A. Rizakisand Dr I. Touratsoglou are preparing.

THE

VIA

EGNATIA

I

147

18. The milestonehas beentransportedto the Florina Museum(no. 265). It was the subjectof a communicationby M.B. Hatzopoulosat the VIIIth InternationalCongressof Greek and Roman Epigraphy(Athens, 3-9 October 1982). It will be publishedin a joint monographby L. Gounaropoulouand M.B. Hatzopoulos,entitledLesMilliaires de la VoieEgnatienneentreHdraclde desLyncesteset Thessalonique. 19. Edson (9) "would place Cellae in the general vicinity of Novigrad or Sorovichin the plain southof Lake Petres"and MacKay (206) "at Novigrad (or the adjacentsite in the fields)". 20. Pappadakis 432, no. 8; seenby Edson7; we couldnotfind it, the house having been demolishedand the stonelost. 21. Edson 10; H, Mac 50 and 54; MacKay 206. 22. MacKay 209 doesnot suggestany emendation.He simply maintains that the figuresmustbe "stretched",that is on the rack of theory. It is better to regardthe Itinerary figures as the basicevidence. 23. Cf. G. Azes, 'H A•ti•t T&v&•X•t•cov M•tx•66vcov,Aristoteles10 (1958) 39, andTh. Vrakas, Tov•toTtxo•xe•ixu•ot e[g•6v vo½6v•Xco•vrlg, Aristoteles 16 (1959) 14 n. 1. 24. In "Nouvelleset Correspondance", BCH 17 (1893) 635. 25. The entryNovigradcouldhavereferredat that time to any of the land up to the outskirtsof Amyntaionand of Ayios Pandelefmon;we learntthis from the early Frenchmap andin conversation with the priestof Vegora. 26. As in H, Mac 51, thoughMacKay 210 proposedto placeit at the hill of Peraia, some two Roman miles farther on, which would have involved

emendingnotonlytheentryin It. Bur&, but alsothename"Ad Duodecimum" to "Ad Quartumdecimum".

27. It is possiblethatthe discoveryof this stretchof the Via Egnatiadates from the nineteenthcentury;seeJ.H.F. Tozer, Researchesin the highlandsof TurkeyII (1869) 369 n. 11; "Mr. Curtis, of Constantinople,informsme that he discoveredlast year, not far from the summitof the passbetweenVodena and Ostrovo, the remains of a Roman road, which was traceable for some distance in the direction of the head of the lake."

28. See also A. Struck, "Die makedonischenSeen", Globus 83 (1903) 219.

Some water is now broughtacrossthe headof the pass,but by an artificial conduit.

29. This is the coursesuggested by P. Theodorides,authorof an important but unpublishedthesison the Via Egnatia (Thessaloniki1973), who kindly put his local knowledgeat our disposal. 30. Edessa Museum no. 220; J.M. Cormack, "Inscriptionsfrom Pella, Edessa and Beroea", APF 22 (1973) 208-9; cf. BCH 76 (1952) 228; JHS 72 (1952) 102; Makedonika7 (1966) 308. Rizari is separatefrom Rizo and not the same, as H, Mac 164 n. 3 had supposed. 31.

Edessa Museum

no. 228.

32. For theseseeH, Mac 19 f. with maps2-8 and 14, andHammond,JRS 192-93.

148

HAMMOND

AND

HATZOPOULOS

33. AlthoughAdams(277 n. 27) observedthat the usualdistancebetween mansioneswas 30 to 33 m.p., he failed to apply that observationto his own suggestion that"mansioClodiana"shouldbe placednot whereHammond(JRS 188) put it, between30 and 31 m.p. from Dyrrachium,but at a point43 m.p. from Dyrrachium(his p. 289). 34. One is thinkingof small units, not of encampingarmies;it is assumed that mountedcourierswith changesof horseswould cover much greaterdistances between mansiones.See Hammond, Alexander the Great (1980) 310 n. 33 for examples. 35. Giannopoulos 635; cf. W. Baege,De Macedonurnsacris(1913) 32. 36. Seenby M.B. Hatzopoulosin 1981. 37. As reportedby the villagersof Vevi. 38. Found,in 1980. For remainsin the region of Petressee Pappadakis 430-32; A.D. Keramopoullos,'"Avaoxatptxa[•@•vvctt•v •,q 'Avco Max•6ovi.a",Ephemeris1932, 97-99; id., "•E@evvat •v Avxtx.qMaxebovi(3",Praktika 1933, 60-63; id., '"Avaoxatpa[xa[ •@evvat•v Maxebovi(3",Praktika 1934, 84-85; Chroniquedes Fouilles 1956, BCH 81 (1957) 600; Th. Vrakas, '"A@9(att'ctl'Ceg 'rofivogoficl))•co@(•Vtlg", Aristoteles35-36 (1961-1962) 86-88. 39.

Now

at the Florina

Museum.

40. Its position reminded one of the Roman baths and watercourseat Bradasheshwhich was identified with mutatio ad Quinturnby Hammond, JRS 188 with Plate IX.

41. Pappadakis432, no. 5 and no. 6; cf. Edson,NB 296 and 297. 42. Pappadakis 431, no. 2; cf., amongmorerecentpublications,G. Daux, "Notes de lecture", BCH 101 (1977) 347-48.

43. Pappadakis 432, no. 7; Edson,NB 306. 44. Edson, "Macedonica", HSCP 51 (1940) 131 and 134; cf. Edson, NB,

and, amongmorerecentpublications, G. Daux, "Notesde lecture",BCH 101 (1977) 348; 350 andfig. 6, with commentsby J. andL. Robert,Bull. dpigr. 1978, no. 281.

45. Eordaea,citedsometimes asa city, isin facta region,cf. H, Mac 109. 46. H, Mac 57.

47. Cf. the studiesof Hasluckand MacKay mentionedabove, and alsoG.

Azes, "•H Ae[3et(•et x•v &@•etlcov Max•btvcov",Aristoteles10 (1958) 22-52; Th. Vrakas, "Tov@tox[xo[ J•e@•xaxot e[gxbvvolxtv•kco@•Vtlg", Aristoteles16 (1959) 13-14; id., "'A@xattxtlXegToOvo[to6•),.co@iVtlg", Aristoteles35-36 (1961-1962) 85-87; Chroniquedes Fouilles 1959, BCH 84 (1960) 767 and Plate6 (p. 766);Ph.Petsas, "X@ovtx& 'A@xato•.oytx• 1959-1965",Makedonika 7 (1967) no. 264; id., "X@ovtx& 'A@9(ato)•oytx6 1968-1970",Makedonika15 (1975) nos. 193, 197a, with Plates 221a, 221b.

48. The findsare reportedin a publicationof the Nomarchiaof Kozani: Fvco@t[da [t/•xbvvo[tbvKo•dvtlg(1970) 164f.; seealsoH, Mac 344 andDeltion 17 (1961-62) 2.216.

49. The oppositewasthe casein the Late BronzeAge andthe Early Iron Age, astheexcavation of AyiosPandele•mon (Pateli)hasshown.SeeH, Mac 340-44.

THE VIA

EGNATIA

I

149

50. Ph. Petsas,"X@ovtx&'A@xcttokoytxd 1966-67", Makedonika9 (1969)

116fig. 16 andPlate83; cf. J. andL. Robert,Bull. dPisr.1970,no. 362;F. Papazoglou,"Inscriptionhell•nistiquede Lyncestide",ZA 20 (1970) 109 and 111 fig. 6. 51. Pappadakis435, no. 17. 52. Edson, NB, no. 303.

53. We wish to expressour gratitudefor help generouslygiven by the following:Mrs M. Siganidou,Miss L. Gounaropoulou andothermembersof the Ephorate,who often accompaniedus in our visits, contributingto the discoveryof importantnew evidence;C. Habicht and C.F. Edson, and the Institutefor AdvancedStudyat Princeton,for permission to useEdson'smaterial; Mrs A. Tataki, Miss A. Panayiotou,Dr A. Rizakisand Dr I. Touratsoglou for sharinginformationobtainedduring a joint expeditionwith M.B. Hatzopoulos;Dr J.P. Adams who sent us copies of the relevant sheetsof the Frenchmapsof theregionandgaveusa previewof his article;andP. MacKay for his usefulcommentaryand suggestions,which helpedimprovethis paper. Photographs are by MargaretHammondandM.B. Hatzopoulos.N.G.L. Hammond expresseshis gratitudeto the British Academy for a ResearchGrant. The OrientationMap wasdrawnby David Cox; theothertwo mapsweredrawn byAntoniaKotsoniandexecutedby EkdotikeAthenonandEkdotikeHellados.

SULLA

AUGUR

Coins and Curiate

Law

Someyearsago, therelatedquestions of whenSullabecamean augur andhow we are to interpretthe auguralsymbolson the coinshe issued at some time in the period 84 to 82 were debatedby E. BadJanand B.W. Frier.• In essence,I believe that BadJanwas correctto date Sulla'sassumption of theaugurateto afterNovember1st,82. However, it would seemthat not all of the argumentshe advancedare of equal merit and a fresh examinationof the sourcesmay help to put his conclusionson a sounderbasis.This datinghas, of course,important consequences for the interpretationof the symbolsand, in the second part of thispaper,I shalltry to setforth what I believeSulla may have beentrying to sayby meansof them. AlthoughI do not acceptFrier's main thesis, it will, I hope, becomeclear that he was, nevertheless, right to seethe symbolismas havinga directreferenceto the contemporary political situation. I. The date at which Sulla becamean augur

Cluesas to the dateat whichSullabecamean augurhavebeensought in a numberof places.I proposeto attemptto demonstrate here that only three sources,ILS 9338, Asconius21C and SuetoniusNero 2.1, are, in fact, of real value in helpingto solvethis problem. It is safeto assumethat Suetoniusde Gramm. 12 may be discarded immediatelysinceit merelymentionsthe auguratebut givesno infor-

mationastothedateof assumption. 2A coinissued bySulla'sgrandson, Q. PompeiusRufus, in 54 showswhat appearto be symbolsof Sulla's augurate together with the decemviral emblems of Sulla's consular

colleague in 88, Q. Pompeius Rufus.3 Frier2.188 tookthisto mean that the coin was proclaimingthat one of the consulsof 88 was an augur then, while the other was a decemvirsacrisfaciundis. This is not necessarilyso. In puttingthesesymbolson his coin Pompeiuswas following a commonRomanpractice.He wishedto draw attentionto some achievement

or distinction of his ancestors in order to shed lustre

on himself and so advance his own career.4 Given that such was his 150

SULLA

AUGUR

151

purpose,it doesnot necessarily followthatthe coincanbe takenasa statementof a given circumstance at a given time. Thus, while the piececouldconceivablyreflectthe positionas it was in 88, we have

nomeans of sayingif it actually doesso.5 Accordingto Appian (B.C. 1.79), Sulla, in 84, complainedto a senatorialdelegationthat he had beendeprivedof a priesthood,which, self-evidently,he had assumedbefore87. Couldthisunspecifiedpriest-

hoodbetheaugurate? BadianandFriercouldnotagreeonthispoint.6 In fact, it is impossibleto tell. Much of the debate hinged on the question:could (or would) Sulla, as an exul, be allowed to retain an

augurate? However,Sullawasnotanexulbuta hostis. 7Essentially an exul is someonewho, usuallybecauseof a capital sentence,renounces his Roman citizenship.He then becomesa citizen of someallied community. From thenon, he is, in Romaneyes, a memberof this friendly communityandis not (nor ever was) a hostis.This meansthat, though barredfrom Rome itself by reasonof his condemnation,he may move freely aboutthe empire and maintainfriendly relationswith his former

home. 8 In thesecircumstances, it is possible to envisage at leastthat a manmightbe allowedto retainan augurate. 9 On theotherhand,a hostis, by definition, is somebodywho has no relations whatsoever with Rome. He is a totally alien beingwho may be killed on sight.If, for any reason, it was necessaryto have parley with him, then this musttake place outsidethe walls. We can understandthen why Cicero remarks that somebodywho was a hostiscould not fulfil any of the functions of a Roman citizen. l0 It thus follows that we have no means

of deducingwhat priesthoodis in questionhere sinceargumentsbased on peculiarconditionsof tenureof an augurateduring exilium do not apply to the case in question. For similar reasonsPlutarch Sulla 35.2-3 is also unhelpful. This

narrates anincident whichoccurred in 81 duringSulla'sdictatorship. 11 Metella fell ill and the priestsforbade Sulla to approachher or have his housepollutedby her corpse.Sulla immediatelydivorcedher and causedher to be removedfrom the premiseswhile still alive. Nothing here hasan especialapplicationto the augurate.Since, as we saw from Appian, he did have some priesthoodor other to which he would by now be restored,he wouldbe boundby therule whichforbademembers

of all the priestlycollegesadtrectare feralia.•2 So far as we know divorcewasnotactuallyrequiredin thesecircumstances, •3sowe may see in this action anotherexampleof Sulla's conspicuous piety which

Plutarchemphasises throughout theLife.J4The dictator,in an excess of zeal, did more than was askedof him.15 We may now turn to thosesourceswhich, as I claimed at the outset,

152

ARTHUR

KEAVENEY

can alone establishwhen Sulla becamean augur. An inscription(ILS 9338), containinga list of priests, tells us that when M. Aemilius Scaurusdied in 89 he was succeededin a priesthoodby L. Cornelius ScipioAsiagenus.Unfortunately,we have no meansof deducingfrom

theinscription itselfif thispriesthood wastheaugurate. •6If,however, it couldbe shownthat Scauruswas an augur,then it would follow that Sulla could not belongto that collegiumfor as long as Scipio held office, sincetwo menbelongingto the samegenscouldnot be members at the same time. •7

Two literary sources,Asconius21C and SuetoniusNero 2.1, enable us to showthat this was, in fact, so. Asconius21C says:

Cn. Domitius,qui consulfuit cumC. Cassio,cumessettribunus plebis, iratusScauroquodeum in augurumcollegiumnon cooptaverat, diem ei dixit apud populum et multam irrogavit, quod eius opera sacrapopuli Romani deminutaessediceret. Crimini dabat sacrapublicapopuli Romani deum Penatiumquae Lavini fierentoperaeiusminusrectecasteque fieri. Quocrimineabsolutus est Scaurusquidem, sed ita ut a tribus tribubusdamnaretur,a xxxII absolveretur,et in his paucapunctainter damnationemet absolutionem

interessent.

While in Suetonius Nero 2.1, we read:

AtavuseiusCn. Domitiusin tribunatupontificibus offensior,quod alium quam se in patris sui locum cooptassent,ius sacerdotum subrogandorum a collegiisad populumtranstulit. Scholars,for the mostpart, appearto haveassumedthatbothwriters are speakingof the same incident and have, in consequence,tried to determine

if the true version of events is to be found in Asconius or

Suetonius. t8 This I believe to be a mistaken view. Both authors are

offering us historicallyplausibleaccountsof two completelyseparate incidentsand there is no real reasonto supposethat any conflationor confusionhas occurred.Asconiustells us how Domitius, angeredby Scaurus'failure to co-opthim to the collegeof augurs,accusedhim aloneon the groundsof someact of commission (or omission?) during religiousceremonies at Lavinium.In contrast,Suetoniusspeaksof a law the tribuneintroducedin orderto revengehimself on the whole collegeof pontiffsfor a refusalto admit him to his father'splace. Neitherauthoris writing a completehistoryof Domitius'tribunate,but

SULLA

AUGUR

153

eachis pickingfrom it an incidentto illustratethe particularpoint he wishesto make and, thus, we have no warrantfor assumingthat they wouldnecessarily bothchoosethe sameincident.It is truethat a wellknown error about Domitius

occurs in the next sentence of Suetonius

but, in the absenceof corroborating evidence,theexistenceof a mistake in one placecanhardlybe saidto justify our postulatinganotherelsewhere.Althoughit hasbeenattacked,Asconius'descriptionof Scaurus

asan auguris perfectlyacceptable. •9By the sametoken,Suetonius, ashasbeenpointedout,2øwouldseemto be quitecorrectin assuming Domitius' father was a pontiff. Moreover, sincea tribunatelasts for a year, there was ample time for Domitiusto bring two prosecutions. Indeed, he may have broughtthree in all, for we also hear of his accusingSilanus(cos. 109) of illegally beginninga war with the Cimbri. 2•

Nor is there anythinguntowardin Domitius' seekingadmissionto two differentpriestlycolleges.It wascommonpracticefor a manwhose fatherwas a memberof somepriestlycollegeto try andjoin another,

oftenwiththeparent'shelp.22Bearingthisin mind,we canblendthe two storiesandoffer a plausiblereconstruction of whathappened.While his father,the pontiff, was still alive Domitiustried to becomean augur but, allegedlybecauseof themaligninfluenceof Scaurus,wasrebuffed. Then, when his fatherdied, he tried to take his place in the pontifical college,only to suffera secondrebuff. He did not havelong to wait for his pontificate,however.Thanksto the changein procedurehe had broughtabout,he appearsto haveenteredonhiswell-attested pontificate

in theverynextyear,afterthedeathof MetellusDelmaticus. 23 So, by arguingfrom what the sourcesactuallytell us ratherthan approachingthem with a view to fitting them into an implausiblea priori hypothesis,we may justifiably claim to have establishedthat Scauruswas an augur. And when he died, Scipio succeededhim in that capacityin 89, and, from thenuntil the lattervacatedhis position, Sulla couldnot becomea memberof the college. Scipio, in fact, continuedto be an auguruntil 82. In that year, he was, however,declareda publicenemy.As suchit mustbe conceded he was treatedvery leniently. Thoughhis name figuredprominently on Sulla's first proscriptionlist, he was allowed to live quietly in Massilia.24The reasonfor this concession,rare in that bloody year, is to be foundin Sulla's own character.Never one to forget a favour, he remembered how Scipiohadshownhimselfwillingto negotiateduring thelatewar andsoforborefrom takingextrememeasures againsthim.25 But this indulgenceshouldnot blind us to the cardinalfact that Scipio

154

ARTHUR

KEAVENEY

was now a hostis and we have seen what this means: he could not fulfil

anyof the functionsof a Romancitizen.Therefore,he, like Sullaa few yearsbefore,wasmadeto forfeithispriesthood. Into thegapthuscreated in the collegiumSulla now stepped. 26 After examiningall of the relevantevidence,we mustconcludethat

Sulladid notbecomean auguruntil afterNovember1st,82.2? II. The meaningof the symbolson Sulla's coins

But, if Sulla did not becomean auguruntil after the Civil War had ended,why did heputauguralsymbolsona coinwhich,it is universally agreed,wasissuedbeforethatwar ended?This is the questionwe must now attempt to answer.

Therehasbeensomedisagreement asto wherethecoinswereactually

minted. 28However,fromanhistorical viewpoint,it is safeto conclude they were intendedfor circulationin Italy. While he was in Greece, Sulla hadno reasonto circulatesuchcoinseitheron economicgrounds or from motivesof self-advertisement. There was no pressingurgency to pay his troops.On the contrary,theywerein a positionto offer him money.With their pocketsstuffedwith the loot of Asia they were able to offer Sulla a loan before embarkingfor Italy and, once they had arrived, their exemplarybehaviourmay, in part, be attributedto the

factthattheywerealreadysatedwithplunder. 29On theotherhand, coin would be needed after Sulla landed in Italy. Those who now flockedto join him would need'supplies which would haveto be paid for. In additionwe may alsonotethat briberywas one of the methods usedto draw Italian waverersto Sulla's side and money in abundance

wouldbe required for that.30Furthermore, although scholars maydisagreeasto how the auguralsymbolsareto be interpreted,nobodywill, I think, deny that a Roman (or Italian) audiencewas envisagedfor

them.Theywouldbetotallyunintelligible to a Greekaudience. 3•Since they had nothing to say to the Greeks, there was thus no reasonto circulate them in Greece.32

But what was Sulla actually trying to say to this Italian audience?

BadJan herebasedhisviewson a theoryof A. Alf61di's. 33Thelatter, in a discussionof the reversetypesof republicandenarii, cameto the conclusionthat types which celebratethe achievements of ancestors were succeeded, only in theperiodafter Sulla, by typeswhichglorified

thepersonal exploits ofthemoneyer. 34SoBadianrejected Frier'snotion that the symbolsin questioncouldrefer to Sulla himselfand thought theymightbe a referenceto anaugurateheldby oneof Sulla'sancestors.

SULLA

AUGUR

155

But objectionscan be raisedto this view. For instance,our (admittedly skimpy) recordshave no mentionof a previousauguratein Sulla's family. More important,recentresearchseemsto showthat references to contemporaryeventshad begun to appearon Roman coins before Sulla issuedhis. Moneyershad alreadyproclaimedthe great victories of Marius. Again Roman successes in the Social War are alludedto in the coinageof the period and, in 87, the hopesof the governmentin

the war againstCinnafindexpression on piecesin thatyear.35Since this kind of precedenthad been set it is obviousthat sooneror later someonewould have sufficiently strongmotivesfor taking mattersa little further and actually proclaiminghis own exploitson a coin. A consideration of the historicalbackgroundand the somewhatpeculiar circumstances underwhich Sulla producedhis moneymakesit at least likely that it was he who took this step. Sulla's coinagewas undoubtedlyinnovatoryin many ways. It proclaimsan origin independentof the Romanmint, it assertsthe right of a Roman commander to mint in the field, and it includes an aureus,

thefirstgoldontheRomanstandard since209.36But, aboveall, these coins were minted by a hostiswho had just broughtto a successful conclusion,on Rome'sbehalf, a greatwar againstMithridates.Plainly, the circumstances which led to this extraordinarysituationmerit further examination.They will, I believe, justify the notion that the symbols on Sulla's coinshave a contemporaryrelevance. The yearof Sulla'sconsulship,88, generatedcontroversy in his own day and has continuedto do so since. His march on Rome sharply divided public opinion. He and his followers saw nothing in the act which was not compatiblewith his positionas consul.Otherswere of a different mind, however, and some went so far, once his year of office was up, as to try and have him deprivedof his commandand broughtback to Rome to face trial for what he had done. Thus were born two views of Sulla's positionwhich were to persistuntil his final victory in 82. He, and his supporters,held that, as a consul,he had at all timesactedcorrectly.His enemiesbelievedhe had committedcriminal acts during his tenure of office. Which view is correctis not of great importancehere. What is of consequence is that both sidesseem to have beenconvincedthat right lay with them and this decidedtheir subsequent actions. Those who tried to strip Sulla of his imperiumearly in 87, acting logically accordingto their beliefs, declaredSulla a hostislater that same year. They denied him suppliesand sent out anotherarmy to replacehim. Later, Carbo wasto declareall of Sulla'sfollowershostes.

156

ARTHUR

KEAVENEY

Sulla, for hispart, continuedto regardhimselfasa properlyconstituted Romanproconsul.Earlyin 87 hedepartedto fightMithridates.Ignoring the fact that he hadbeenmadea publicenemy,he prosecuted the war with the utmostvigour, at the end conductedthe peacenegotiations with the enemy and made provisionsfor the better orderingof Asia. To furtheremphasizethe pointhe attackedbothFlaccusandFimbria, whom the senatehad designatedas his successors, on the groundsthat they were illegally intrudingin his provincia. To crown all of this, as it were, he then senthome, as any proconsulwould, a completereport on what he had done. Receivingno reply, he wrote oncemore, this time a personalletter, recountingall he had done for Rome from the timeof theJugurthine warupto theMithridaticconflict,layingparticular stresson his exploitsin the latter. His reward,he complained,wasto be made a hostisand have his propertydestroyed.He promisedhe would soonreturn to wreak vengeanceon thoseresponsiblefor this stateof affairs. He did, however, make it clear that his quarrelwas with the Cinnansalone. The rest of the citizensand the newly enfranchisedItalians had nothingto fear from him. The senatein alarm sent ambassadors to negotiatewith him about his positionand for a time there was a possibilityof a settlement.Eventually,however,the talks failed and Sulla then made war on the Cinnansin orderto gain what

hefelt wasrightfullyhis.37 One factemergesclearlyfrom all of this:Sulla'sstatuswasthe great issueof the day. He consideredhimselfto be still a Romanproconsul and he intendedto force thoseat Rome to recognizethe fact. If this

couldbe donepeacefully,somuchthe better.If not, he wasprepared to go to war over the issue. Given then that Sulla wishedto establishthe justice of his claim in the face of considerable oppositionandwaspreparedto advanceit by meansof letters,ambassadors, negotiations andultimatelya civil war, it seemshighly unlikelyhe wouldneglectto do so by meansof coins as well. Consideringthe situationin whichhe now foundhimselfand the objectiveshe was strivingto realise,it is difficult to see why he would choose to advertise an ancestor's achievement

rather than his

own. His family had beensunkin obscurityfor generations.What one of his ancestors(even the famous P. Cornelius Rufus) could claim

exploitsto matchhis?What needhad he to recall one of their meagre glories?What purposewould sucha reminderserve?Was it not more naturalthathe, thegreatest representative of hisfamily,shouldadvertise hisown claimto be a legitimateproconsul,something he himselfnever doubted,by puttingon his coins, both denarii and aurei, symbols illustratingthatproconsulate andthe brilliantvictorieswon duringit?

SULLA

AUGUR

157

And, knowing their background,we can now see alsojust who in Italy was meantto seethe coins. First of all, therewas the senate.It will be recalledthat Sulla had said he had no quarrelwith the Roman people as such. This, of course,also meant that he bore no ill-will towardsmany of the senators,ratherthe opposite.He recognizedthat many of them had done their best to composethe quarrel between himselfandCarboand, afterhe had arrivedin Italy, theycameto join him in his camp.Thesemen, therefore,cansurelybe numberedamong thoseat whom the coin would be directed.They were to be given yet anotherstatementof what Sullabelievedto be hisposition.At the same time Sulla intendednot only that the coinsshouldencouragebut that they shoulddemoraliseas well. As we shall see, his symbolismcontainedan impliedthreatwhichis preciselywhatonewouldexpectfrom somebodywho hada deadlyfacility for probingan enemy'spsychological weaknessand unsettlingthe opposition.Those who electedto remain with Carbowere to have a furtherreminderof the proconsul's pretensionsand his determinationto realisethem. Contemplationof thesefactsmightalsohelpwaverersto makeup theirminds,andindeed desertions did continue for much of the war.

It might be askedhow the senatewould learn aboutSulla's coins, since Rome was in the heart of enemy country.We shall have no difficulty in envisagingthis if we bear in mind an observationmade by W.V. Harris, Rome in Etruria and Umbria (1971) 256. He points out that the areaswhich supportedSulla and thosewhich backedhis enemiesare not to be regardedas hermeticallysealedcompartments. Given thattherewastraffic betweenthetwo, it is not strainingcredulity to suggestthat somesamplesat leastof Sulla'shandiworkfoundtheir way into enemyterritory,thereto be displayedas yet anotherexample of the proconsul'sarrogance.And we may also claim that an even larger audience,which did not actuallyseethe coins,will have heard somethingof their contents.I list below someof the possibleways in which Sulla's coinscould have found their way into Cinnan, and ultimately into senatorialhands. (a) Both Cinnansand Sullansfuriouslylobbiedthe italians in the winter of 83 and we are told that money changedhands(App. B.C. 1.86). Someplacesseemto have beenvisitedby the agentsof both sides(cf., e.g., Plut. Pomp. 6.2-6) and the Cinnanscould thushave seen Sullan coins.

(b) it is not implausibleto suggestthat Scipio learnedof the coins duringhis negotiations with Sullaandthathe wouldhavebroughtword of themback to Rome. Here again,one sourcespeaksof the consul's

troopsbeingbribed. 38

ARTHUR

158

KEAVENEY

(c) Commanders in the field, such as Albinovanus, were able to

communicate with the enemyandwe may suspectthatSullanpartisans in Rome could do likewise. The youngerMarius certainlyseemsto havethoughttheycould.As furtherproofof the degreeto whichRome wasopento Sullanpenetrationwe maydrawattentionto how remarkably

wellinformed theproconsul wasaboutthemoodofScipio's troops. 39 (d) We may also suggestthat non-combatants passedto and fro betweenthe two areas.Sulla'sorderto depositbondsin the city before he hadevencapturedit (Livy, Per. 86), if not a pieceof bravado,may be a reflection

of this situation.

It will alsobe rememberedthatSullahadprofessedgoodwilltowards the Italianswhile abroad.Once he returnedhomehe proceededto woo them strenuously.In the winter of 83, he usedeverypossiblemeans to bring as many of them as he could to his side and even went so far asto concludea treatywith someof them(Livy, Per. 86). The Italians, therefore, must be numberedamong those at whom the coin was directed. It will be obvious

that much was at stake here. The more senators

one had on one's side the greater the veneer of respectabilitythere would be lent to one's cause. And if a man could not hope to entice everybody,he could still hopeto unsettlethe oppositionby giving a clearindicationof his determinationto seethat causetriumph.Again, the more Italians Sulla could win to his cause,the greaterhis chances of winningthiswar. In thesecircumstances, then, it wasnecessary for Sulla to advertiseas widely as possiblethe justiceof his claim, and coinswere obviouslyone of the meansby which he might do so. Having thus consideredSulla's belief that he was still a Roman proconsuland his need to convinceothersof the correctness of that belief, we mustconcludethat it is extremelyunlikelythe auguralsymbols could refer to an ancestor.Rather they must refer to his own positionat thistime. But how exactlydo they do this?It appearsmost plausibleto suggestthat they refer, in someway, to the power of a

consulor proconsul to taketheauspices whileon campaign. 4ø On the morningof his inaugurationthe consulwas obligedto take theauspices in orderto discoverif thegodswerewell disposed towards his holding the imperiumwhich had been bestowedon him by the people. If the signswere good then he automaticallypassedhis lex curiata de imperiowhich confirmedhis imperium.The latterwas now iustumjust ashe himselfwasiustus.Theseceremonies wereof supreme importance.Good omensmeant that the gods approvedof a man's imperiumas a whole. It wouldthenfollow thathe couldexpectfavour-

SULLA

AUGUR

1•9

ablesignswhenhe performedtheauspication whichwasrequiredbefore each act, at home or abroad, during his tenure of office. Failure to securethe approvalof heavenmeant that a person'simperiumcould not be confirmedby the lex curiata. He would, of course,still hold it, but auspicesfor all his acts would be unfavourableand he could accomplishnothing.Oncethe imperiumhadbeenmadeiustumthe magistratehad otherreligiousceremoniesto perform.Undoubtedlythe most importantof thesewere thosewhich took placeprior to his departure on campaign.Heaven already approvedof his imperiumand of the exerciseof its civil power. It was now necessary to find out if the gods also favouredthe exerciseof the military power. So auspiceswere taken once more and, if all went well, the consul made vows on the

Capitol for his imperiumand the republicin general.Favourableauspicescouldthenbe expectedon campaign.Shouldhis successes warrant it, then on his returnthe consulmight celebratea triumphand go

in procession to theCapitolto givethanksto thegods. 41 This then is how the properlyconstitutedconsulor proconsulwould behave. A contemporary,aware of the great controversywhich raged over Sulla's status, would, with the coin in his hand, soon realise what

Sulla was saying.He was just sucha proconsul,had alwaysbehaved

assuch 42andwouldcontinue to do so.On themorningof hisinauguration he had taken the auspices.When they provedto be favourable he had beenable to have his imperiumconfirmedand madeiustumby the lex curiata. In other words, the godsapprovedof his holding the imperiumand as a directconsequence of this he couldlook forwardto favourablesignsand a successfulissueto all of the individualactshe might auspicateduringhis tenure.And the godswere pleasedthat he shouldexercisethe military aswell asthe civil powerof the imperium. When he departedfor war the signshad beenfavourableand led him to expecthe would receivefavourableauspiceson campaign. Now comesthe crucialissue.For muchof his campaignSulla was, in the eyes of certainpeople, a hostiswhile he himself continuedto think of himself as a proconsul.Which view was right?Sulla's symbolismleft the world in no doubt. If we wishedto pressthe matter, we might say the augural symbolsrefer in a concretefashion to an individualact of auspication.The rod depictsthe delineationof the templumand the jug the drawing of lots. It may not, however, be necessaryto go this far. For us a crown on a coin symbolisesregal power;we do not necessarilythink aboutwhat a king doeswith it. The vital pointto graspis surelythatthe symbolsdrawattentionto themost importantcharacteristic of the imperiumiustum:the right to take the

160

ARTHUR

KEAVENEY

auspicesand to expecttheywouldbe favourable.But if Sulla hadbeen a hostiswe wouldexpectthe godsto shunhim and sendunfavourable signs.They did no suchthing. The omenswere goodandgreatvictories were won in consequence. The two trophieson the coin representthe

magnificent victorieswonat Chaeronea andOrchomenus43and, on a wider perspective,tell the world of the crushingof Mithridates.Men might say Sulla was a hostis;the godsevidentlythoughtdifferently. The rest of Sulla's symbolismon no. 359 is in harmonywith this view. The inscriptionIMPER ITERV(M) draws attentionto the fact that only a man who held an imperiumwhich was iustumcouldbe so hailed. The designationITERV(M) is to be relatedto Sulla's recalling of hispastdeedsin his secondletter. Oncebeforehe hadbeensohailed for vanquishingMithridatesin Cilicia in 96. Nobodyhad questioned

thevalidityof hisimperium then.44Why shouldtheydo sonowwhen he had won even greatersuccessagainstthe sameopponent?And if the gods smiled on Sulla, there was one deity in particularwho took a specialinterestin him: Venus. She, as mother of the Roman race, favoured him, not so much for himself, as Apollo did, but as ,the championof her descendants. As Rome's proconsulhe was labouring

onbehalfof herchosen raceandsosheblessed hisenterprises. 45Hence it wasfitting sheshouldappearon theobverseof the coinaccompanied

by a Cupidbearingthepalmbranchof victory. 46 Likewise, the other coin of Sulla (no. 367) also advertisesthe fact

thatSulla'simperiumwasiustum.Therewe havedepictedon thereverse a triumphatorcrownedby a flying victory.The message is clear:like any otherproconsulSulla had mountedthe Capitolon the day of his departureto make his vows. And, againlike any otherproconsul,he fully intended,when he returned,to celebratea triumphand go up thereoncemoreto give thanksfor thecampaignhe hadjust successfully fought. Nothingthat was fitting for a man with iustumimperiumto do would be left undoneby him. The head of Roma on the obversewas most probablyintendedto remindmen of the res publica on whose behalf he had fought, and whoseproperlyconstitutedofficer he continued to be.

Finally, we can now seehow the coinscontaineda promiseas well as a threat. Sulla had clearly enunciatedhis positionand his intention to behavein a mannerbefittingit. Obviouslythosewho would accept this had nothingto fear from him but, by implication,anybodywho triedto stophimenteringRomein triumphcouldexpecttobecrushed. It is my conclusion,therefore,that Sulla's coinswere intendedto proclaimto the world that he was a proconsul of Romewith a iustum

SULLA

AUGUR

161

imperium.The auguralsymbolsplayedtheir part in this by reminding menof the mostimportantcharacteristic of suchan imperium:theright of the proconsulto auspicateindividualactson campaign,from which auspicationfavourableomens might be expected,foretellingthe vic47

tories which were to ensue.

Darwin College University of Kent at Canterbury

Arthur Keaveney

APPENDIX

The lex curiatade imperio,the auspicesand the religiousceremonial of the consul

It is, I think, obviousthat the interpretation of Sulla's coinagewhich I have offeredabovedependsto a large extenton the view I take of the nature of the lex curiata, the functionof the auspicesand the significance of thereligiousceremonialthatthe consulwasrequiredto perform.Sincetheseare controversial mattersI proposeto justify in detail here what I have statedin summaryin my text. Beforedawnonthemorningof hisinauguration theconsul-elect rose

fromhisbedand,in thepresence of augurs,tooktheauspices. • If these provedfavourable, hethen,asMommsen pointedOUt, 2summoned the curiateassembly himself.A. Magdelain,contra,is not persuasive. 3 Amongothershe wouldappearto havemisunderstood Cic. Leg. ag. 2.274andLivy22.35.4.Thelatteractually dealswithanincident where normal electiveproceduresfor the consulatewere not followed and anywayrefersto the centuriateand not to the curiateassembly.We may alsoaddthatMagdelainignorestheevidenceof passages suchas

Cic.Rep.2.25. In Republican timesthecuriateassembly wascomposed of thirtylictorswhostoodby awaitingtheresultsof the auspices. s It wasin thisbodythatthe consulpassedhiscuriatelaw.6 It is clear from this that the !ex curiata did not, as has sometimes

beenmaintained,conferthe powerto take the auspices.This power (andtheconsul'simperium)wasgivenby thepeopleandwe havejust seena consulexercise it beforepassing hiscuriatelaw.7The!excuriata, in fact, confirmeda man'simperiumandin the late Republicit was felt that the !ex itself and the procedures associated with it couldbe tracedbackto thetimeof thekings.ThusCiceromakesit clear(Rep. 2.25, 31, Leg. ag. 2.26) he regardedthe comitiacuriataasbeingan assemblywith real powersin the regal period.It could,on its own

162

ARTHUR

KEAVENEY

initiative, withdrawfrom a man, on a secondvote, an imperiumgiven by the first. A concreteillustrationof this belief is furnishedby Dion. Hal. (2.3-6). He describeshow Romuluswas electedby the cornitia curiata. He thenhadto take the all-importantstepof discoveringif the godswere kindly disposedtowardshis holdingthe irnperiurn.When the signsshowedthat they were, he went before the assemblyfor a secondconfirmatoryvote. In Republicantimes,a manthusconfirmed would then be called iustusand his irnperiurnwould be a iusturnirn-

periurn.8 So, it was generallyheld that it was preciselythis procedure,with the inevitablemodificationswroughtby time, which was observedin the Republic.Insteadof a king chosenin the cornitiacuriata, it was now a consulin the cornitiacenturiata.The auspiceswere still of vital importance.If a man did not obtainfavourableauspicesfor his irnperiurn, it meantthat it was displeasingto the godsand he would not receivegoodsignsfor any act whichmightneedauspicationduringhis term of office. More immediately,it would mean that he would not obtainthe lex curiata which would confirmhis irnperiurnand make it

iusturn.9 Thustheconfirmatory powerof thelexcuriatastillcontinued and the law was still held to be necessaryfor irnperiurn-holders even thoughthe cornitiacuriatahadall but vanished,beingrepresented, as we saw, by thirty lictors. Cicero tells us that it had dwindledaway almost to nothing. It only retains its form (species)and that in an imperfectway. He clearly differentiatesbetweenit and the cornitia centuriataand tributa. They still have real power; the cornitiacuriata hasnone.They are real assemblies with real votes;the cornitiacuriata is a rubber-stampgathering.It meetsonly to hear the resultsof the

auspices andif thesearegoodit automatically passes thelexcuriata.•o So, the lex curiata was not now alex passed,as menthoughtit once had been, by the people. Its grantingwas, in fact, the immediate consequence of favourableauspices.Therefore,it is clearthatthe most importantpart of the proceedings was the takingof the auspices.All

depended on it.• Because of thisoursources lay particular stress on it and Cicero saysthe cornitiacuriata met auspiciorurncausa. We may sum up, then, by sayingthat in the late Republicmen believed that, once, the cornitiacuriata had had the power to vote a secondtime on a man's irnperiurnand confirm it by alex curiata. The lex still had this confirmatorypower but, as the cornitiawas now powerless,itsgrantingwasdependent entirelyonthecandidate'sobtain-

ingfavourable auspices. •2Theimportance whichattached tothesecere-

SULLA

AUGUR

163

moniesmay be gaugedfrom the fact that a furoreinevitablywascaused

whenever anybodyneglected them.•3 After passingthe lex curiata, the newly inauguratedconsulreceived the salutatioof the peoplein his house.He then performedhis first official act, which was to go in solemnprocessionwith his lictorsto

theCapitol.Thereheofferedsacrifice andmadehisvowstoJupiter. 14 Now, after the requiredauspication,the new consulheld a meeting of the senatein the temple of Jupiter. As religious businessalways took precedenceover secular,this meetingfirst concerneditself with fixing the date of the Latin Festival--no consulcould leave for his provinceuntil this was celebrated.After this, the senatecouldconcern itself

with

secular

business

if it wished.

The

new

consul

was also

required to make sacrifice to the Penates publici and Vesta at Lavinium. •5

Specialceremonies,too, were necessarywhen the consulleft to go

to war.•6 The imperium,in its entirety,whichgavebothcivil and military power had alreadybeengrantedto him by the peopleand, as we have seen, divine approvalhad also beenobtainedfor it by means of the auspices.However, the consulhad sofar only exercisedthe civil powerof the imperium,sincethe military couldnot be usedwithin the

pomerium. •* It wasfelt to benecessary for himto discover if thegods were well disposedtowardshis exercisingthemilitary power. He therefore took the auspicesfor his campaignas a whole and, if they were favourable,he went to the Capitolto makevows for his imperiumand

theRepublic in general. •aHe didnot,despite whatseems tobegenerally believed,taketheauspices on theCapitol.19 We have seenthat when a consul,by meansof the auspices,had won approval from the gods for his imperium as such it was still necessaryfor him to auspicateacts, suchas the summoningof the senate,performedin the city by virtue of the imperium.Likewise,

althoughauspicesfor the military powerwere held throughoutthe

campaign, 2øindividual actson thatcampaign, suchasthegivingof battle,hadto be auspicated on the daytheywereperformed. 2• If on anyof theseoccasions theconsulreceivedunfavourable signsit would meanthatthe auspices for his militarypoweras a wholewereflawed

andhe wouldhaveto returnto Rometo repeatthem.22Shouldthis happenin a far-offprovince,whenreturnwasimpracticable, theconsul wouldthendesignate an areaof landasRomeandwouldrepeatthem in the ternplumthuscreated. 23 After makinghis vows, the consulthendonnedhispaludamentum

164

ARTHUR

KEAVENEY

(military cloak) and left the city. He might not returnuntil the period

of hisimperium hadexpired.24 Thedonning of a specialcloakandthe prohibitionon return clearly show that the military power of the im-

periumwasfelt to be differentfromthecivil.25 If the consulon his return celebrateda triumph, he would go in procession to the Capitolandgive thanksto thegods.He alsosacrificed once more at Lavinium. 26 NOTES

1. B.W. Frier, 'Augural symbolismin Sulla's invasionof 83', ANSMN 13 (1967) 111-118(=Frier 1); E. Badian,'Sulla'saugurate',Arethusa1 (1968) 26-46 (=Badian 1); B.W. Frier, 'Sulla's priesthood',Arethusa2 (1969) 187199 (=Frier 2); E. Badian, 'A reply', Arethusa2 (1969) 199-201 (=Badian 2). Since theseexchangesthe only scholarto touchon the matter has been J.R. Fears, 'The coinageof Q. Cornificiusand auguralsymbolismon late Republicandenarii',Historia 24 (1975) 598. AlthoughI believeFears'general view of auguralsymbolismto be correct,his treatmentof thisparticularmatter appearsto be defective.After first declaringhe acceptsBadian's view that Sulla was not an augurin 84-82, Fearsthengoeson to say the proconsulput the symbolsof his augurateon the coinshe issuedat that time. The coinagein questionis no. 359 in M.H. Crawford,RomanRepublican coinage, 2 vols. (1974). Sulla's other coin (no. 367) will also figure in our discussion.

2. Pliny, NH 2.144 mentionsa marvelloussignvouchsafed Sullaedictatori, but doesnot say if he was an augur. 3. Crawford, no. 434.

4. On the developmentof personaltypesseeCrawford, op. cit. 2.725-734. A.H.M. Jones 'Numismaticsand history', in R.A.G. Carson and C.H.V. Sutherland(eds.), Essaysin Roman coinagepresentedto Harold Mattingly (1956) 15-16, expresses somedoubtsas to how effectiveimperialcoin types mightbe asan instrumentof propaganda.Cf. alsoW.E. Metcalf, 'The Roman Empirefrom Augustusto A.D. 284', in R. Carson,P. BerghasandN. Lowick (eds.), A surveyof numismaticresearch1972-1977 (1979) 186-187. Undoubtedly,thereis needfor discretionin interpretingcoin types--a point stressedby Crawford, 'The Roman Republic', in Carsonetc., op. cit. 175. However, given the well-knownfondnessof the Romansfor makingmuchof ancestralachievementsand virtues, it is not too difficult to envisagethat Pompeius'coin aidedhim to somedegreein his searchfor advancement.After all, he wasaddressing a politicallyawareaudiencewhowereinterested in such things,cf. (e.g.,) Sail. Jug. 85.4-10. Likewise, to anticipatea little, I shall be arguing,on the basisof its historicalbackground,that Sulla's typeswere intendedto conveya politicalmessageto his contemporaries. Indeed,Jones' analogywith modempostagestampswhich 'mainly reflect the mentalityof

SULLA

AUGUR

165

the post-officeofficials' is particularlyapt here, for Sulla'scoinsproclaimhis own policiesand not that of the Governmentof the day. 5. Badian 1.44 n. 5 also suggestedthat a coin of Faustus(Crawford, no. 426.3) mightbe takenasreferringto Sulla's augurate,but this may not be so, cf. Crawford, ibid. 6. Frier 1.116; Badian 1.37. 7. Cf. R. Bauman, 'The hostisdeclarationof 88 and 87 B.C.', Athenaeum 21 (1973) 270-293.

8. SeeJ.L. Strachan-Davidson, Problemsof theRomancriminallaw ( 1912) 2.16-74 (especially24-40) for a carefultreatmentof the questionof exilium. 9. There is somesupportfor this notionin our ancientsources,cf. Frier 1.116; Badian 1.37.

10. Cic. Cat. 4.10 with Strachan-Davidson, op. cit. 1.104 and A.H.J. Greenidge,Romanpublic life (1901) 6, 290, 292. 11. On this date and its possiblesignificancefor anothercontemporary problem see A. Keaveney, 'Young Pompey: 106-79 B.C.', AC 51 (1982) 131-132.

12. Gell. 10.15; Sen. Cons. ad Marc. 15.3; Dio 54.28.4, 56.31.3; Tac.

Ann. 1.62. The sourcesmake it clear that pollution arosefrom attendinga funeral. It did not arisefrom the carnageof battle sincemanypriests(Sulla included)commandedarmieswithoutdifficulty. For example,in the Tacitus passageTiberiusobviouslydid not objectto Germanicusleadingan army. It was solelyagainsthis adtrectareferalia that the emperorsethis face. 13. Cf. (e.g.) Frier 2.192. 14. See the remarksof T.F. Carney, 'The deathof Sulla', AClass4 (1961) 74n.

91.

15. Nor, self-evidently,can the passagebe usedas evidencefor Sulla's holdinga pontificate,despiteBadian 1.39-40. 16. Badian 1.42 n. 14, 2.191; Frier 2.191. 17. Dio 39.17.2 with Badian 1.31-36, 2.201. Frier 2.191-192 contra is not

convincing.He doesnot appearto give due weightto the evidenceof App. B.C.

1.63-64.

18. Badian 1.29-31, 2.200-201; Frier 2.190; MRR 1.562 n. 7; R.M. Geer,

'M. Aemilius Scaurus',CP 24 (1929) 292-294; G. Szemler,The priestsof the Roman Republic(1972) 123-124. An exceptionis G.V. Sumner, The oratorsin Cicero'sBrutus(1973) 100. Sincehis views do not appearto have receivedwidespreadacceptance(cf. (e.g.) K.R. Bradley, Suetonius'life of Nero: an historicalcommentary (1978) 29-30), but are, in my opinion,of supremeimportance,I shalltry to establishthe casefor themmore fully than he himselfhasdone. It may be notedthatthosewho hold 'the unitaryview' do not appearalwaysto havecompleteconfidencein it, cf. (e.g.) Geer, op. cit. 292-293: 'On readingthesetwo passages onesuspects [my italics]thatthe occasionof the two acts,sofar asthey weretheresultof personalresentment, was the same.' Frier 2.190 is more sure, but offers no reasons.

19. Geer thoughtthe chargeof mishandlingthe ceremoniesat Lavinium

166

ARTHUR

KEAVENEY

could not refer to somethingScaurushad done as consul,sincethat was too long ago (115). This would then mean that the chargesScaurusfaced related to somethinghe did there as pontiff, since augurswere not presentat the ceremonies.Thus, Asconiusis wrongto call him an augur.However, Roman nobleswere notoriousfor rakingup old crimesas subjectsfor prosecution,cf. (e.g.) E. BadJan, Studies in Greek and Roman history (1964) 206-234. Moreover, althoughGeer doesnot appearto be right in assumingthat augurs did not play a part in theseceremonies, no priestcould,in fact, be tried apud populumfor malfeasance.So it would appearplausibleto suggestthat it was, after all, for somethinghe did as consulthat Scauruswas prosecuted,since we certainlyknow that any man who held that office was obligedto make sacrificeat Lavinium, cf. Obsequens 54; Val. Max. 1.6.7; Livy, Per. 55, with G. de Sanctis,Storia dei Romani 4.1 (1953) 169-170 and Frier 2.190 (after Mommsen). See further n. A 15 below. 20. Geer, loc. cit. 21. MRR 1.559. 22. Badian 1.42 n.22.

23. MRR 1.564-565. We have no meansof knowingwhen the repulsae took place. Any one of the following reconstructions is possible,but none affectsthe validity of ourthesis:(a) bothrepulsaeoccurredbeforethe tribunate; (b) one or other did; (c) both repulsaeoccurredduringthe tribunate. 24. Oros. 5.21.3; Cic. Sest.7. For Sulla'spowerto declareScipioa hostis seeA. Keaveney,Sulla: the last Republican(1983) 161. 25. App. B.C. 1.85; Cic. Phil. 12.27. The fate of Scipio's fellow consul Norbanusis an instructivecontrast.He ill-treatedSulla'speaceemissaries and was later pursuedto Rhodesby the dictator'sagents.To avoid falling into their handshe committedsuicide(Livy, Per. 85; App. B.C. 1.91). He was pursuedas a hostis,not an exul, cf. above. 26. Cf. BadJan1.38. He alsotoyed with the notionthat Sulla could have had himselfexemptedfrom the law concerningtwo membersof a gensin the sameco!legium,but rightlyrejectedit. As we havejust seen,it is unnecessary. Frier 2.197 is unsatisfactory on thesematters. 27. We haveno meansof knowingwhat wasthe priesthoodof whoseloss Sulla complainedin 84. Badian 1.39-40 suggests it might be the pontificate, but a moment's

consideration

will show that the nature of the evidence which

preventedusfrom deducingit wasan auguratealsoinhibitsusfrompostulating that it was a pontificate,cf. n. 15. 28. CompareCrawford,op.cit. (n. 1) 1.80,373,386 withBadian1.42n. 10. 29. Plut. Sulla 25.3-4, 27.4; Vell. Pat. 2.25.1.

30. Vell. Pat. 2.25.2; App. B.C. 1.86. It may seem,at first sight,bizarre thata Romanproconsul shouldcoinmoneyin orderto bribepeople.It appears considerably lessso whenwe bearin mindthe unprecedented stateof affairs nowexistingwhichwe shalldiscussbelow. It wouldalsoappearsafeto assume that anythingwhich was requisitionedon the marchthroughCalabria and Apulia(seepreviousnotewithtext)waspaidfor. In view of Sulla'sprevious

SULLA

AUGUR

167

chronicfinancial difficulties (cf. (e.g.) Plut. $ulla 12.4-8) it seemsplausible to suggestthe new coinswere usedfor this purpose. 31. For somefurtherremarkson Sulla's effortsto explainhis positionto a Greek audiencesee A. Keaveney, 'Sulla and the gods' in C. Deroux (ed.), Studiesin Latin literature and Roman historyIII (1983) 63-64. 32. I do notthinkSulla'ssoldierswouldneedto be persuaded of thecorrectnessof their chief's motivesor actions.It may be notedthat Crawford, op. cit. (n. 1) 1.81 n. 1 does not believe the coins were intended for Eastern circulation.

We should also bear in mind that the coins Sulla struck in Greece

were imitationsof Atheniannew-styletetradrachms (discussion andreferences in Crawford, ibid.). 33.

Badian

1.27.

34. 'The main aspectsof politicalpropaganda on the coinageof the Roman Republic', in Carsonand Sutherland,op. cit. (n. 4) 63-95. 35. Crawford, op. cit. 2.730-731. 36.

Frier 2.187.

37. A detaileddiscussion of thesematterswill be foundin my two articles, 'Who werethe Sullani?'Klio 66 (1984) 126-137and'What happened in 88?', Eirene 20 (1983) 53-86. See alsomy Sulla (n. 24) 56-77, 117-124. 38. See App. B.C. 1.85• Plut. $ulla 28.3. 39. See App. B.C. 1.85, 88, 89, 91. 40. See P. Willems, Le droit public romain (1883) 239-240.

41. A detaileddiscussion of thesematterswill be foundin the Appendix

to this article.

42. It is worth repeating,perhaps,that we are not here concernedwith whether,in thelightof Sulla'sbehaviour in 88, thiswas,legallyspeaking, a justifiedclaimornot.Whatweareattempting to showis themessage, whether trueor false,thatSulla•himselfwastryingto convey.Cf. n. A24 below. 43.

Frier 1.112 n. 5.

44. Badian 1.46 n. 64• Th. Mommsen, Histoire de la monnaie romaine (1870) 2.440 n. 1. B adian here also refutes the notion that Sulla would consider

histitleimperator asthesolesource ofhisauthority afterbecoming ahostis. 45. ForSullaandVenusseefurthermy 'Sullaandthegods'(n. 31)60-65. 46. In my view Crawford'streatmentof thiscoin(op. cit. (n. 1) 1.374) is most unsatisfactory,sinceit seemsto be basedon theorieswhich I believe to

be erroneous (cf. Appendix)andis imprecisein its formulation.Nor do some of hisotherobservations onthesemattersinspireconfidence. Because no. 367 formsthelargerpartof Sulla'scoinagehe concludes (p. 387), 'thechoiceof it in preference to no. 359 is presumably significant.'We arenottold,however,

in whatwayand,in truth,it is difficultto seehow.Sullamadehisposition clearfromthe startandneverwaveredfromit. We shouldexpect,therefore, thatthecoinswouldreflectthissituation and,asI argue,thisis whattheyin factdo. Bothof themproclaimthe samemessage: thevictorieswonby the manwithimperium iustum.Equallyunlikelyis Crawford'ssuggested equation (p. 388)of theheadof Apolloonsomecoinsof theperiodwithlibertas.There

168

ARTHUR

KEAVENEY

is not a shredof ancientevidenceto supportthis idea. On Apollo andSulla seemy 'Sulla andthe gods'(n. 31) 56-60. Nor wouldeverybodyacceptthat Sulla's coinsshowa new individualisticattitudetowardsthe respublica (Crawford, op. cit. 2.732). See furthermy Sulla (n. 24) passim. 47. An earlierdraftof thispaperwasimprovedby somepertinentcomments fromProfessor T.V. Buttrey.The presentversionhasbenefitedfromtheadvice of ProfessorE. Badianand the anonymousreadersof thisjournal.

Notes to Appendix A1. On all of theseceremoniesseeDion. Hal. 2.5-6. The time of rising: Th. Mommsen,Le droitpublic romain (1893) 1.116. A2. Op. cit. 2.259, 269. A3. Recherchessur l'imperium (1968) 26-30. A4.

See further below.

A5. G.W. Botsford, The Roman assemblies(1909) 198. A6. Livy 9.38.15; Cic. Rep. 2.25, cf. J.J. Nicholls, 'The contentof the

lex curiata', AJP 88 (1967) 265. Botsford'sstatement(op. cit. 191) that a separateauspicationwas necessary for the curiateassemblyis erroneoussince in the casehe cites(Livy 9.39.1) repetitionwasnecessary becausethe auspices were unfavourablewhen first taken(Livy 9.38.15-16). A7. Auspicesand imperiumcomefrom the people:Cic. Leg. 2.17, Div. 2.76; Festus43L. The divisionbetweenauspices andimperiumis well brought

outbysuchpassages asCIL 12626;Livy22.1.5,cf. Nicholls,op.cit.260-263. The theory that the power to take the auspicesderivesfrom the curiatelaw hasbeenexpounded mostrecentlyby Magdelain,op. cit. 16-36(seeespecially 19-20) and (less fully) by R. Develin, 'Lex curiata and the competenceof magistrates',Mnemosyne30 (1977) 59. Magdelain baseshis thesisalmost entirelyon two phrasesof Cicero:curiata tanturnauspiciorumcausaremanserunt (Leg. ag. 2.27) and illis ad speciematque ad usurpationeravetustatisper XXX lictores auspiciorumadumbratis(Leg. ag. 2.31). These, interpretedas they stand, divorcedfrom their context, do indeed seem to lend powerful supportto Magdelain's theory.When, however,we considerthewholeof Cic. Leg. ag. 2.26-31, togetherwith some other sources,we will find a very different storyemerging,cf. below. A8. The term iustumimperiumwill be found in Livy 22.1.5, while iustus comesfrom Gell. 13.15.4 (a difficult passagewell discussed by Nicholls, op. cit. 270-274). On theconfirmatorypowerof thelexcuriataseefurtherNicholls, op. cit. 274-277. A9. See, for example,Livy 9.38.15-16, 21.63.5-221.7. Mommsen,op. cit. (n. A1) 2.282 n. 2, rightly emphasizedthat no auspicesmeant no lex. This was deniedby Magdelain,op. cit. 41, althoughthe text he citesto refute Mommsen(Livy 22.2.5) is fatal to his own theory and supportsthat of the older scholar.Cf. Develin, op. cit. 60. Cic. Phil. 5.16-45, Leg. ag. 2.30, tells us that the curiatelaw was necessary for oneengagingin military affairs

SULLA

AUGUR

169

while Dio 4.43 addsthat it was also requiredin order to hold cornitia.This was so becauseonly one who had his irnperiurnconfirmed(iusturn)by the curiatelaw as a resultof obtainingthe god's favourthroughbenignauspices couldperformthesefunctions.Withoutthatpowerhis imperiumcouldnot be called iusturn,auspicesfor all his actswould be unfavourableand he could accomplishnothing(Livy 22.1.5). See furthern. A24. A10. Leg. ag. 2.26-31. A11. In the caseof Flaminius(Livy 21.63.5-22.1.7) we may observethat emphasisis laid upon the auspicesas being the most importantpart of the proceedings andthelex curiataisreferredto onlyin passing(22.1.5). Flaminius undoubtedlypossessed the irnperiurnbut feared,notthatthe lex curiatawould be withheldfrom him, but that his auspiceswould be falsifiedby the augurs presentand that, in consequence, the irnperiurnwouldbe declareddispleasing to the gods(Livy 21.63.5-6). His drasticsolutionto the problemwasto avoid taking them altogether. A12. It is true that another tradition recordedby Livy and Dion. Hal. (referencesand discussion in Magdelain,op. cit. 30-35) concerning the regal periodwouldseemto havenothingto sayabouttwo separatevotesfor a king. This need not upsetour thesis,however,which restson the fact that Cicero and Dion. Hal. (supportedby Livy) are both talking of what they conceived the curiateassemblyto be, as well as informingus of what the connection betweenthe lex curiata and the auspiceswas in their day. Whetherthey are right or wrong in thinkingthat the procedureof the Republiccouldbe traced back to the kingsdoesnot affect the argument.For speculationon the possible real originof the doublevote, seeR.M. Ogilvie, A commentary on Livy Books 1-5 (1965) 733. A13. See the incidentsgatheredby Botsford, op. cit. (n. A5) 191-195

(althoughhis conclusionsare unsound). A14. Ovid, Fasti 1.79, Pont. 4.4.23-35; Livy 21.63.7-8. A15. Ovid, Pont. 4.4.35-36; Livy 21.63.5-8; Gell. 14.17.9; Cic. Postred. ad quir. 5.11; Macrob. Sat. 3.4.11.

A16. Theseceremonies werepreciselythesamefor proconsuls withproroguedirnperiurn,cf. App. B.C. 2.18; Cic. Div. 1.29, 2 Verr. 5.34. As will

become obvious in a momenttheirobjectwassuchthattheactof prorogation wouldnot absolvea man from his obligationto performthem.Nor doesthe

factthatthesenate gradually usurped thepeople's rightto givetheseirnperia meanthattheceremonies werethenabandoned. Magdelain,op. cit. 52-54,in accordance withhisthesisthatthemilitaryimperium wasgivenat a separate ceremony fromthecivil, is of theopinionthatthesepro-magistrates actually

performed theirceremony outside thecity. Unfortunately, thistheoryis not securely based(cf. n. A25) andtheonlycaseof an extra-urban auspication Magdelaincanfind is thatof Octavianat Spoletumin 43. This, however,took placeat a time whenthe Republican constitution wasbreakingdownandin fact seemsto be repugnant to normalcustom,cf. Livy 21.63.10. A17. SeeMagdelain,op. cit. 43-45.

170

ARTHUR

KEAVENEY

A18. Livy 21.63.9; Cic. 2 Verr. 5.34. A19. This view hasbeenmaintainedby scholarsfrom Mommsen(op. cit. (n. A1) 1.114) to R. CombOs,Imperator (1966) 387, 389. The sources, however,makeit clearthatthetakingof auspiceselsewherewasthenfollowed by the vows on the Capitol, cf. (e.g.) Livy 21.63.9, 22.1.6, 42.40.1. Festus 15L alonespeaksof auspication on theCapitol,but thishasnothingto do with ceremonieswe are considering,cf. A. Alf61di, Early Romeand the Latins (no date) 119.

A20. This may be seenfrom Cic. Div. 2.67-77 and Serv. on Aen. 9.24. A21. Macrob. Sat. 1.3.7; Gell.3.2.10. Hamspicywas alsoperformed.For both ceremoniesseeLivy 9.14.3-4. A22. Livy 8.30.1-2 A23. Serv. on Aen. 2.178, pace Magdelain, op. cit. 52. It may be noted

thatwhenCicerospeaks (Div. 2.76) of thosewhowagewarwithouthav. ing the power to take the auspices,he is referringto privati cum imperio, cf. P. Willems, Le sdnatde la Rdpubliqueromaine(1878-1885) 2.557-559 andW.F. Jashemski,The originsand historyof the proconsularand thepropraetorian imperiumto 27 B.C. (1950) 21-24. A24. Paludamentum:Livy 21.63.9. Prohibitionon returning:Cic. 2 Verr. 5.34. Magdelain, op. cit. (n. A3) 46-48, is surelycorrectto suggestthat the holdingof thecomitiacenturiatawouldalsoinvolvetheexerciseof themilitary power of the imperium.The consulwould, as he suggests,take the auspices for this military act within the city and then summonthe comitiaoutside,cf. his explanationof Gracchus'error in 163. It may be remarkedthat, sinceit is more than likely Sulla had alreadyperformedhis ceremoniesfor departure to the East, it would appearthat his marchon Rome violatedthe prohibition we havebeenconsidering.He himselfseemsto havethoughthis dutyasconsul to protectthe stateoverrodetheseconsiderations. See further'What happened in 887', Eirene, cit. n. 37.

A25. Magdelain,op. cit. 54, believesthattheauspicescanconferimperium. The auspices,however,do not conferanythingon anybody;they are merely themeansby whichonedetermines thewill of heavenwithregardto a particular act. It is because of this mistaken belief and also because he has overlooked

the fact that the entireimperiumhasbeengrantedby the peoplein the comitia centuriataand been approvedby heavenat the taking of the auspicesthat Magdelain, op. cit. 43, thinksthat the consulreceivedthe military power of his imperiumwhen he took the auspicesprior to departure.In fact, auspices takenbeforedepartureare clearlyto gain divine approvalfor the exerciseof the military power of the imperium, since this power was felt to be distinct from the civil. There is no evidenceto supportDevelin's suggestion (op. cit. (n. A7) 61) that the consul,beforedepartingfor war, passeda secondcuriate law.

A26. Livy 14.2.5, 42.49.6. [A. Giovannini,Consulareimperium(1983), cameinto my handsonly when this article was alreadyin proof. Thus I can only make a brief commenton

SULLA

AUGUR

171

someof the pointshe raises.Rejectingthe MommsenJanspatialview of the

divisionof civil and military power, which I have implicitlyfollowed, he argues(9-15) thatthe phrasedornimilitiaequehasa meaningwhich'n'est pas territorialemaisqualitative'.Military andcivil aspectsaredifferentiatedwherever they are and the porneriurnis not necessarilythe boundarybetweenthe two. What I have had to say about imperiurnand the lex curiata would, I think, harmonisejust as well with this notion(if it shouldproveto be correct) as with the older. In this regard,it may be notedthat thereis a largemeasure of agreement betweenuson themilitaryimportance of thedepartureceremony (16-19, althoughhe doesnotseetheconnection with thelex curiata),theunity of the irnperiurn(27-30) and the importanceof the auspicia(33-37). I am not convinced, however, that he has establishedthe case for the existenceof two

curlatelaws or for the curiatelaw as a granterof the auspices(44-56).]

DEMOGRAPHIC REFLECTIONS ATHENIAN CITIZENS

ON THE NUMBER 451-309 BC

OF

In this articleI wish to arguethat the Atheniancitizen populationafter the introductionof Pericles"citizenshiplaw wasdoomedto be stationary or to decline,andthat any significantincreasein the numberof citizens living in Attica was ruled out. I will openmy accountby statingthe basicline of my argument. (a) Pericles'citizenshiplaw of 451, reenactedin 403, prescribedthat only childrenof citizenscould becomecitizens. (b) This law was strictly enforcedapart from a shortperiod during the PeloponnesianWar. (c) Naturalizationof aliensby decreewasdemographically insignificant, and most of the naturalizedcitizenswere princesand political leaderswho becamehonorarycitizensanddid not take up residencein Attica.

(d) In Athensasin all Greekpoleismigrationshada veryconsiderable influenceon the size of the population. (e) Immigrantsbecamemeticsand not Atheniancitizens,apartfrom the very few who were naturalized. (f) There was a constantand demographicallysignificantemigration from Athens, of citizens who became metics in other cities or cleruchs in Athenian

settlements abroad.

(g) Before ca. ^D 1750, a naturalgrowthof a populationexceeding one per hundredper year is unlikely and can only be acceptedif we have very gooddocumentation for it. (h) The few sourceswe have for Athensindicate, if anything,a decline

in the number

of citizens

from 451 to the end of the fourth

century.

(i) Given a limitednaturalgrowthof the citizenpopulationanda not insignificantemigrationfrom Athensof meticsand cleruchs,this emigrationmusthave neutralizedany possiblenaturalgrowthof the citizen population. (j) It is importantto distinguishbetweenthetotalnumberof Athenian citizens (including metics in other cities, cleruchs and naturalized foreigndignitaries)andthenumberof Atheniancitizenslivingin Attica.

172

THE

NUMBER

OF ATHENIAN

CITIZENS

173

For the runningof the democraticinstitutionsit is the secondnumber that matters.

(k) The number of Athenian adult male citizens in the later fourth centuryhasbeenestimatedat 21,000 by somehistoriansandat 31,000 by others.In this articleI do not commitmyselfto preferringthe one figureto the other.But in a futurearticleI hopeto bringnew arguments to bear on this question.

I will now openmy accountwith somedemographic reflections.There is a historicalaxiom aboutthe growth of populations.Until recent years,it was generallyaccepted,and it is still maintainedby many historians.The axiomcan be statedthus:a war, a plagueor a serious famine may well cut down a populationto, say, two thirdsor even a half of its former size; but after the catastrophe,if no furtherdisaster occurs,thepopulationwill, in thecourseof a few generations, replenish itselfto its formersize, whichis determinedby theamountof foodstuffs that can be suppliedby productionand/orby trade (accordingto the prevailingculturalpattern).An ancienthistorianwho still holdsthis view is, for example,Paul Cartledge.In Sparta and Lakonia (1979) he states(p. 222): "Elementarycomparativedemographyshowsthat, otherthingsbeingequal,a populationwhichsuffersevena majorloss of life from a disasterquicklyreconstitutes itself." If we apply this principleto classicalAthens,we are facedwith a seriousproblem:we haveevidenceof an adultmale citizenpopulation

in 431 amounting to ca. 40,000+ ,• andevidence,accepted todayby mostscholars,of a citizenpopulationin thelaterfourthcenturyamount-

ingto 21,000adultmales. 2 If we accept(withmostscholars) thatthe fourthcenturywasa relativelypeacefulandprosperous period,thetwo figuresareincompatible. It is easyto explainhowthenumberof citizens dwindledfromca. 40,000 + in 431 to, say,20,000 in 403, butaccording to the axiom statedabovethe citizen populationoughtto have grown to something like itsformersizein thecourseof thefollowing80 years. Therearetwopossibleescapes: (1) to questioneithertheca. 40,000 + in 431 or the 21,000 in the later fourth century;(2) to questionthe axiom and maintain,first, that it is wrong, and second,that if it were correct,it doesnotapplyto a citizenpopulation of theAtheniantype. I will beginwith a shortdiscussion of the figures.In SO 56 (1981) I have arguedthat Gomme'sestimateof 47,000 male citizensin 431

is basedon wrongassumptions. 3 My calculations implythat (e.g.) 40,000 is a possiblealternative.On the otherhand, it is impossibleto

174

M.H.

HANSEN

pressthe figure muchbelow 40,000, and, more important,we cannot precludethepossibilitythat(e.g.) 50,000 or 55,000 is therightfigure; in which casethe gap betweenthe numberof citizensin 431 and a centurylater becomeseven more conspicuous. The sourcesfor the later fourth centurygive us more latitude for conjecturinga highernumberof citizens,in orderto reducethe discrepancyas comparedwith the fifth-centuryfigure. When Athenssurrendered to the Macedoniansin the autumnof 322, an oligarchicconstitution was introducedandpoliticalrightswere restrictedto citizenswith propertyworth2,000 drachmasor more. We havetwo differentreports of the effect of this censuson the numberof full citizens. According to Diodorus18.18.4-5, only 9,000 Athenianspassedthe limit of 2,000 dr., whereas 22,000 were disfranchisedand forced to emigrate to Thrace. But Plutarch tells us (Phocion 28.7) that only 12,000 were disfranchised,and only someof theseemigratedto Thrace, whereas othersstayedin Athens. So, any historian,accordingto his beliefsanda priori assumptions, canprefereitherthe 31,000 reportedby Diodorusor the21,000 reported

byPlutarch. Gomme argued infavour ofthehigher figure, both inh!s

Population ofAthens andin hislaterarticlein JHS.4Oneof Gommes

reasonsfor preferringthe 31,000 was that he couldnot accepta static fourth-centurypopulation,with no increasewhatsoeverin the number of citizens. Let me quotea few lines from his article in JHS: "Jones convinceshimselfthatthecitizenpopulationof Athenshadbeennearly staticsince400, while in the fifth, from 480 to 430, he says 'that it musthavegrownat a prodigiousrate', in spiteof muchlossby emigration .... But, as arguedin The Population of Athens, I do not believe in a static fourth century either .... The only reasonthat Jonesgivesfor not believingthat in the fourthcenturythe population of Athensincreasedis becausehe prefersPlutarch'sfiguresaboutthe restrictionof thecitizen-bodyby Antipaterin 322 ( 12,000disfranchised)

to thoseof Diodorus (22,000disfranchised, 9,000remained). "5 Without committingmyself, yet, to preferringone numberto the other,I shallargueagainstGomme'sa priori assumption that, evenin ancientsocieties,peaceand prosperitymustlead to a (rapid) growth of the population. Within thelastfifteenyearsa majorbreakthrough in historicaldemography has taken place. A new methodof family reconstruction has made it possibleto obtain a very full insightinto the demographyof any communitywhichkept a full recordof baptisms,burialsandmarriages.The parishregistersof severalEuropeancountries(Germany, France, and Scandinavia)have been provedto be far more reliable

THE

NUMBER

OF ATHENIAN

CITIZENS

175

sources thanmanyhistorians assumed only a generation ago.6 The resulthasbeena mostnecessary revisionof our a priori assumptions aboutlife expectancyand populationgrowthbefore ca. 1750. Early modemEuropeis theperiodmostthoroughlystudied,buta considerable effort hasalsobeenmadefor medievalandfor Romanhistory,although

thesources herearemuchinferiorto theparishregisters. 7 Theimpact of the new historicaldemographyhasonly just begunto leaveits mark on Greek history, and it is worth mentioningthat we can still find

historians, asforexample Anthony Snodgrass orOswynMurray,8who, on very slightevidence,are willing to acceptan annualincreaseof a closedpopulationin archaicAttica amountingto 4% per annum,which is morethanfor Mexicoin the20thcenturyAD.Moreover,thisincredible increaseis combined with thebeliefthatfemaleinfanticidewasperhaps 9 not

uncommon.

I suggest thatthesoundest basisfor aninvestigation of thepopulation of ancientGreeceis the startlingresultsproducedby historicaldemographersfor early modemEurope,ca. 1500-1750. We mustacceptthat life expectancyat birth was perhapsas low as 20 years,and probably never over 30, and that the long-termpopulationfluctuationsnever resultedin an annualincreaseof 1-4 per hundred,but rather of 1-4 per thousand;whereasthe short-termfluctuationsmay showsometimes a drasticdecline, sometimesan annualincreaseperhapsamountingto

1 perhundred, andsometimes a stationary population. •oLetmeadduce two examples,onerelatingto a smallparishpopulationand oneto the populationof a small nation, viz. Iceland, which, in 1701, conducted the first populationcensusin modemhistory. One of the most thoroughlystudiedEnglishparishesis Colyton in southeastDevon. The graphof baptismsandburialsfrom 1550 to 1830 showsa surgeof populationgrowthdown to ca. 1640. A plaguedrastically cut back the population,and for nearly a centurytherewas a surplusof burialsover baptisms.The populationwas slowly declining

evenafterthe 17th-century catastrophe hadcometo an end.• The populationof Icelandshowsa long-termdeclinefrom 1200 to 1800, but interrupted by importantshort-termfluctuations; e.g., in the 14thcenturywe havefor almosta centuryanannualpopulation growth

of about0.5 perhundred peryear.•2Thesetwoexamples showthat, even after a disaster,the populationmay remainat a reducedlevel for a long period without any significantincrease,or any increaseat all, whereasin othercaseswe may have a short-termpopulationincrease of somesignificance. A verysimplecalculation mayalsoillustratehowslowlythepopultion of the world increasedbeforethe 18th century.Let us assumetht our

176

M.H.

HANSEN

planet, ca. AD 1, was inhabitedby only two persons,and that they propagatedthe humanspeciesat an annualrate of two per hundredper year (which is certainlynot excessivefor the period after 1750). The worldpopulationwouldthen,by 1980, amountto threehundredmillion

billion(3 x 1017= 300,000,000,000,000,000) humanbeings. 13 The historicalaxiom statedat the beginningof this papermust then be modified as follows: "Elementarycomparativedemographyshows that a populationwhichsuffersa majorlossof life from a disastermay reconstituteitself (probablyat a rate not exceedingan annualgrowth of one per hundredper year), but we have many examplesof reduced populationswhich were stationaryfor more than a centuryafter the disaster."

Returningto Athens, if we assumea citizen populationof 20,000 around400, which is indeedonly a guess,and an annualgrowthrate of 1 per thousand,the resultis slightlymorethan21,000 citizenswhen the democracywas abolishedin 322/1, which is what Plutarchreports. On theotherhand,if we assumean annualgrowthrateof 5 perthousand, which is alsopossibleunderfavourableconditions,we end up with a total male citizenpopulationof almost30,000 in the late fourthcentury, and may find supportfor Diodorus' figure. However, an annualincrease of 0.5 per hundredper year may have beenpossiblefor the population of Attica as a whole, but not for the Athenian citizen population,for reasons which I shall now discuss.

In 451 Periclesintroduceda citizenshiplaw accordingto which an Athenian could be a citizen only if both his parentswere Athenian citizens.This law was disregarded in the laterphasesof thePeloponnesian War, 14but it was reintroducedafter the restorationof the democ-

racy15andwasstrictlyenforced throughout thefourthcentury.What are the demographicconsequences of sucha law.9 It is well known that demographicchangesare determinednot only by the relationbetweenfertility and mortality,but alsoby migration. Now, it is importantto emphasizethatHellaswassplitup into600-700

different poleis, 16andthatthefreepopulation inallpoleis wascomposed of two elements, citizens and foreigners (in many poleis called

metoikoi). 17Apartfrom Sparta,the Greekpoliswascertainlynot a societyof stay-at-homes, sufferingfrom xenophobia.On the contrary, large numbersof Greekswere on the move all the time. Apart from all the sourcestestifying to foreignersin individualpoleis such as Athens,Corinthor Megara, it is apparentfrom Aristotle'sPoliticsthat meticsconstituted animportantelementin thepopulationof thestandard

Greekpolis.•8And to conduct an investigation of population growth

THE

NUMBER

OF ATHENIAN

CITIZENS

177

in a Greekpolis withouttakingmigrationsinto accountis an operation doomed to failure

and a waste of time.

In many medievaland modem societiesimmigrationand emigration may oftencancelout, with the resultthat migrationsdo not drastically changea demographicanalysisbasedon fertility andmortalityfigures. For a Greek polis I can also imaginethat emigrationwas balancedby immigration,but in thiscasethe neutralizingeffectappliesonly to the populationas a whole, not to the citizenpopulation. If a citizenmovesfrom, e.g., Corinthor Megaraor Athensandtakes up his residencein anothercity, the numberof citizensin thepolis he leavesis reduced,but the numberof citizensin his new polis is not correspondinglyincreased, since he will become a metic and not a

citizen. 19So,a citizenpopulation of theAthenian typewill alwaysbe reducedby emigration,but never increasedby immigration,and, consideringthe importanceof migrationfor the Greekpoleis and the slow naturalgrowthof populationsbeforeca. AD 1750, the inferenceis that a citizen populationof the Atheniantype could never grow. It was

doomedto be stationaryor perhapseven declining,if the citizenship laws were strictlyenforced. But was the citizenshiplaw strictlyenforcedin the fourthcentury? As far as the sourcesgo, it was. The Athenianshad a whole set of

publicactions aimedat intruders whoattempted to•ass as Athenian citizens.The mostimportantwas the graph• xenias,•"which seemsto have been appliedfrequently. Moreover, we know that the Athenians in 346/5 conducteda major revisionof all the local citizenregisters, and that many were struckoff the lists. It is ironical, but true, that the man who proposedandcarriedthe decreeorderingthe generalrevision

borethenameDemophilos. 2• But the citizenshiplegislationdid not precludenaturalizationby

decree. 22Wasthisoftenpractised andof anydemographic importance? During the first yearsof the restoreddemocracywe haveevidenceof three block grantswhich may have had someeffect on the numberof citizens:

1. In 403/2 the Atheniansconfirmedtheirprior grantof citizenship

toallSamians (IG II 2 1.51-55).SinceLysander hadrecently overthrown theSamiandemocracy andlefttheoligarchs in power,23it isreasonable to assumethat some, perhapsmany, of the exiled Samiandemocrats

tookupresidence in Athensandexercised theirAtheniancitizenrights. 2. In 401/0 an honorarydecreewaspassedfor the Heroesof Phyle, bestowingcitizenshipon thosewho had joined in the attack on the

Peiraeus (IG II 2 10, Osborne D 6). A fragmentary list of namesis

178

M.H.

HANSEN

appended to thedecree,andaccording to Osborne'sreconstruction there is room for no less than 70-90 naturalized citizens. D. Whitehead

has

now argued(LCM 9.1 (1984)) that all the ca. 900-1,000 foreigners mentionedin this decreeacquiredAtheniancitizenship. 3. Shortlyafterwards,in 400, Andocidesreportsthat someThessalians and Andrianshad been naturalizedin order to make up for the shortageof citizensafterthe war (Andoc. 1.149). Andocides probably refersto block grantsratherthan to individualgrants;but rhetorical exaggeration of a few individualgrantsis a possibilitywe cannotrule out.

After 400, however,them are no more block grants,unlesswe take the honours to the Elaeousians of the Thracian Chersonese in 341/0 to

be a blockgrantof citizenship (IG II 2 228=Tod174=Osborne D 15). But I follow Osbornein assuming"that the decreeis not so much a grantof citizenshipas an assimilationof the statusof the Elaeousians to that of kleruchs" (Commentaries83).

The only attestedblockgrantof the fourthcenturywas abortive: afterthe Atheniandefeatat Chaeronea,Hyperidesproposed andcarried a decreeprescribingthatcitizenrightsbe grantedto all metics(Lycurg. 1.41:Hyp. fr. 32 (Ps.-Plut.)and33 (Suda)),thatall slavesbemanumitted (Lycurg. 1.41; Hyp. fr. 32 (Rutilius, Ps.-Plut., Longinus, Anonymus)and33 (Suda)),andthatamnestybegrantedtoexiles(Hyp. fr. 32 (Rutilius)),to atimoi (Lycurg. 1.41; Hyp. fr. 32 (Dem. 26.11 and Anon. ad Hermog.) and 33), and to disfranchisedcitizens(Hyp. fr. 33). This decree,however,wasimmediatelyindictedasunconstitutionalby Aristogiton (Hyp. fr. 32-43;Aristogiton,fr. iii; Lycurg.1.36; Dem. 26.11). Hyperideswas acquitted(Hyp. fr. 32; Aristogiton,fr. iii), butthedecreewasnevercarriedintoeffect,owingto theconclusion of peacewith Philip (Dio Chrys. 15.21). So all unquestionable grantsof the fourthcenturywere individual grants,to whichI now tum. In the fourthcenturya citizenshipdecree hadto be passed by thepeoplein onesession andratifiedby a quorum of 6,000 at thebeginningof the followingecclesia.The procedure for ratificationwas devisedso that, probably,only one citizenshipdecree wasratifiedper session.The possibilityof having(e.g.) tengrantsof

citizenship proposed andcarried inonesession seems toberuledout.24 For the period403-322we haveevidenceof someonehundredgrants of citizenship.Abouta fourthare epigraphically attested,the restare

mentioned in literarytexts. 25Admittingthe fragmentary preservation of our sources,I guessthatthe Atheniansduringthe 82 yearsof the seconddemocracymusthavepassedseveralhundredcitizenshipde-

THE NUMBER

OF ATHENIAN

CITIZENS

179

crees,perhapsasmanyastenin a year. But manygrantsof citizenship werehonorificandpolitical,e.g., the grantsto Dionysiusof Syracuse, Ariobarzanes,Cersebleptes, Philipof Macedon,etc. Theybecamehonorarycitizens.The numberof naturalizedmeticsis muchsmaller:the

bankers Pasion andPhormion maybementioned asobvious examples. 26 So thedemographic importanceof naturalizationseemsto be negligible. After 400 there were probablyno more block grants,and only a few individual grantsto men who were either meticsor xenoi and who took up residencein consequenceof the grant. The Athenian citizen bodymay be describedasa peculiar"closed"population,whosedemographic determinantswere fertility, mortality and emigration,but not immigration or naturalization. Now, can it be demonstrated that Atheniancitizensemigratedfrom Athens to such an extent that it must have neutralizedany possible increaseof the citizenbodyby naturalgrowth?I believeso, andI will discussthe followingfour groupsof possibleemigrants:(1) exiles;(2) mercenaries;(3) roetics;and (4) cleruchs. (1) Permanentexile was the penaltyfor assaultwith intentto kill (if the victim was an Athenian citizen) and for intentionalhomicide (if the

victimwasa metic). 27Far morefrequently, however,exilewasthe result of a death sentenceagainsta citizen who did not standhis trial

but wasconvicted in absentia. 28This typeof self-imposed exile to avoid capital punishmentis well known in numerouspolitical trials. Furthermore,citizenspunishedwith atimia were allowed to stay in Attica, but oftenpreferredto go into exile ratherthansuffera humiliating

anddangerous lossof rights. 29Owingto theastonishing frequency of political trialsinAthens,3ø theexiles,inanyyear,probably outnumbered the naturalizedimmigrants, and so exile, in itself not an important demographicfactor, at leastneutralizesthe small positiveeffect which naturalizationmay have had on the growthof the citizenpopulation. (2) In the fourthcenturyGreek warfarewas increasinglydominated

by mercenaries. 3• Historians dealingwith Athenianhistoryregularly refer to Atheniangeneralsin foreignservice,e.g., Iphicrates,Chabrias andChares,or to foreignmercenariesin Athenianservice.But we must not forget the other side of the picture: that an increasingnumberof Atheniancitizensservedabroadin the armiesof barbarianprinces.Let me mention three cases.

Apart from Xenophon,at leasteight otherAthenians,and probably manymore,joinedthe Ten ThousandandfollowedCyruson hismarch

againstArtaxerxes. 32 When Chabriasfought in Egypt ca. 360, his mercenaryforce must

180

M.H.

HANSEN

have includedan Atheniancontingent;for in Egypt, nearMemphis,a dedicationhas been found which was probablyset up by ten of his mercenaries. Five are Athenians, the other five come from Boeotia,

Caryanda, Corinth,CyreneandNisyros. 33 When AlexanderinvadedAsia, largenumbersof Greekmercenaries were in the serviceof the PersianKing. In the battleof the Granicus, the numberof Greekmercenaries is reportedto havebeenca. 20,000. Most of themwerekilled, but 2,000 weretakenprisoner.The Athenian contingentseemsto have been particularlylarge, sinceAthenssent severalembassiesto Alexanderaskingfor their release.In the spring

of 331Alexander eventually granted thepetition. 34 The growingnumberof Athenianmercenaries servingabroaddid not necessarily leadto a permanent reductionin thenumberof Athenian

citizens.Someof the mercenaries returnedto Athens. 35Many were probablyunmarried,and if they were marriedthey undoubtedlyleft their familiesin Attica. So the long-termeffecton the sizeof the citizen populationmay havebeenrathersmall. (3) The term"Athenianmetics"is invariablytakento denoteforeigners who stayedin Athensfor a periodand often took up residencein Attica. This importantgroupof freeforeigners hasalwaysattractedthe interestof historians.A recentexampleis David Whitehead'sexcellent

monograph, Theideology of theAthenian metic. 36But "anAthenian metic" can also denote an Athenian

citizen who leaves Athens and

becomesa metic or a xenosin anothercity. This reversephenomenon is oneof the neglectedproblemsin Athenianhistory.It is barelymen-

tionedbyhistorians andnoonehasdevoted a bookoranarticletoit.37 What I can do here is only to emphasizethe problem, to presenta selectionof sources,and to discusssomeimplications. In the literarysourcesthereare casualreferencesto Athenianswho becamemeticselsewhere.Platoreportsthat Euthyphro'sfamily were

farmers in Naxos(P1.Euthyph. 4C).38In Antiphon 5, thevictimHerodes andhis family wereprobablyformerAtheniancleruchswho stayedon

asxenoiin Mytilene(Ant.5.29).39Philonwasa meticin Oropus during theoligarchyin thelatefifth century(Lys. 31.9). The sophistPolycrates settledfor many years, perhapspermanently,in Cyprus(Isoc. 11, hypoth.).One of Demosthenes' guardians,Aphobus,movedto Megara and lived thereas a metic(Dem. 29.3); andpoorLeocrates,who fled Attica after Chaeronea in 338/7, was a metic first in Rhodes and then

in Megarabeforehe, in 331/0, returnedto Athensand was charged with treasonfor having left his countryin time of danger(Lycurg. 1.21). In a fragmentof a fourth-century forensicspeechpreservedon

THE NUMBER OF ATHENIAN

CITIZENS

181

papyruswe hearthatAntiphanes,probablyan Athenianemigrant,lived with hisfamily in Selymbriaandmarriedoff hisdaughterto anAthenian emporos,who took her back to Athens(P.Oxy. 2538). Now, the reasonwe know about theseAthenianemigrantsis that they eventuallyreturnedto Athens. If they had emigratedwith their families and settledelsewherepermanently,it is unlikely that they would have left any traces in our texts. We have to admit that the literary sources,principallyforensicspeeches,are unsuitablefor an investigationof Athenianemigrants. So we mustturn to the epigraphicalevidence,lookingfor Athenians outsideAttica. Proxenydecreesand dedicationsare, of course,of no

valuefor thisinvestigationri øWhatwe haveto lookfor is tombstones found outside Attica, inscribed with the names of Athenian citizens.

Attestationsof womenor severalrelatedpersonsrecordedon the same stonewill be our bestindicationthat we haveto do with true emigrants and not with an Athenianwho accidentallydied beforehe got back to Athens. But in most cases, I believe, a tombstoneis evidence for an

emigrant.Runningan eyeoverepigraphical publications andanunpublishedfile of AtheniansoutsideAthens,which ChristianHabichtkindly

allowedme to use,I canpresent a list of tombstones. 4• The list is by no meanscomplete.It is only meantas an illustration of the importanceof a usuallyneglectedtype of sourcefor studying the neglectedproblemof emigrationfrom Athens.Sometwenty-five namesis perhapsnot a very impressiveresult,comparedwith the fact that someten yearsago I counted490 Athenianmetics(in the usual

sense)buriedin Atticain the fourthcenturyalone, 42 whereasthe tombstones listed above cover several centuries.

But we must bear in

mind that Attica is betterexcavatedthanany otherpart of the Hellenic world and that the Athenianinscriptionsare by far the bestpublished andmoststudiedof all inscriptions.Much of the evidenceI am looking for is no doubtstackedin small local museumsawaitingpublication, and much more is not yet excavated.Furthermore,the Athenianswere in the habit of putting records,both public and private, on stone, whereas other Greek

sites that have been well excavated

often come

up with an astonishinglysmall number of inscriptions,tombstones included.CorinthandOlynthosareobviousexamples.Thereasonmust be that the Olynthians and the Corinthians were not used to com-

memoratingthe dead by settingup a stonestele with the name in-

scribed. 43So theinformation aboutAthenian, s buriedoutside Athens will neverbeasgoodastheevidencewe havefor meticsburiedin Attica. My superficiallycollectedlist of Atheniansburied outsideAthens

182

M.H.

HANSEN

indicatesthat there was a constantflow of emigrantsfrom Athens to otherGreek cities. In all periodsthe immigrantswould probablyoutnumberthe emigrants,andI havelittle doubtthatAthensattractedmore

meticsthanAthenians everleftAthensto gotootherpoleis. 44Nevertheless, the emigrantsmay be a demographicallysignificantfactor, and its negativeimpact on the numberof Atheniancitizensmust be taken into accountin all studiesof the citizen population. (4) The mostimpressiveemigrationof Atheniancitizensfrom Attica is connectedwith the sendingout of cleruchies.By 392 Athens had reoccupiedSkyros,LemnosandImbros,andthe conquestwasacknowl-

edgedby theKing'sPeacein 387/6.45In the sameperiodDelosfell intoAthenian hands. 46A navalinscription records cleruchs to be sent to an unknowndestination in 370/69.47Threecontingents of cleruchs were sentto Samos:in 366/5, in 361/0 and in 352/1.48Potidaeawas

conquered in 364/3 (?) and settledwith cleruchs in 362/1.49When CharesdestroyedSestosin 353, thewholeChersonese exceptfor Cardia

hadto acceptAthenian cleruchs. 5øFinally,in 325/4theAthenians sent a squadronto the Adriatic to found a colony, which was probablya

cleruchylike all the otherAtheniansettlements of thisperiod. 51 For the fifth centurytherehasbeensomedoubtwhetherthe cleruchs really settledabroador, at leastin the caseof somecleruchies,stayed

in Atticaandwereabsentee landlords. 52For thefourthcenturythere can be no doubtthat the Atheniancleruchsemigratedfrom Athenswith

theirfamiliesandtookupresidence in theplacestheyweresentto.53 It is difficult to estimatehow manycitizenswere sentout ascleruchs. If Lysander, in 405/4, allowed the Athenianson Skyros, Lemnos and Imbros to stay and did not force them to go back to Athensas he did

elsewhere, •4 the recoveryof theseislandsmaynothaveentailedthe sendingof new cleruchs.But accordingto some literary sourcesone

of thecontingents sentto Samosinvolved2,000men,55anda fragmentary catalogueof cleruchs,probablyto be connectedwith the second contingent, indicatesthat at least several hundred were sent on this occasion.56The total number of Athenian citizens sent out as cleruchs

musthaveamountedto severalthousand,whichimpliesthat, especially in the 360s and 350s, more than a tenth of all adult male citizens were

removedfrom Attica and had no longerany possibilityof participating in the runningof its democraticinstitutions.They were, of course,still citizens. Their childrenall had demotics, 57 which indicatesthat the cleruchsprobablyhad to take their sonsback to Attica to have them inscribedin their demes.So cleruchieshadno automaticnegativeeffect on the total number of Athenian

citizens.

And if the cleruchs or their

THE

NUMBER

OF ATHENIAN

CITIZENS

183

descendantsreturned to Athens after a period of emigration, the cleruchiescannotevenhavehad any permanentnegativeeffecton the numberof Atheniancitizensin Attica. So we mustask:what happened to the cleruchssentout in the fourthcentury? The cleruchsin Potidaeawere sentback to Athensby Philip when

hecaptured thecityin 356.58TheSamiancleruchs wereexpelled from the islandin 322/1.5 Many of themreturned to Attica,but we know that at least one did not, viz. Neocles, the father of Epicurusthe philosopher.He went to Samosas a cleruch, probablyin 352/1. His son Epicurusreturnedto Athens in 323/2 and served as an ephebe,

whereas hehimselfmovedonto Colophon, wherehedied.© Lemnos, probablyfollowedby Skyrosand Imbros, brokeaway from Athensin

314/3.6• Someof theinhabitants mayhavepreferred Athenian citizenship and returnedto Athens,but mostundoubtedlystayedand were no longer Athenian citizens. The Athenian cleruchson the Chersonese may have sufferedthe samefate as thosein Potidaea:they may have been sent back to Athens when the Macedoniansconqueredthe area; but we do not know.

Our sourcesare, as usual,inadequate,andsothe inferencesmustbe cautious:manyof thefourth-century cleruchsundoubtedly returned,or weresentback,to Athens,but at leastsomeof themstayedor moved on to other parts of the Hellenic world. The result was a net loss of citizens. It must have been difficult for them to send their sons home

to Atticato be inscribedin their demes,andif they failedto do so, it is easyto imaginethedifficultiesthechildrenor grandchildren of the originalcleruchsmusthavehadin beingacknowledged astruecitizens

if theyeventually returned.Emigrants in largenumbers cannoteasily all be broughtbackto theircountryof origin. Let us sumup. Exile was not demographically significant,but was probably commonenough to neutralizeany positive effect which naturalizationmay have had on the number of citizens. Mercenaries

mayhavereturned to Athens(if theysurvived),butevenso,bymarrying late or by leavingtheir wivesfor someyears,they may havereduced thefertility of someAthenianwomen.The sameholdstrueof Athenians who,evenfor a shortperiod,weremeticsin otherpoleis;andAthenians who left Atticawith theirfamiliesto takeup residence elsewhere must certainlyhavehada permanentnegativeeffecton the sizeof thecitizen body. Most important,in someyearshundredsor even thousandsof

Atheniansleft Attica as cleruchs,and it is very likely that manyof them (or their descendants) never returned.

DuringthePeloponnesian War theAthenianpopulation wasexposed

184

M.H.

HANSEN

to plague (430-26), to severecasualtiesin war (e.g., in 424, 413 and 405) and to famine (405-4). In the fourth century,the annualgrowth rate of the citizen populationmay have been in the rangeof 0.5-1.0 per hundred,butthe emigrationto otherpoleisor to cleruchiesprobably neutralizedtheeffectof thenaturalgrowth,andwasnotcounterbalanced by immigrationor naturalization.In the first half of the fifth century

the sources indicatea rapidgrowthof the citizenpopulation, 62but I follow CynthiaPattersonin believingthat the principalreasonwas "the admissionor entry of non-Atheniansinto the demes and phratries,

resulting in thecreation of newAthenians". 63Aftertheintroduction of Pericles'citizenshiplaw any furtherincreasewas probablyruled out, and the citizen populationwas doomedto decline(in the later fifth century)or to remain stationary(in the fourth century).

TheSpartan oliganthropia is a well-known phenomenon. 641believe that a similar (but much slighter)oliganthropiamust be assumedfor Athens,dueto the strictrulesfor acceptingnew citizenscombinedwith

a continuous emigration fromAttica.65 M.H.

Universityof Copenhagen

Hansen

NOTES

1. Thuc.2.13.6-7. Cf. A.W. Gomme,"The Athenianhopliteforcein 431", CQ 21 (1927) 142-50; idem, Thepopulationof Athensin thefifth andfourth centuriesB.C. (1933) 4-5; idem, A historicalcommentaryon ThucydidesII (1956);idem,"The population of Athensagain",JHS79 (1959)61-8;A.H.M. Jones, Athenian democracy(1957) 161-80; E. Ruschenbusch,Athenische lnnenpolitikim 5. Jahrhundertv. Chr. (1979) 137- 40; R.P. Duncan-Jones, "Metic numbersin PericleanAthens", Chiron 10 (1980) 101-9; M.H. Hansen, "The numberof Athenianhoplitesin 431", SO 56 (1981) 19-32. 2. Diod. 18.18.4-5; Plut. Phocion28.7 (numberof citizensin 322). Ath. 272 C = KtesiklesFGrHist 245 fr. 1 (numberof citizensin 312-09). Cf. Gomme (1933) 17-19; (1959) 67-8; Jones76; S. Isagerand M.H. Hansen,Aspectsof Atheniansociety(1975) 11-14; E. Ruschenbusch Anh. 1; idem, "Epheben, Bouleutenund die Btirgerzahlvon Athen um 330 v. Chr.", ZPE 41 (1981) 103-5; idem, "Noch einmaldie BtirgerzahlAthensum 330 v. Chr.", ZPE 44 (1981) 110-12;P.J. Rhodes,"Ephebi,bouleutae andthepopulation of Athens", ZPE 38 (1980) 191-201. 3. Gomme 1981. My objectionsagainstGommeare: (1) the 16,000 in the

reservesis nota populationfigurebut an army figure. Consequently, thetotal numberof oldest,youngestandmetichoplitesmaybe (e.g.) 20,000 or perhaps even25,000. (2) There is no evidencethatthe oldestandthe youngestcompriseonly citizensof hoplitecensus.The phrase6oot 6•,i'c•tt riocry(Thuc.

THE

NUMBER

OF ATHENIAN

CITIZENS

185

2.13.7) goesonly with [texoi•ctov andnot with x•v wto•:ti•:tov.So the oldestand the youngestprobablyincludethe oldestand youngest of thethdtes.(3) The numberof thdtesisunknown.Gommearbitrarily estimates18,000, on the a priori assumptionthat the poor must have outnumberedthe propertiedcitizens.But the dividingline betweenzeugitaiand thdtesis certainlydifferent from the dividing line betweenpoor and propertied. The numberof thdtesmay have beenas low as (e.g.) 5,000 or as high as 25,000. We simply do not know. 4. Supra n. 1. 5. Gomme(1959, 66-7), followedby Rhodes.Jonesis followedby IsagerHansen,by Ruschenbusch (supran. 2) and by mosthistorianswho mention the numberof citizensin passingwithoutdiscussing the problem. 6. E.A. Wrigley, Populationand history (1969); T.H. Hollingsworth, Historicaldemography(1969); L. Henry, Manuel de d•mographiehistorique (1967); A.E. Imhof, Einfiihrungin die historischeDemographie(1977); M.W. Flinn, The Europeandemographicsystem1500-1820 (1981). 7. Cf., e.g., P.A. Brant, Italian manpower225 B.C.-A.D. 14 (1971); K. Hopkins, "On the probableage structureof the Romanpopulation",Population Studies19 (1966) 245-64; B. Frier, "Roman life expectancy:Ulpian's evidence", HSCP 86 (1982) 213-51, especially247-50. 8. A. Snodgrass,Archaic Greece (1980) 23; O. Murray, Early Greece (1980) 65, 107. Let me addthatapartfrom the exaggerationof the population growth I much appreciateSnodgrass'ideason populationin ancientGreece. What I criticizeis thequantification, notthegeneralthesisof boththesestudies. 9. Cf. Murray 111. On infanticidecf. L.R.F. Germain, "Aspectsdu droit d'expositionen Grace", RD 47 (1969) 177-97; D. Engels, "The problemof female infanticide in the Greco-Roman world", CP 75 (1980) 112-20; M. Golden,"Demographyandtheexposure of girlsat Athens",Phoenix35 (1981) 316-31; W.V. Harris, "The theoreticalpossibilityof extensiveinfanticidein the Graeco-Romanworld", CQ 32 (1982) 114-16. E. Eyben, "Family planning in antiquity", AncSoc11-12 (1980-81) 12 19. 10. Wrigley 70; Flinn (supra n. 6) 76-7. 11. E.A. Wrigley, Mortality in pre-industrialEngland: the example of Colyton, Devon, over three centuries", Daedalus 97 (1968) 246-80. 12. J. Steffensen,"Islands Folkema•ngdegennemTideme", Medicinsk Forum16.5 (1963). Fornumerous otherexamplesof slowlygrowing,stationary or decliningpopulationsin medievaland early modemEuropecf. the works referredto in note6, whichgivefull references to thespecificinvestigations. 13. P.C. Matthiessen,BefolkningensVa•kst(1979) 140. The formula, ac-

cordingto Matthiessen, is 2 x e198ø x .02. 14. Arist. Ath. Pol. 26.4. Cf. CynthiaPatterson,Pericles'citizenshiplaw of 451-50 B.C. (1981). For its disregardseeIsoc. 8.88. 15. Ath. 577B; Schol. Aeschin. 1.39.

16. Cf. E. Ruschenbusch, Untersuchungen zu StaatundPolitik in Griechenland vom 7.-4. Jh. v. Chr. (1978).

186

M.H.

HANSEN

17. Cf. H. Homm½l, "Metoikoi", RE 15 (1932) 1454-5. 18. Arist. Pol. 1275a7, 12; 1277639; 1326a20, b21.

19. It is apparentfrom Arist. Pol. 1275623 that in the fourth centurythe ordinaryGreekpolis(i. e., mostpoleis?)hadthesamerequirements for citizenship as the Athenians. 20. Lys. ft. lxxi (Saupp½);P.Oxy. 2537 verso32 (referencesto threelost speeches by Lysiasdeliveredin graphaixenias);Isa½.3.37; Hyp. ft. xv (Jensen);Din. ft. v (Conomis);Dem. Ep. 3.29; In additionto theseknown applications, the action is referredto in Arist. Ath. Pol. 59.3; Lys. 13.60; Dem. 24.131; 39.18; 40.41; 49.66.

21. Dion. Hal. Isaeus617; hypoth.Dem. 57; Schol. A½schin.1.77; Harp.

s.v. •)ta•O•lq•toL•; Dem. 57.77, 15, 30 et passim;Aeschin.1.77, 86, 114; 2.182; Isae. 12 passim. 22. Cf. A. Billheimer, Naturalizationin Athenianlaw andpractice (1922); M.J. Osborne, Naturalization in Athens. I. Documents, II. Commentaries (1981-82). 23.

Xen.

Hell.

2.3.6-7.

24. Cf. M.H. Hansen,"How manyAtheniansattendedtheecclesia?",GRBS 17 (1976) 125-6. The possibilitythatmorethanonegrantcouldbe ratifiedin a sessionis discussedin M.H. Hansen, The Athenian ecclesia (1983) 21-2.

25. For grantsof citizenshipin literarysourcescf. Billheimer(supran. 22) 110-28; list of naturalizedcitizenswith someadditionsin Hansen, GRBS (supra

n. 24) 128 n. 52. For grantsof citizenshipin epigraphicalsourcescf. Osborne

(supran. 22)D 7-14, 16-28;I believethatIG 112 239canbeadded tothelist. 26. Dionysios(OsborneD 10); Ariobarzanes(Dem. 23.141,202); Cersebleptes(Dem. 12.8; 23.203); Philip of Macedon(Plut. Dem. 22.4); Pasion (Dem. 59.2; 45.46); Phormio (Dem. 46.13). 27. Athenians:Lys. 3.38; metics:Lex. Seg. 194.12-13. 28. Cf. M.H. Hansen,Eisangelia (1975) 35-6. 29. Lys. 20.35; Isoc. 16.47; cf. M.H. Hansen, Apagoge, endeixisand aphegesis(1976) 59 with note 21. 30. Cf. Hansen(supra n. 28) 64-5, dealingonly with strategoi. 31. H.W. Parke, Greek mercenary soMiers (1933); G.T. Griffith, Mercenariesof the Hellenisticworld (1935). 32. 33.

Cf. Parke 28. CIG 4702.

34. Arr. Anab. 1.14.4 (ca. 20,000 mercenaries);1.16.2 (ca. 2,000 prisoners); 1.29.5 (first Athenianembassy);3.6.2 (secondAthenianembassy). 35. Isae. 4; Men. Aspis. 36. PCPhS Suppl. 4 (1977). 37. Only D. Whiteheaddiscusses brieflythepossibilityof Atheniansliving in Chalcis as metics, in "IG 12 39: 'Aliens' in Chalcisand Athenianimperialism",ZPE 21 (1976) 255-6. 38. Euthyphro'sfatherwas probablya cleruchwho stayedon in Naxosas a xenos after the Athenian

defeat in 404.

THE NUMBER

OF ATHENIAN

CITIZENS

187

39. I take Herodesand his family to be Atheniancleruchs(cf. Hansen (supra n. 29) 124) who stayedon in Lesbos(asxenoi) after ca. 424 (cf. Brunt (infra n. 52) 81-4). 40. Basessignedby Attic sculptorsare anothertype of sourcewhich must be treatedcautiously.Some of the sculpturesmay have beenexportedfrom Athens,in othercasesthe sculptormay haveworkedfor a shortperiodabroad and returned to Attica when he had finished his commission. Cf. A. Stewart,

Attika (1981), passim 41. Abdera: ©ctX•ctOZoq'A0svcffoq. Saec. v. SEG XV 420. Aegina: 'A[k/•ctv]6p• 'A[k•]•6v6pov 'A0•lVCt•, Xctrp•.No date. IG IV

Aegina: Aegina:

IG IV Alexandria: Alexandria:

159

'AvxtoxdtZrlq'Axdtp[•ov 'A0rlVCffoq.No date. Archaic (?). IG IV 50 'HXt6&opoq Mooxicovoq 'A0rlvcffe, Xcffpe.No date. 116

'Axo)O•obcSpov zoo 'Aptcs'rdtpzov 'A0rlvaiov b/IToooq. 2717 SEG XX 504

'Aox•,vlm.dt[6o]v. ToOZqvcovo•'A0vlvct•ov•rp•o•3•v (To½). 215/47 SEG XX 505

Anaia:

Chalkis:

Zcfozpct[zoq]TrlXo•cX•[ovq]Ei'recffoq.Saec.iii. JOEAI 11 (1908) Beibl. 163 Adttxxoq ©06tocovoq'A0vlVCffoq.Saec. iii-ii. SEG XXVII

566

Chios or

Erythrai:

A6'r6Xv•coq'AozucX•:ovq'A0vlVCffoq. No date.'A0•v5 20 (1908) 244.69

Ephesos: Eretria: Eretria: Eretria: Halonnesos:

'Ava•t•O•zBq •q•wvoq 'AOBva•oq.No date.Inschr. Eph. VI 2210A Xatogaq Bdzz•voq 'A0•va•oq. Saec.ii. IG XII 9.838 Xato(ov 'A0gva•oq. Saec. vi. IG XII 9.296 ABgqzOtoq Hoogtb•v•ov. Saec. ii/i. IG XII 9.289

--ZIA--•OZTH XXV

Halos:

['A]0gva•a (?). Hellenistic? SEG

1000

Karpathos:

Xato•aq 'AOBva•oq. No date. IG IX 2.120 ['A]ya0•xxov zoo •tXo•pdzsvq 'AOBva[•ov

Kyzikos:

(?) q]p•mv. No date. IG XII 1.986 KzBo•aq Btavopi6o 'AOBvaroq. Saec. iv. Inschr. Kyzikos I 271

Mytilene:

EO•X0ev Bu0i==ou 'A0•va•oq. Saec.iv. IG XII 2.306; cf. Kirchner, PA 5453

Mytilene: Naxos:

'Apstv•aq AvotrpSzovq 'AOBvaroq.No date.Saec.iv? IG XII 2.30?; cf. PA 6?3 'Avzt6Xou ' Hpa•Xsibou 'A0•vaiou. Hellenistic?IG XII

5.84

188

M.H.

Olbia:

HANSEN

'Ap•ce•0fov Kvl•0toof6xov'A0vlvaiog.Saec.iii. Inscr. Olbiae (1968) 98

Rhodes: Rhodes: Tenedos:

'A•0po•S/otog 'A0vlvaiog.No date.IG XII 1.387 Actg& 'A0vlva/a. No date.IG XII 1.388 II•[tqot•,oq'A0rlva•oq.No date.IG XII 2.642

42. Cf. IsagerandHansen(supran. 2) 217-19:Attic gravestelaiof thefourth century BC for Athenian metics. The epigraphicalevidencefrom Corinth, Olynthos,and other excavatedcities indicatesthat meticsand foreignerson the whole tended to follow the custom of the place, regardingfuneral monuments,and not that of their patris. 43. In Corinth,only some400 inscriptions antedatingthe Romanimperial periodhavebeenfound. They havebeenpublishedin CorinthVIII 1 (1931), ed. B.D. Meritt, andVIII 3 (1966), ed. J.H. Kent. In Olynthosthe excavators foundsofew inscriptions thattheycouldnot evenfill a fascicleof theexcavationreport.Theyhavebeenpublished in AJAandTAPAandin otherperiodicals. 44. Cf. Whitehead(supran. 36) 5: ' ... as longas we remainignorantof crucial variables ... it is mere pretenceand delusionto imagine that the economicanddemographic factsof life threwup anAthenianmetoikiaanywhere but in Athens."

45. Andpc. 3.12, 14; Xen. Hell. 5.1.31.

46. IG II 2 1634 (390/89, 389/8). 47. IG II 2 1609.88 if. For the date cf. R. Sealey, Phoenix 11 (1957) 95-9.

48. TimotheusconquersSamos:Isoc. 15.111; Dem. 15.9. Cleruchssentin 366/5 (Diod. 18.18.9; Arist. fr. 611.35); in 361/0 (Aeschin. 1.53 with Schol.); and in 352/1 (Philochorus,fr. 154). Cf. furthermore:Arist. Rhet. 1384632-5;

Strabo 14.1.18; Diog. Laert. 10.1; Diod. 18.8.7; IG II 2 1952; 1437.20; 3207.20.

49. TimotheusconquersPotidaea:Isoc. 15.113. Cleruchssentto Potidaea in 362/1:

IG II 2 114--Tod

146.

50. Cleruchs sent to the Chersonesein 353/2 (Diod. 16.34.3-4; IG II 2

1613.297-8.Cf. Dem. 23.103; 7.43-6) andin 343/2 (Dem. 8.6, hypoth.1-2). Cf. furthermore:Aeschin. 2.72; Isoc. 5.6; Dem. 12.16; IG 112228.13-16 = Tod 174.

51. IG II 2 1629.165-272= Tod 200. The decreerefersto/1 elq x6v 'Afp/.av 6txot}c/.a(175-7); cf. 6 ol}cto'rqq}cai o[ f:xot}cot(224-5). The establishment of theva6oxa0[tov is thereforeby mosthistorians calleda "colony" (cf. Tod). But vauo'r60[tou [oi}c]•/•ot)}ca'r0to}c•uao0•v['ro]q(220-2) points to an Athenianoutpost,not to an independentcolony. Furthermore,the leadersof the cleruchs sent to the Chersonese are called o[}cto'ra/. in IG I12 1613.297-8.

So the terminologyis not strict,and I believethatthe settlementin the Adriatic was a cleruchy. 52. Cf. Jones(supran. 1) 176; P.A. Brunt, "Athenian settlements abroad in the fifth centuryB.C." AncientSocietyand Institutions(1966) 84. 53. E.g. Aeschin.1.53 (Samos);Dem. 7.10 (Potidaea);Aeschin.2.72 (the Chersonese).

THE NUMBER

OF ATHENIAN

CITIZENS

189

54. Plut. Lys. 13.3. For new regulationsfor the cleruchson Lemnoscf. Hesperia40 (1971) 162-73 no. 23. 55. Strabo 14.1.18; Arist. fr. 611.35. 56. IG II 2 1952.

57. Cf. (e.g.) Michelno. 832 (inventoryfromthetempleof Hera in Samos); IG XII 8.63 (dedicationmade by 20 councillorson Imbros). 58. Dem. 6.20; 7.10; Diod. 16.8.5. 59. Cf. Chr. Habicht, "SamischeVolksbeschliisse der hellenistischenZeit",

MDAI(A) 72 (1957) 156-69 nos. 1-2; idem, "Der BeitragSpartaszur RestitutionvonSamoswfihrenddesLamischen Krieges",Chiron5 (1975)45-50;R.M. Errington,"Samosandthe LamianWar", Chiron5 (1975) 51-7; E. Badian, "A commain the historyof Samos", ZPE 23 (1976) 289-94. 60. Diog. Laert. 10.1. For Euthyphro'sfathercf. supra note 38. 61. Diod. 19.68.3, cf. IG XII 8.7' 6 6flgOq6 'AOBvtxkovxf)v •v Mup•vxit

replacedby 6 6flgoq 6 Mu0tvaftov. 62. Cf. Patterson(supra n. 14) 40-81 with further references. 63.

Cf.

Patterson 70.

64. Cf. the excellentdiscussion of thisproblemin Cartledge(suprap. 173) 307-18.

65. This paperwas first presentedas a lectureread at HarvardUniversity on March 10, 1983. I shouldlike to thank E. Badianfor inviting me to give the lecture,for his critical remarksandfor his carefuleditingof the final version. Furthermore,I shouldlike to thank the Institutefor AdvancedStudy in Princetonfor appointingme a VisitingMemberfor Spring1983,the Commission for EducationalExchangebetweenDenmarkand the United Statesfor appointing me a FulbrightScholarfor thesameperiod,andtheDanishResearch Council for the Humanitiesfor supportingme with a grant-in-aid.