American Journal of Ancient History 9781463237424

Citation preview

American Journal of Ancient History

American Journal of Ancient History

4.2

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 4.2 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 1979 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 978-1-4632-0669-7

Printed in the United States of America

TABLE

OF CONTENTS

P.J. Parsons:The Burial of Philip II? ...............................

97

EugeneN. Borza: Some Observationson Malaria and the Ecology of Central Macedoniain Antiquity ................................

102

John Humphrey: The Three Daughtersof Agrippina Major ............

125

Luciana Aigner Foresti: Zur Zeremonieder Nagelschlagungin Rom und in Etrurien

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

June W. Allison: Additional

Note

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

144

1:57

Frank J. Frost: The Dubious Origins of the 'Marathon' . ............... E. BadJan:The Name of the Runner: A Summary of the Evidence ....

159 163

G•za Alfi51dy:Review-Discussion:Ronald Syme, Roman Papers ........

167

Michael B. Walbank: The Family of PhilagrosErchicusand the Cult of Asklepiosat Athens ...................................

186

TABLE

OF CONTENTS

P.J. Parsons:The Burial of Philip II? ...............................

97

EugeneN. Borza: Some Observationson Malaria and the Ecology of Central Macedoniain Antiquity ................................

102

John Humphrey: The Three Daughtersof Agrippina Major ............

125

Luciana Aigner Foresti: Zur Zeremonieder Nagelschlagungin Rom und in Etrurien

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

June W. Allison: Additional

Note

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

144

1:57

Frank J. Frost: The Dubious Origins of the 'Marathon' . ............... E. BadJan:The Name of the Runner: A Summary of the Evidence ....

159 163

G•za Alfi51dy:Review-Discussion:Ronald Syme, Roman Papers ........

167

Michael B. Walbank: The Family of PhilagrosErchicusand the Cult of Asklepiosat Athens ...................................

186

THE

BURIAL

OF PHILIP

II?

POxy XV 1798, an unimportant Alexander history related to the Romance, was published by Grenfell and Hunt in 1922. They at once recognisedthe Introductionas probablyan odd newversionof Philip II's death and its immediate consequences.No one has since looked at it. Papyrologistslike Wilcken and Bilabel based their restorationson the editio princeps (as did Jacoby for the whole history: FGrHist 148)• and N.G.L. Hammond recently produceda new restoratiomwith important historicalconsequences (GRBS 19 (1978) 343-9), without inspectingthe papyrusor evenreproducingthe originalapparatus.It washigh time for a new edition by a specialistwithout a historicalpartipris. Our thanks are due to Mr. Parsonsfor performingthe task in hisusualexemplaryway, and to him and Dr. R.A. Coles for the photograph (original size) we are reproducing; also to my colleague A. Henrichs for advice. Editor

POxy XV 1798fr. 1 hasbeentakento describethe murderand burial of Philip II; and recent discoverieshave given it a particular interest.At ProfessorBadian'ssuggestion,I offer a new collation of the original, as a firmer basisfor future speculation. The Papyrus. At sometime after the first edition, an unknown hand

(mostprobablyMr. Lobel's)joined fr. 1 to fr. 17,sothat fr. 17.1continues fr. 1.11; the combinedfragment containsparts of fourteen linesfrom the foot of a column.

Line-lengths.The completelines offr. 44 rangefrom 14 to 19letters;15 to 17 letters

commonest.

The line-beginningsof fr. 44 are not vertically aligned, but move graduallyleftward as the columndescends; in fr. 1+17, if the slopewasthe same,the first letter of line I will have stood about two letter-spacesfurther

right than the first letter of line 14 (supplementsmust allow for this). Fr. 1+17.9, 13 and 14 show short blanks after their last visible letters, and

in 9 and 13theselettersare prolongedtowardsthe right; it is likely though not certain

that these are line-ends.

Scribalpractice.The scribeuseddiaeresis,but no accents,breathingsor elision marks; and no punctuationexcept paragraphusand sometimesa shortblank to mark the beginningof a sentence(but the spacingof lettersis so irregular that a visible short blank cannot be assumedto serve as ¸ 1981 by E. Badian. All rights reserved.

97

98

P.J. PARSONS

punctuation, unlessthe context makes it clear). Iota adscript is always written when required. Elided vowels are simply omitted in 8(&) and prepositions;scriprioplena of otherwordsfr. 44 ii 11,ii 13and (at a pause)i 11, iv 8. The division of syllablesover the line-endsfollows normal rules (but note fr. 44 iv 4/o-rosy not POxy 1798 fr. 1+17

1

5

10

foot

THE

I

BURIAL

OF

PHILIP

II?

99

] , foot of vertical,anotherpoint of ink high up to the right before omicron(gammaor tau?).a.., probablyto be combinedasmu. [,

pointatline-level; lowerpartofupright, withendofhorizontaljoih•ng from the left at mid-height(e.g..z¾). 2 [ ], the spacewould allow rho. [, probablyalpha (too angularfor epsilon).

4 . [, bitsof ink mostsuggesting theleftsideoflambdaor mu(0po•zl [[3-, rrep•0po•zl [[3-?);though,giventheirregularity of thewriting,nu may not be excluded.

5

. [, lambda or mu.

6 ]., tail of alpha, lambda or similar.. [, point at line-level,well to the right (e.g. foot of tau; or letter beginningafter punctuation-space). 7 ]., right side of alpha, lambda, mu.. [, vertical on the edge (nu suitable).

8 ]., probablyfoot of diagonal,and rightvertical,of nu (junctiontoo low for alpha iota); then space(punctuation?). 9 ]., tracesof upright on the broken edge.

11 . [, probablyalpha(z.•[q•,?). 12 ]., endof horizontaljoiningsigmaat two-thirdsheight(epsilonrather than alpha?).. [, part of an upright (of vowels, epsilon or eta: ?). 13 ]., top of shortvertical,probablynu (too shortfor iota). 14 ]., probablyfoot of diagonal,and lower part of right vertical,of nu. Christ Church

P.J. Parsons

Oxford

For comparison,we are reproducingthe editioprinceps,with itsapparatus. See next page.

(Editor)

100

P.J. PARSONS

OX YRH

YNCHUS

PAP YRI

XV

(Ed. B.P. Grenfell and A.S. Hunt)

1798. Anonymous work on Alexander the Great Fr.

44

14.3x34.3

cm

Late second century Ft.

1

Fr.

17

10

Fr. 1. The mentionof a theatrein 1.2,in conjunctionwiththeburialof. ]trrrrovin 11.8-10,leaveslittle room for doubt that this fragmentrefersto the death of Philip, but the details are unfamiliar. Philip's assassinwas Pausanias (Diodor. xvi. 94, Justin ix. 6), for whosenamethere seemsto be here no place;moreover,according

to Diodorushe waspursuedand killed forthwithby o[ rrz•[ 'r6,•Hz•8•xxo•,• who •¾xz,rr•o•wrz• &,•zO, o,•. Apparently, then, the object of 0½rrz'r•rr0t,•[t•0½,• is some other person,whoseidentityis obscure;cf. Justinxi. 2.l Prima illi curapaternarum exsequiarumfuit ; in quibusante omnia caedisconsciosad tumulumpatris occidi iussit.

I sqq. The length of the lacunae is estimatedon the basisof 11.8-10,which can be

THE

BURIAL

OF

PHILIP

II?

101

restored withprobability. In 11.1-4.oBq•[$]¾I [$¾.cot0]$(x.[O]cot x0•[ 0•xe[),ul [*e •:ouq(or -co[q) 8]e may be suggested. 5. ]iv: ]co¾ is not possible, and]0•vis unlikely.The doubtful• maybe 6. Boththislineand1.9 look asif theywerecompleteat theend,butthereisnot marginenoughto be certain.If 1.6 endedwith-xe,it wasrathershorterthanits neighbours.

7. The spelling&r•o•:ur•0•co seems to benovel;-c,Sx0•o• is a poeticform. Fr. 17.4. Somecaseof e),0•:[l*og presumably.

SOME

OBSERVATIONS ON MALARIA AND ECOLOGY OF CENTRAL MACEDONIA

THE

IN ANTIQUITY

In the first volume of his history of Macedonia, Hammond describesthe central (Emathian) plain of Macedonia in the time of Philip II thus:

One imaginesthat the central plain was at that time intensively cultivated by men and women who walked as far as 20 km from the citiesto their fields in the plain and lived in the fieldsat the busiest seasons.The climate was healthy, and malaria wasas yet unknown.• How different this seemsfrom the desolate and malaria-ridden country

widely portrayedby nineteenth-and early twentieth-centuryvisitorsin the regiombeforethe modernagriculturalrevolutionchangedthe face of the landscape.At another time I hope to deal at lengthwith the matter of cultivation in the œmathianplain in antiquity. At present,however,it may be usefulto attempt to understandsomethingof the physicalconditionof this plain, which wasthe heartlandof the Macedoniankingdom.There are two immediateconcerns:the incidenceof malaria in the regionin antiquity, and the topography of the plain itself.

The history of malaria in Macedonia isessentiallythe sameasthe history of the diseasein Greece.There are three methodsof determining the prevalenceof malaria in antiquity. First, a considerablebody of testimonyabout the affliction survivesin the ancient writers. Second,modern techniquesin

paleo-pathologyas a branchof the historyof medicineenableus to trace the spread of this scourge.Finally, one may argue by analogy from the modern experiencewith malaria where it can be demonstratedthat the ecologicalconditionswhich nurture the diseasehave not changedsignificantly from antiquity. Malaria is an infection of the blood by a minute plasmodiumparasite.2 The parasitesmultiply rapidly and destroy red cells. Victims normally sufferseverefevers•generalmalaise,and sometimesdeath,dependingupon the age and general health of the victim and the particular speciesof plasmodiumparasite.Until recentlythe agentresponsible for transmitting the diseasewas unknown. European colonial interest in the tropical malaria belts of Africa and Asia in the late nineteenth century led to

102

OBSERVATIONS

ON

MALARIA

0

10

I

I

103

20 I

30 J

miles

EMATHIA

ICENTRAL MACEDONIA) 4TH

CENT. B C

Ed,

Cs014 S

.//'•..(Nea• •f.(YI4A/A/Its4) .

Nikom_edeia) ,d*

•"f

"...

8eroia

THERMAIC o

enb

MT. OLYMPUS

Dion

GULF

104

œ.N. BORZA

considerablescientificinvestigationinto the affliction. In July 1898, a British researcherworking in India. Ronald Ross,describedthe resultsof his observationsin a letter to a friend: "Malaria is conveyedfrom a diseased person or bird to a healthy one by the proper speciesof mosquito and is inoculatedby its bite."3-Ross'discoveryled to his beingawardedthe Nobel Prize for medicine

in 1902. 4 Italian

scientists learned later that the mos-

quito genusAnopheleswas the culprit, and recentresearchhas shownthat of the nearly 3000 known speciesof mosquito,about 375 areanopheline,of whichmore than 70 are vectorsof four speciesof humanmalaria.5 Several anophelinespeciesand subspecies are responsiblefor human malaria in Greece.6 The Anopheles'habitat and breedinggroundsare well known: they are mainly (but not exclusively)marshyareas,regionsof humidity and still or slow-movingwater,wherethereare meandaily temperaturesabove 60ø F--conditions necessaryto incubate the mosquito. These conditions are conduciveto endemicmalaria in everyhabitablecontinentin the world. Modern Macedonia• with its marshy areas in the Emathian and Strymonian plainsand along the lower Axios river, is well-known as a malaria center.7 In Macedonia malaria wasespeciallyseverebecausethe two main vectors,A. sacharoviand A. superpictus,are complementary.A. sacharovi breeds in marshes,and in Greece has adapted to the brackish salt-water marshesin coastalareas, where it flourishesin early and mid-summer. It is a mosquito which bites man readily, and is the chief vector of malaria in regionswhere it exists.A. superpictusis a secondaryvector, and is often found in the Mediterranean basin and Near East in conjunctionwith A. sacharovi.A. superpictusis a late-summerand autumn breederin foot-hill streams which are reduced in volume

and flow-force

at that time of the

year. These two malaria vectors,adapted as they are to different types of water surfaceswhere each may breed and thrive in conditionsunfavorable to the other, combine to produce a transmissionseasonwhich in Greece lasts from April to November. The transmissionis thereby intensified through a long seasonover a range of ecologicalconditions;the result is that only those areas of Macedonia marked by high elevationsand cool temperatures would be free from malaria. Moreover, the winter pause in transmissionand occasionaldrought amidst normally abundant rainfall interrupts the transmissionprocess, which in the tropics enablesthe adult population to build limited immunities to the disease.Much of the Balkans,including Macedonia, has thus sufferedthrough the deadly"endemic-epidemic"cycle which hasproved so costly to human health and life (see notes 6 and 7). A malaria crisisstruck the area in the early twentiethcentury. In 1916-18 Macedoniabecamea major military front, and the susceptible personnelof the British, French and German armies were laid low. The situation was

OBSERVATIONS

ON

MALARIA

105

exacerbatedby the fact that only a few yearsbefore, in the aftermath of the Balkan Wars, about 150,000 Greek refugees, many from malaria-free regionsof the Balkansand Asia Minor, had settledin Macedonia and had becomeinfected.The paralysisof three modern armiesand a large civilian population before the ancient disease was a tragic lesson in human vulnerability. Malaria has disappeared from Greece only since the early 1950s. 8

Ross'successin identifyingthe mosquitoas the vehiclefor transmitting the malaria parasite,and the rapid medicaladvanceswhichfollowed his discovery,resultedin an especiallyhigh interestin the subjectin the early twentieth century. Impressedwith Ross' work, the Cambridge classicist W.H.S. Jones produced two volumes investigating the prevalence of malaria in classicalantiquity.9 Jones was correct in calling malaria a "neglectedfactor" in ancient history; ironically, his work has had much more influence on the students of epidemiology and the history of the disease than on the classical scholars for whom it was intended.

The main

criticism of Jones'work is that, having collectedthe ancient testimoniaon malaria, he drew a number of inferencesabout social, moral and economic

decline that appear naive to a modern public. Jones' views about the "degeneration"of the Greek character,the "lossof brilliance"after ca. 400 BC as the result of malarial infection, are based on outmoded racial and

social premises.•0 Ross himself subscribedunreservedlyto Jones' views, and argued that malaria was responsiblefor the "decline"of a vigorous Greek civilization.And no lessan expertthan the presentleadingscholarof the paleo-epidemiologyof malaria restatedJones'thesisthat after malaria becameendemicca. 500 BCit was probably responsiblefor the downfall of

Greek civillzation.• Simplisticcultural notions like "degeneration"and "decline"aside,the value of Jones'work was hiscomprehensivecollection of testimony from the ancient medical and non-medical writers. All subsequentaccounts of the history of the disease in ancient Greece ultimately dependon Jones'Malaria and Greek history. The weight of the ancient testimonia is impressive.It is collectedand commented upon in Jones in detail, and there is no need to set it down here;•2a summary will suffice. The common ancient Greek word for fever waspyretos(rrupe'r6C), and by thefifth centuryBCit is clearthat, exceptin a few specialcases,the useof the word normally refersto malaria.•3"Fever" was well known after the mid-fifth century, and that it was often malarial fever (as opposedto other types) is evident for two reasons.One is the ancients' associationof the affliction with marshy areas and seasonal attacks,correspondingto modern experiencewith the disease.The other is the description of fever stages,degreesof severity (until recently the modern diagnostic terminology for malaria was based mainly on the

106

E.N. BORZA

phrasesin the Hippocratic corpus), and the condition known as splenomegaly (enlargement of the spleen), which is one of the observable symptomsof the malady. It is obviousfrom the Hippocratic corpusand from the medical writers of the following four centuriesthat the ancients were so well acquaintedwith malaria as to describesymptomsin terms clear enoughfor modernmedicalscientiststo recognizewithout doubt.TM

Two challengesto Jones'view have emergedin recent years. One utilizes modern techniques in paleo-pathology; the other reveals underlying assumptionsthat the physicalenvironmentof the Emathian plain was not productive of malaria. Both argue that Greecewas in fact virtually free from malaria in the classicalperiod. First, J. LawrenceAngel has offered the materialsof physicalanthropology-skeletal remains in particular--to suggestthat the evidenceof patho-physiologicaladaptation in responseto malaria is lacking for classicalGreece.15Angel'sthesisisbasedon the connectionhe seesbetween malarial environment,boneenlargementand thalassemia.Thalassemiais a blood disorder resultingin severeanemia. It is a recessivegenetictrait which can affect the individual's ability to resist malaria. Like a related disorder, sicklemia("sickle-celldisease"),thalassemiaprovidesthe victim with a level of protectionagainstthe malaria parasite.It wasobservedlong ago, for example, that Blacks in Africa and the American South seemed generallymore resistantto feversthan Whites who worked in mosquitoinfested areas. Many Blacks are traditionally affected by sicklemia, a debilitating and often lethal genetic blood disorder. Sicklemia harbors fewer malaria parasitesby incidentallyprovidingan unfavorableenvironment for the parasite.•6 Like sicklemia, thalassemiais often fatal, and would logically disappearthrough natural selectionby the death of the host, but it curiouslypersistsamongpopulationswhich are also subjectto malaria. By a mechanismnot understoodat the moment, the presenceof malaria infections activates certain resistant factors which provide a protectioncounteractingthe defectof thalassemiaand othergeneticblood disorders.A malarious population is sometimescharacterizedby a relatively high frequencyof other blood disorders,especiallythose red-cell abnormalitiesthat seemto exist in a symbiosiswith malaria.•7That is, the continuingexistenceof a population survivingsuchafflictionsas thalassemiaand sicklemiamay indicatethe presenceof malaria as well, even while malaria may alsoexistin populationsnot sufferingfrom otherblood disorders.

OBSERVATIONS

ON

MALARIA

107

A third important disorder often coincideswith malaria, the glucose-6phosphate-dehydrogenase (G6PD) deficiency. Known as "fayism", this hereditary blood defect producesa severetemporary anemia when the affectedindividual ingestsor inhalesthe pollen of a broad bean(viciafava). Exposure to the fava producesa deficiencyin the vital G6PD enzyme which normally arreststhe deterioration of red-cellmembranes;the result of favismis a transientacutehemolyticanemia.•sAn advantageous effect of the malady is that, as in the casesof thalassemiaand sicklemia,a hostile environment for the malaria parasite is also created;that is, an individual may continue to be affected by the geneticblood disorder, but will suffer less from

malarial

infections.

There is a considerableliterature on the use of the fava among Greeks and other eastern Mediterranean peoplesas the focal point of ritual, cult and kinship-systems. •9There isalsoa notorioustaboo on beansto be found in the philosophicalwritings, especiallyamong the Pythagoreans? It is problematic whether the existenceof widespreadbeliefs and practices relatedto the beansuggests that a significantsegmentof the populationof the easternMediterranean sufferedfrom a G6PD deficiencyoften associated with endemic

malaria. TM

Within a population, individualsor groupsmay or may not be affected by these blood disorders, including thalassemia.In those who are thalassemic,however,an adaptive individual physiologicalresponseto their condition may occur:certain of the body'sbonesenlargeas the internal spongymatter producing red cellsmust of needincreaseto combat disease. The enlargedboneconditionis knownasporotichyperostosis. On the basis of skeletalexaminationsAngel reportsthat there is a substantiallyhigher incidence of porotic hyperostosisamong farmers living in marshy areas--e.g., the sixth-millennium BC Macedonian settlementnear Nea Nikomedeia in the southwesternpart of the Emathian plain--than among those living on higher, drier ground. Angel claims a generallyhigh incidence of porotic hyperostosisamong personswho lived in the malarial belts of the Old World, although the statisticalbasisfor this view is not made

clear.

Angel also suggests(again on the basis of skeletalexaminations) that there have existed fluctuating levels of malaria in the eastern Mediterranean,causedby two factors:one is the periodicdraining of swampland; the other is the occasional variations in climate, which result in lower sea

levelsanddrierlandin normallymarshyareas?2During such"dry"periods the incidenceof malaria would at least in theory decline. Now, to the crux of Angel'sargument.The incidenceof porotichyperostosisappears to be variable from time to time. For example, the

108

E.N.

BORZA

prehistoric Macedonian farmers at Nea Nikomedeia were presumably malaria-ridden, accordingto Angel, becausetheir bonesshowevidenceof porotic hyperostosis,an indication of the thalassemiawhich is sometimes symbioticwith malaria. Angel claims a 50% incidenceof hyperostosis among the late Paleolithic boneshe has studied,but the incidencedrops gradually to only 8% for the Greek BronzeAge, 4% for the Archaic(early Iron) Age, and then to virtually "no malaria" for the classicalperioddown to Ca. 300 BC.23There is a marked upward swing in the Hellenistic era (10%), and by Roman timesmalaria has again becomeendemic(24%) in the eastern Mediterranean.

Any attempt to establisha necessaryrelationshipbetweenmalaria and thalassemiais a complex problem, and there are some difficulties with Angel'sthesis.24First, the sampleof bonesstudiedis small, beinglimited to the availability of materialsfrom a few archaeologicalinvestigations.The lack of a broadly-basedstatistical sample--in both geographicaland chronologicalterms--makes one feel insecurein acceptinga thesisdescribing fluctuationsin a major, widespreaddiseaseovera longperiod of time.25 Next, Angel assumes"that much of this anemia [that which may have causedthe bone enlargement]was thalassemia".26But thalassemiais not necessarilya function of climate or environment,and Angel appearsto have overlooked the fact that thalassemia(which always producesbone enlargementas an adaptiveresponseto infection)is also prevalentin some non-marshy areas where the bones may enlarge to combat some nonmalarial infections.Indeed, one moderndiagnosticstudyin modernGreek villageshasshownthat thefrequenciesof thalassemiain malariouslowland villages and non-malarious mountain areas were approximately the same.27Further, evenif the anemiacausingboneenlargementwereproven thalassemia--which

it is not--there

are other than malaria-related

factors

which can causeporotic hyperostosis. 28 In sum, there is an apparentcoincidencebetweenporotic hyperostosis and marshyareasin parts of the Old World as evidencedin a few ancient sites (as Angel points out); these regions may or may not have been malarious.The existenceof boneenlargementmay or may not beevidence of thalassemia,and the existenceof thalassemia,evenin a marshyarea, is not a proof of the existenceof malaria (seenote 27). That is, wecannotuse the bones alone to show the existence of malaria.

Angel has suggestedthat there is a correspondence betweenthe datable low incidence of porotic hyperostosisand minor climate fluctuations resulting in drier climate (hence fewer marshesand less malaria). As evidence he cites the work of Denton

and Porter

on climate variations. 29

Denton and Porter discuss "neoglaciation", that is, the post-Ice-Age climate fluctuations evidencedby the occasionalretreat and advance of

OBSERVATIONS

ON

MALARIA

109

glaciers.Glacial growth and shrinkageover the past five millennia reflect the slight reversibleshifts from the warmer/wetter range to the cooler/ drier range.3ø Pollen records, Carbon-14 dating• botanical analyses, changesin the habitat of man and animals, agricultural data and, in the historical period, eyewitnessaccounts,are among the kinds of evidence used to document theseclimatological phenomena. Denton and Porter's data indicatethat the first millenniumBe wasmarked by a glacialadvance whichpeakedabout 500 BC.Thus the climatewascoolerand drier than that of the Hellenisticperiod whichfollowed. By ca. AD 500 the glacierswerein marked retreat beyond even presentconditions. Angel seesa correlation betweenthe incidenceof porotic hyperostosis(malaria'?.) and glacialfluctuations(climatevariations).He usesthe climatedata to suggesta reasonfor the lower incidenceof bone enlargement:that is, a drier climate produces less marshland, hence less malaria. 3•

That there appearsto be a correlationbetweenAngel'sdata on porotic hyperostosis and minor climatechangesis not in dispute.But Angel goesa step further in arguing that marshinesswas further reduced becausethe Mediterraneansealevel was 3-4 meterslower than presentlevels,as some Mycenaeanand classicalsitesare now 1-2 metersbelowsealevel.Whatever the value of sea-levelargumentsin archaeology(e.g. local land-subsidence can also produce underwater sites, for which see p. 111 below), the particular caseAngel makesis dubious. For example,he usesthe caseof underwater"Mycenaean and Classical"(my italics) sites,even though the data published by Denton and Porter show that much of the second millennium BC was a period of relative glacial retreat, that is, higher sea levels according to Angel's argument (are we to supposethat the Mycenaeans deliberately constructed their sites underwater?), whereas classicalglaciers were at the peak of advance (lower sea levels). One cannothave it both ways.32Angel'sthesisis not supportedby the evidence he cites.

Finally, Angel correlatesthe declineof malaria (presumablyevidenced by porotic hyperostosis)with the increasein population and more progressivemethods of agriculture between the prehistoric and classical periods.33One must weigh this againstthe view that with an increasein populationand consequentstrippingof forestand cover-landfor agricultural use,natural surfaceabsorptionand drainageare interfered with. The result is an expansion of the breeding grounds for anopheline mosquitoes.34Sucha transformationof the countrysidefor humanusedoesnot necessarilyproducemalaria, for the mosquito-vectormustbe infestedwith the deadly parasite;but one may suggestthat, on ecologicalgroundsalone, the classicalera offeredmore potentialfor the scourgethan did the earlier periods.

110

E.N.

BORZA

III

We may now turn in detailto the matter of the sealevelin antiquity, both for its generalinterestand for its connectionwith Macedonian marshland, particularly in the Emathian plain. In his History of Macedonia, Hammond wrote (I 145): In antiquity the level of the seain the Mediterranean was some5 feet lower than it is today, and this meansthat the high flood levelin the central plain, e.g., in the vicinity of Alorus, was5 feet lower and that the rivers had that much more fall.

The questionof sea-levelchangessinceantiquity isvexedand a matter of continuing concern to archaeologistsand oceanographersalike. For his view that the level of the sea was lower in antiquity Hammond cites the evidencein his earlier article on the battle of Salamisand in his Ei9irus.35 The evidencein the Salamis article consistsof referencesto suggested North Sea level changesand opinionsof local Greek seamen.Hammond admitted that the matter was problematic,and assumed(for the sakeof his discussionof the topographyof the Straitsof Salamis)that the sealevelhas risen 5-6 feet since antiquity. In his account of the battle of Marathon published twelve years later,36 Hammond made no referenceto sea-level changesin an otherwisedetailed discussionof topography. At Marathon five or six feet lessdepth of water would have significantlyaltered the coastline of that shallow bay and perhaps affected both fresh-water marshesand the sea-water lake. We cannot assumethat Marathon Bay in 490 BCwasas it istoday, but that the levelof the Aegeandroppedfive or six feet by the time of Salamis a decade later. In Epirus Hammond cited six different pointsalong the westerncoastof Greece, between the Gulf of Arta and the Gulf of Valona, where ancient

remainscan be seenbeneaththe surfaceof the sea.Hammond'sargumentis based entirely upon the existencetoday of these submarineremains, all consistently5-6 feet below the presentsurface,and he concludesthat the sea level was from three to five feet lower in antiquity. This view figures prominently in Hammond's reconstruction of the topography of the Emathian plain and the central Macedonian coastline.It evencorresponds (although he doesnot mention this himself)with his view statedelsewhere (seenote l) that Macedonia wasfree from malaria in Philip lI's time, on the assumption, of course, that a lower sea level would have produced more efficient alluvial drainage• hence lessmarshland in the Emathian plain. There now existsa body of scientificresearchinto the matter of sea-level changes. N.C. Flemming and others have devised a method using the

OBSERVATIONS

ON MALARIA

111

resourcesof both archaeologyand geology, which attempts to establish averagesea-levelmeasurementsexclusiveof wave and tidal fluctuations. Analysisof the placementand type of coastalsitesfrom antiquity, when joined with a studyof attendant land forms with a known geologichistory, has producedsomeconclusionsabout the ancientsealevelrelativeto local coastlines. Flemming studied 69 Aegean sites;this survey, when taken togetherwith his earlier researchin the westernMediterranean, showsthat the last great sea-levelchange occurredin the Mediterranean 10-11,000 yearsago, and that sinceabout 2000 BCthe levelof that seahasbeenwithin a few centimetersof presentconditions? The world-widesealevelhasnot increasedmore than about 30 cm (about one foot) in the last 3000 years. Any ostensiblelargechangeasevidencedby numeroussunkensitesis more likely the resultof local volcanicand tectonicsubsidence, commonin this geologicallyactivepart of the world. It is thuslocalearth movementwhich accountsfor the apparent rise in the Mediterranean sea level; it may be understoodthat this subsidence of landformsis relativeto a virtually stable sea level. 38

One final note on climate.That minor periodicclimatechangesoccuris beyonddispute;beyondthis little elseis certainabout climatefluctuations. Since precisemeasurementsof climatologicalconditionsare lacking for ancienttimes,we are totally dependentupon the occasionalobservationof phenomenain our written sources,the dating of organicmaterials,botanical analyses,changesin human and animal habitat, agricultural data and mountain glacier variations. These data can indicate long-term trends or abnormal variances,but in the absenceof detailedrecordstheycan provide only guidelinesfor describingthe climatic history of a singleregion like Macedonia.

39

The literature of antiquity, however,suggests that ancient Greece'sclimate wasnot significantlydifferent from today's.Hesiod'sgrowingseasons in Boeotiacould providea guideto a modernplanter;Theophrastus'plants still grow in the sameregions,though many are reducedin number owing to human mismanagement;someforms of wildlife are long extinct, but domesticanimalscontinueto flourishin the sameregionsas in antiquity;the capesat Malea and Athos still blow fierce, and as recentlyas 1971 the author witnessedAthens"crownedin violet", as Pindar put it.4øAberrations aside, we can assumethat the climate of the Greek peninsulain antiquity was about as it has been in modern times, when malaria was endemic

in Greece.

Thus the views of Hammond and Angel on the physical conditions conduciveto malaria, and of Angel on the link betweenthe bone evidence and the disease,seemunconvincing.It seemsmore prudent to acceptthe testimony of the ancient writers about the prevalenceof malaria. It is

112

E.N.

BORZA

unlikely that so much preciseinformation could be transmittedabout a diseasethat did not exist. Moreover, it will be shownbelow that the ecology of the Emathian plain in the classicalperiodwasaboutasit wasin theearly twentieth century (except for some coastlinealterationsresultingfrom alluvial processes),when the region was notorious as a malaria center. IV

When then did malaria enter into Greece?The original home of Anopheles may have been the Ethiopian regions,and it may have precededman's existenceboth thereand in othertemperateclimates.4•It undoubtedlytook time for the relationshipbetweenthe plasmodiumparasiteand the host Anophelesto develop;preciseecologicaland physicalbalanceisnecessary. The mosquito-borneinfectionis thoughtto haveprogressed from its East African genesisdown theNile valleyand eventuallyinto the Mediterranean and Near East. 42 It is clear that the disease was endemic in the Greek world

by the fifth century,sinceit waswell known after mid-centuryby both the writers of the Hippocratic corpusand non-medicalwriters alike.43 To search for referencesto malaria in the pre-classicalperiod is to confrontthe usualproblemof historicalresearchin that era:thepaucityof literary evidence.Nonetheless--andto risk an argumenturnesilentio--it is curious that our best early source for practical matters, Hesiod, who complainsabout everythingelse,fails to mention the fever in his work. Perhapshis town of Ascra was high enoughabove the Boeotian plain (where malaria wasrife in the early twentiethcentury)on the slopesof Mt. Helikon to have escapedthe affliction, although modern Anopheleshas spreadits scourgein Greekvillagesup to 600 m elevation.It isalsopossible that malaria had not yet arrived in Greece. Whatever the case may be, Hesiod is silentabout the feverswhen we would expecthim to be otherwise. Also silentare a seriesof votive tabletsat Epidaurusdating from the early classical period. Sigerist summarized some seventy casesof afflictions noted on these tablets, and Bruce-Chwatt concluded that none referred to malaria .44

If it iscorrectto suggestthat thereis no evidenceof malaria in at leasttwo early sourceswhereonewouldexpectsomeindicationof itsexistence,and that there is clear evidenceof malaria in fifth-century Greece, we may postulatesomepost-Hesiodicand pre-Hippocraticcircumstancethat was favorableto the introductionand consequentspreadof malaria in Greece. A singleevent may supplythe answer:the entry into Greecein the early fifth century BCof a large number of Asians, many of whom camefrom areas of the Near East with endemic malaria. 45 Did Xerxes' armies or the

military and administrative personnel who precededthem carry the

OBSERVATIONS

ON

MALARIA

113

scourgethat would eventuallyattract so much attention from the Greek medicalwriters?Perhapslow incidencesof malaria had alreadyexistedin parts of Greece.One assumesthat malaria inevitablywould havebecome endemicin Greeceduring the courseof its inexorable spread.Yet one is temptedto suggestthat the Persianarmy may have hastenedthe process, especiallyin thosemarshyareaswherethe hitherto uninfectedanopheline mosquitoesalready flourished, which, now having an infected Asian population to prey upon, spreadmalaria to the local population.46The residentsof northern Greece,especiallythe inhabitantsof what in modern times was the malaria belt of Thrace and Macedonia, may have suffered morefrom the bite of the newly-infectedmosquitoes thanfrom thepassage of a vast Asian army through their land. There are other possiblesourcesfor the infectionand for its spread.The gradual strippingof the land of its natural covercreatedmosquitobreeding-grounds. 47Increasedcommercewith malariousregionsplus the occasional mingling of large groups of people from throughout Greece at panhellenicfestivalsmay have contributedto the spreadof the disease. There can be no certainty about thesematters,but if we acceptthe notion that malaria was endemicin Greeceat leastby the fifth century--whatever its origin--and we believe that the Macedonian environment was as conducive to the affliction then as in modern times, we have little reason to

doubt that it was a factor in Macedonian history. What effect this scourgehad on the Macedonians is difficult to say. The diseasetakes its greatesttoll among those who are newly exposed,as was the casewith the highly susceptiblerecently-arrivedpopulationsin Macedonia in the early part of this century. Long-term inhabitants in heavily endemicareasdevelopsomeformsof resistance,but theseare idiosyncratic and weak, as studiesof modern Greek mortality ratesand incapacitation among Macedonian villagers have shown. An infected population is unhealthy.The major effectof malaria on a population--after producinga high infant mortality rate--is to reducethe residents'work efficiency.As modern experiencethroughout the world has shown, populationshave lived and worked for centuriesin malariousregions.4sOne suspectsthat in antiquity the inhabitantsof Macedoniakept to the highest,driestground available, living and working on the terrace lands bordering the deadly marshes which

were the center of ancient

Emathia.

49

Macedonia shareswith the rest of Greecethe fact that only a small part of the total land area is suited to agriculture. The combination of scant rainfall, sharplyerodedland forms whichdo not hold soiland water,and a

114

E.N. BORZA

large proportion of mountainousterrain reducesthe amount of tillable and grazing area. These factors are somewhat mitigated in Macedonia, however, by its large alluvial plains, relatively more abundantrainfall throughout the year, and cultivableterracelands on mountain slopes.Moreover, Macedonia'smajor rivers flow year-round, permitting both natural and artificial irrigation for crop and pasture land. As an agricultural area Macedoniais, comparedwith muchof Greece,blessedby nature.The large mountain ranges also provide abundant well-watered summer pasture slopesand basins,a phenomenonwhich has sustainedMacedonian agriculture for millennia. Until the agricultural revolution of the past halfcentury, Macedonialooked muchasit had in antiquity. Two major natural factors particularly affectingriver-plain cultivation in central Macedonia are climateand sealevel. We havesuggested that both in moderntimesare virtually what they were in antiquity. As for human attemptsto alter the landscape,Hammond claimsthat Philip II was responsiblefor a flood control project in the Emathian plain.50No direct evidencefor thisexists,althoughTheophrastus(de Caus. Plant. 5.14.6) mentionsthat Philip drained and reclaimedthe land around

Philippi. This area, the lowerpart of the plain whichbeginsaboveDrama and runs down to the Kavalla coastal ridge, is hardly analogousto the central plain. The Philippi plain is an ill-drained alluvial basin,fed by the numerousstreamsfalling from nearby mountains.The plain'sdrainage system narrows to a single stream where the Angitis river piercesthe Pangaion-Menikion mountain barrier to flow into the lower Strymon plain. The drainage of the Philippi plain was accomplishedwith relative easein modern times by the erectionof a pumping station at the Angitis bottleneck.5• If Philip was responsiblefor draining the Philippi plain shortlyafter he took it in 356 BC,he may havesolvedthedrainageproblem in someequally simplefashion. The Emathian plain, however,is a different matter. It is huge and complex, fed by ever-flowingmajor rivers, and resistsany singleeasy method

for flood

control

and reclamation.

In modern

times it took the

combinedeffort of the most advancedAmerican and Greek hydraulic engineeringskillsand considerableamountsof moneyto alter river courses and drain swamps,a processoccupyingmuch of the two-decadeperiod 1920-40and continuingfor severalyearsafter the end of World War II and the Greek Civil War. One doubtsthat ancientMacedoniantechnologyand the royal purse were up to this formidable task, even on a more limited

scale.That the regionaround Philippi was reclaimedmay or may not be true. In either case,Philippi'sdrainagecannot serveas an analogfor the great central plain. If the sealevel were severalfeet lower in antiquity, it might be arguedthat drainagewas naturally more efficientthan it is today, and thusthe Emathian plain waslessswampy.But we haveseenthat the sea

OBSERVATIONS

ON MALARIA

115

level in historical times has remained virtually unchanged;moreover, the silting processby which the plain was formed (for which see below) indicatesthat it was ill-drained in antiquity. It is thus best to assumethat the area remained undrained, as it was until the modern program of reclamation.

The historical geographyof the central plain's metamorphosisover the centuriesis too complex to deal with here.•2 As late as the classicalperiod an inlet of the Thermaic Gulf extendedquite far westinto what becamethe central plain, muchas the seastretcheseastwardtoday to form the Gulf of Salonica(seeMap). No evidenceof prehistoricsettlementhasbeenfound in the area?3 and all the known sitesof the historical period--including Aigai (Vergina), Beroia, Mieza, Edessaand Pella--lie on the adjacent terrace-land.•4 The alluvial activity of the four rivers which flow into the region--the Haliakmon from the southwest,the Moglenitsas (Loudias) draining Almopia, the Axios and the Gallikos (anc. Echedoros)from the north--plus innumerable small streams draining the slopes of Mt. Bermion, gradually began to silt up the inlet. In the fifth and fourth centuries•3cEmathia wasstill largelya seainlet and marshes,with the main route from Tempe and the Pierian coastal plain hugging the adjacent piedmont. It was not until Roman times that lower Emathia was able to supporta road directly acrossthe deltasfrom Pieria to Salonica?•and that the plain assumed roughly the form it has retained until the twentieth century: a large marsh with a nuclear lake of varying dimension (Lake Loudias in antiquity; mod. Yiannits•i) connectedto the sea by a river.56 We may thus envisagethe central Macedonian homeland as the fertile terracesabove the swampy Emathian plain. Farmers tilled the slopes,or drained patchesat the marsh'sedge?7probablyavoiding the central swamp whereverpossible.Pasturistsutilized the mountain meadowsabove, and game and timber was widely availablefrom the nearby slopes.It was, by Greek standards, a prosperous region, but awkwardly arranged in its central portions. As local populations increasedand the foreign demand for Macedonian metalsand timber grew, the needboth for cultivable land and for securityagainstGreek encroachmentsgave riseto an expansionof Macedonian

interests in all directions. •8

Institute of Classical Studies, London/ The PennsylvaniaState University

Eugene N. Borza

NOTES

1. A Histor.v of Macedonia I: Historical geography and prehistory (Oxford 1972) 160.

2. The literature on malaria is extensive, most of it, as one might expect. on

116

E.N.

BORZA

technicalmatters relating to treatment of the diseaseand to mosquito-eradication programs. Useful general accounts include: Gordon Harrison, Mosquitoes, malaria and man: A history of the hostilities since 1880 (London 1978), with a frightening final chapter showing that, because of political instability and the breakdownof publichealth services,malaria isagainbecomingendemicin partsof Asia and Africa: L.W. Hackett, Malaria in Europe. An ecologicalstudy (London 1937); and Paul F. Russell, Mang mastery of malaria (London 1955). On the originsand early spreadof the affliction seeL.J. Bruce-Chwatt,"Paleogenesis and paleo-epidemiologyof primate malaria," Bull. Worm Health Org. 32 (1965) 363-87; this is a comprehensiveaccountof what is known about the transmissionof malaria in antiquity written by the (then) Chief of Researchand Technical Intelligencefor the World Health Organization. For a valuabledescriptionof the diseaseseeBrian Maegraith in Adams and Maegraith, Clinical tropical diseases 6 (Oxford 1976), chap. 16. 3. Letter reprinted in full in Russell:seeesp. p. 60. 4. Ross' own account of these matters can be read in Sir Ronald Ross, Memoirs

(London 1923). 5. Bruce-Chwatt

365: and see next note.

6. The taxonomic classification of anopheline mosquitoesis a continuing problem. As an indication of the increasinglysubtle distinctionsbrought to this study by taxonomists.the number of known separatespeciesof mosquitoeshas risenfrom about fourteen hundred in 1932 to nearly three thousandin 1973;seethe statisticscitedin Kenneth L. Knight and Alan Stone,A catalogof the mosquitoesof the worm (CollegePark, Maryland 1977)I. Part of the difficultyin determiningthe vectorsfor malaria in Macedonia restsin the changingclassificationof anopheline mosquitoes,and, in particular, the precisetype and location of the speciesA. maculipennisand its various subspecies,once thought to be vectorsin Greece;see Marston Bates, "Anophelines of the palearctic region" 420, 422 and 426; M.F. Boyd,"Epidemiologyof malaria. Factorsrelatedto the definitivehost",table 104; and L.W. Hackett, "Conspectusof malaria incidence in northern Europe, the Mediterraneanregionand the Near East"788, all in M.F. Boyd(ed.), Malariology (Philadelphia and London 1949), 2 vols. The most recentcomprehensiveclassification(Knight and Stone, Catalog; see indices)makesA. maculipennis'statusuncertain, but clearly identifiesAnopheles anopheles sacharovi and Anopheles cellia superpictus as the main vectors in Macedonia.SeeespeciallyHackett (pp. 795-97and 1422in Boyd,Malariology) for a descriptionof the ecologicalrelationship betweenthe two species. 7. Great Britain, Admiralty, Naval Staff, A handbook of Macedonia and surrounding territories (London 1920) 65: "Malaria is notoriously the disease which is the scourgeof theselands."As late as 1936malaria was(excludingdeaths ascribedto senility)the fourth leadingcauseof death in Greece,behind pneumonia, tuberculosisand intestinal diseases.The malaria death-rate (75-80 per 100,000 population) wastwenty timesthat of any other Europeancountry of the time. An observernoted that on any one day in easternMacedonia up to 5.6% of the village population was incapacitatedduring the malaria season.In the summerof 1936, 69% of the infants in thoseMacedonianvillagesunder observationwere infected.It

OBSERVATIONS

was estimated that about two million

ON MALARIA

of Greece's seven million

117

inhabitants

were

malarious.SeeGreat Britain, Admiralty Handbook, Naval Staff, GreeceI (London 1944) 170-75 and 270-80, and the studiescited by Hackett, in Boyd, Malariology 796. For a basicdemographicsurveyof the incidenceof malaria in modernGreece, seeM.C. Balfour, "Malaria Studiesin Greece.Measurementsof Malaria, 1930-33", American Journal of Tropical Medicine 15 (1935) 301-30. 8. Malaria wasalso endemicin the easternsectionsof Yugoslaviain the 1930s and spreadquicklyasthe resultof the government's policyof shiftingsoldiersfrom one provinceto another. See H.E. Sigerist, The Sociologyof Medicine, ed. M.I. Roemer (New York 1960) 97. 9. Malaria, a neglectedfactor in the history of Greeceand Rome (London 1907), and Malaria and Greek history (Manchester 1909). 10. E.g., Malaria and Greek history 101-108. ll.

Bruce-Chwatt

377.

12. Malaria and Greek history,passim. A usefulepitome of Jones'testimonia, conclusions and historicaltheoriescanbefoundunderhisnameas"The prevalence of malariain ancientGreece,"in Diseasesin Antiquity, ed. Don Brothwelland A.T. Sandison(Springfield, Ill. 1967) 170-76.

13. In modernGreek "malaria" is either e16dispyret6s(kX&8• rc•pzv6•)or elonosia 14. E.g., Bruce-Chwatt377; Harrison 1; Russell80-82; Sigerist301-302;Adam

Patrick,"Diseasein antiquity:AncientGreeceand Rome",in Diseases in antiquity 238-46; Douglas Guthrie, A history of medicine(London 1945) 57; Arturo Castiglioni,A histor), of medicine,trans. E.B. Krumbhaar (New York 1940) 163 and 170; and Henry E. Sigerist,History of Medicine II (New York 1961)328-30. 15. J.L. Angel, "Porotic hyperostosis,anemias, malarias and marshesin the prehistoric eastern Mediterranean", Science 153 (1966) 760-63 [henceforth Science]; "Porotic hyperostosisand osteoporosissymmetrica", in Diseasesin antiquity 378-89; and "Ecologyand population in the easternMediterranean", Worm Archaeology4 (1972) 88-105[henceforthWorm Archaeology].It mustbe emphasizedherethat agreementamonganthropologists and biologistson manyof thesetechnicalmedicalmattersis lacking. Malaria itselfis a complexdisease;its mechanisms and pathologicaleffectsare not completelyunderstood.In thefollowing pagesI cannotclaimto haveavoidedcontroversy, asany seriousreviewof the medicalliteraturewill confirm. I hope,however,to haveprovidedfor my fellow ancienthistoriansa fair summary of the main trends,and to have shownthat there is little to suggestthat malaria was absentfrom Greecein classicaltimes. 16. For a detailedreviewof the processes of acquiredand innateimmunitiesto malaria, see Carol Laderman, "Malaria and progress:some historical and eco-

logicalconsiderations", SocialScienceandMedicine9 (1975)587-94,withcomprehensivebibliography. 17. Ibid.

589 and 592.

18. Ibid. 488-89, with bibliographyfor the mechanicsof the process. 19. Summarizedin A.C. Andrews,"The beanand Indo-Europeantotemism", Amer. Anthropologist 15 (1949) 274-92. 20. Walter Burkert, Lore and sciencein ancient Pythagoreanism,trans. E.L.

118

E.N.

BORZA

Minar, Jr. (Cambridge, Mass. 1972),passim,esp. 183-5.The fava-Pythagoreanlink is, however, no aid in attempting to date the introduction of malaria into classical Greece. Pythagoras was a sixth-century figure• but the earliest firm date for the writings of the Pythagorean schoolin which the bean is mentioned is the fourth century. See Burkert, 97-120, and Holgor Thesleff, An introduction to the Pythagorean writings of the Hellenistic period, Acta Academiae Aboensis, Humaniora XXIV.3 (Abo 1961) 3045. 21. Paul B. Burke, Jr., of Clark University, has recentlyargued in his detailed study of favism and the Pythagorean taboo that long-term exposureto malaria led the peoplesof antiquity to developa substantiallore about the bean. I am grateful to Dr. Burke for permitting me to seeand usea draft versionof hispaper,"Malaria in the ancient world: Prolegomenonto an ecologicaland environmental study", presentedat the 1978 meetingsof the American Philological Associationand now being prepared for publication. 22. Worm Archaeology 98. 23. Ibid. 94-95 and 100 C-.. malaria may have actually disappeared."):also Diseasesin antiquity 384, and Science 760-63. 24. Angel wasaware of someof the problemsinherent in the attempt to connect hyperostosiswith thalassemiaand thus malaria: seeDiseasesin Antiquity 381-84. 25. The statistical bases for Angel's conclusions raise some questions about method. The table in Science,which servesas the foundation for the hyperostosismalaria link, givesthe following data for Greece(my summary): Site Nea Nikomedeia

Date 6000 •c

No. skeletal remains 45

Kephala, Kea

3000 •c

37

Corinth

2400 •c

21

Lerna

1800 •c

149

ClassicGreece[sic]

450 •c

115

The frequenciesof hyperostosisvary among theseremains:what can be concluded from such variations? The data are drawn from one site in Macedonia,

two in the

Peloponnesus,one on Kea island, and one general category. They represent a chronology ranging from the early Neolithic to the classicalperiods, but at widelyseparatedsites.Further, there is no topographicalevidenceto provide a common frame of referencefor anophelineecologybeyond a proximity to marshes.With respectto the aforementionedwe have already seenthat at leastone malaria vector (A. superpictus)is a stream,not a marsh,breeder.An examinationof the statistical tables in Angel's 1967 (Diseasesin Antiquity) and 1972 (Worm Archaeology) articles reveals that the data are based on the same few sites (plus some new excavation material), and arranged virtually in the same manner. The method

seemsquestionableon the groundsthat a constantgeographicalor chronological referenceis lacking. If, for example, one were to take a singlesite (or region) and examineskeletalremainsrepresentinga long periodof time--say, severalcenturies --it might be possibleto producea history of the health of the inhabitantsof that locale. Moreover, if severalsucharea-analysesexisted,we would be in a position to

OBSERVATIONS

ON

MALARIA

119

speculateabout a large region such as Greece. But to utilize materials from a handful of sitesscatteredthroughout the Aegean world over a 5500-yearperiod--a samplelacking any constantfactor--does not, on methodologicalgroundsalone, seem convincing. Further, the statisticalsampleis small, and percentagefigurescan be misleading. In one case(Angel's report on Neolithic human remains in "Excavations in the Franchthi Cave, 1969-71. Part II", Hesperia42 (1973) 277q82)it is statedthat 31% of the Neolithic skeletonsat the Franchthi cave showedevidenceof hyperostosis, and were probably malarious.This is an impressivefigure until one realizesthat the evidence is drawn from the remains of four personsout of a total of thirteen discovered.Similarly, we may note conclusionsderivedfrom four personsat early Bronze Age Corinth, and two skeletonsfrom the late BronzeAge levelat Episkopi, Cyprus.The percentagesdrawn from the observablehyperostosisin thesehuman remains serve to plot curves upon which Angel bases his theories about the incidence

of malaria

in the entire eastern

Mediterranean.

One of the oddestexamplesof this method occurswhen Angel describesadult porotic hyperostosisat the 4% level("trace to slightdegree", World Archaeology 101) in the modern period, eventhough there is abundant first-hand evidencethat much of the Balkans (especiallyGreece) was malarious in the extreme in recent times (see note 7 above). And it will be recalled that this same 4% level for hyperostosiswas given for the Greek archaic period when malaria, accordingto Angel, was declining to virtual non-existencein the classicalperiod (see note 23 above).

26. World Archaeology 97. 27. G.L. Fraser et al., "Thalassemias,abnormal hemoglobins and Glucose-6Phosphate Dehydrogenasedeficiencyin the Arta area of Greece:Diagnostic and geneticaspectsof completevillage studies",Annals of the New York Academyof Science! 19(1964)415-35.The studiesacceptedby Angel,Diseasesin antiquity385, offer a contrary view. 28. See Laderman 589-90, who pointsout that to explain porotic hyperostosisin the New World, Angel posits causes other than malaria: iron deficiency and prolonged lactation, restrictedchildhood diet, dysenteryand responseto other parasitic infections. Laderman comments:"With so many choicesbefore us, why should we attribute porotic hyperostosisin two groups of early farmers to thalassemiaalone. 9With a changefrom huntingandgatheringto sedentaryfarming [preciselythe stageof developmentof theNea Nikomedeiafarmers]thereare many opportunitiesfor anemia to develop,includingparasitismand infectionof all sorts, as well as dietary deficiencies." 29. World Archaeology89 and 100. George H. Denton and StephenC. Porter, "Neoglaciation," Scientific American 222.6 (June 1970) 100-110. 30. See Denton and Porter's diagram (p. 107) of estimatedaveragefluctuations of mountain glaciers. 31. World Archaeology 89 ff. We have already noted that the presenceor absenceof marshland per se is not a contributing factor to the incidence of thalassemia,and also that boneenlargementcan resultfrom other than thalassemic conditions.

120

E.N.

BORZA

32. Roughlyspeaking,thereis a constantamount of free water on thisplanet.It existsin the atmosphere,in seas,lakes,rivers,etc., and in the form of iceand snow. The balancebetweenthe amount of water in ice sheetsand glaciersand in the sea will shift dependingupon the temperature of the general climate. I accept the view that the change in the level of the Mediterranean in historical times has been no more than 30-50 cm; see W. Gordon East, "The destruction of cities in the Mediterranean lands," The sixth J.L. Myres memorial lecture (Oxford 1971) 4-5, following the archaeological-geological surveysof N.C. Fietaming, for which see below.

33. As far as Angel'sargumentson the efficacyof man-mademarshdrainageare concerned,the only evidence we possessis Theophrastus' statement (de Caus. Plant. 5.14.6) that Philip II reclaimedthe water-loggedplain of Philippi. There isno evidencefor drainage in Emathia, and the Lake Kopais regionof Boeotia(heavily malarious in modern times) remained marshy following the breakdown of the Mycenaean drainage system.Pollen analysisin the Kopais region showsthat the forestshad been decimatedsinceat least Bronze Age times; seeJ.R.A. Grieg and J. Turner, "Some pollen diagramsfrom Greeceand their archaeologicalsignificance", Journ. Archaeol. Science I (1974) 177-94. Malaria may be the "fever" mentioned as having struck the region not long after the battle of Chaeroneia (338 BC);seeTheophr. Hist. Plant. 4.11.3. 34. Laderman 587-88 and 592-93• and L.W. Hackett, "Distribution of malaria",

in Boyd, Malariology 722-35, esp. p. 730. 35. "The battle of Salamis,"JHS 76 (1956) 32-54, esp.35-36, and Epirus(Oxford 1967), seeindex under "Sea level, changesof." 36. "The campaign and battle of Marathon", JHS 88 (1968) 13-57. Revised versionsof both the Salamisand Marathon papersappear in Hammond's Studies in Greek history (Oxford 1973), but nothing there affectsthe presentdiscussion. 37. N.C. Flemming, N.M.G. Czartoryska and P.M. Hunter, "Archaeological evidencefor eustatic and tectonic components of relative sea level changesin the South Aegean",Marine Archaeology,ed. D.J. Blackman,Colston Papers,Vol. 23, Proceedingsof the 23rd Symposiumof the ColstonResearchSociety(Bristol 1971) 1-63. Also seeFlemming, Citiesin the sea(New York 1971),App. I I, and "Changes of land and sealevel in the Aegeanarea sincethe BronzeAge"• BICS21 (1974) 15557. The view that the seareachedcloseto its presentlevel about 5000 yearsago is found in H.E. Wright, Jr., "Glacial fluctuations, sea-levelchangesand catastrophic floods," in Atlantis:fact or.fiction?ed, Edwin S. Ramage (Bloomington, Ind. and London 1978) 161-74.

The matter, however, is being debated. Opposed to Flemming's view, and arguing that the Mediterranean sea level has been rising at the rate of about one meter per millennium sinceabout 6000 BC,is John Bintliff, Natural environment and human settlementin prehistoricGreece,BritishArcheologicalReports,Suppl. Ser. 28i (Oxford 1977) 13-26,althoughin hispreface(p. 1) he suggests that he feels lessconfident about his argument than when he first wrote it. Bintliff's thesisis summarizedin his "New approachesto human geography.PrehistoricGreece:A casestudy",in F.W. Carter (ed.), An historicalgeographyof the Balkans(London, New York and San Francisco 1977), chap. 3.

OBSERVATIONS

ON MALARIA

121

38. E.g., the block of the Peloponnesusis now tilted relative to its positionin antiquity, and apparentlysoutheastern Englandis sinking,with London some12 feet lower today than it was in Roman times. In Greecethe effect of land subsidence causedby seismicdisturbanceswas clearly recognizedby the excavators of the (now) underwater harbor site at Kenchreai on the east side of the Corinthian isthmus.See Robert Scranton and JosephShaw, "Changesin relativesealevel", App. E in Robert Scranton,JosephW. Shawand Leila lbrahim, Kenchreai,eastern port of Corinth I. Topography and architecture (Leiden 1978). The authors have compileda usefullist (with sources)of earthquakesattestedin the regionand have added an outline of the effect of these seismic disturbances, which have resulted in a

subsidence of more than two meters since Roman times. In some cases, however,

earth movementshave causeda relativerisein land forms, as in the example of the isthmusof the Athos peninsula,which is nearly 14 metershighertoday than when Xerxes cut a canal through it. An important methodologicalpoint emergesfrom these studies:the evidence of a few scattered submarine remains from antiquity cannot, without corroboratinggeologicaldata, be usedalone to prove that there has been a general rise in the sea level of the Mediterranean. 39. For example, a comparisonof Denton and Porter's data (op. cit., n. 29) showingestimatedaveragefluctuations of mountain glaciers(an indicator of shifts between warmer/wetter and cooler/drier climate patterns) with pollen records from the plain of Philippi showsno correlationbetweenclimatechangeand the two brief periods(ca. 1900-1300BCand ca. 1050-500BC)duringwhicha few olivesgrew in the region.See Denton and Porter 107,and GriegandTurner (cit. n. 33),passim. For a review of additional studiessupportingthe notion that the climate of Greece today is much as it was in Classicalantiquity, seeBintliff, Natural environment 51. 40. Although the famous honey-beeshave fled Attica's Mt. Hymettos for relief from the quarrying operations,the small purple Hymettan flowers that occasionally suffuseAthens' atmospherewith a reflected pale violet light in the late afternoon sun still exist. For Hesiod'scommentson the seasonsseeErga 383ff., 415 if., 448ff., 479ff., and 564ff. While we cannot pinpoint Hesiod'shome-town of Ascra, the region of Mt. Helikon whereit waslocatedis still "bad in winter, difficult in summer, good at no time" (Erga 639-40). On the abundance of pine, oak and beech--major Macedonian woods today--in antiquity, see, e.g., Theophrastus, Hist. Plant. 3.8.713.9.2,6: 3.10.2• 4.5.515.2.1. For a useful surveyof Macedonian timber resourcesseeE.C. Semple, Thegeographyof the Mediterranean region. Its relation to ancient histor.v (London 1932) 276-77. The olive tree, which characterizes the "Mediterranean" climate par excellence, is not a regular feature of the Macedonian scene.Mention of its existencein Macedonia is lacking in our ancient sources(except for a fragment of Theopompus in Athen. 3.77E, which is too complexto deal with herebeyondpointingout that it is mainly nonsense, speaking as it doesof double-bearingfruit treesand fig trees,vinesand olivetreesproducing fruit in the middle of Spring). Except for the southernslopeand the peninsulasof the Chalcidice, the olive is rarely seenin Macedonia, the climate being too harsh and unmitigated by Mediterranean sea-breezes.The northern limit of olive trees 1 have observedlies along the northwesternChalcidiccoastabout 20 milessouthof Salonica. (This, of course, excludes the single tree which the Director of the

122

E.N.

BORZA

ArchaeologicalMuseum in Salonicaclaims to have planted and nurtured in the Museum courtyard!) For a general survey of plant life in Macedonia see W.B. Terrill, The Plant-life of the Balkanpeninsula.A phytogeographicalstudy(Oxford 1929).

As for fauna, many of the wild and domesticatedanimalsof antiquity--goats, sheep,cattle, boar, pigs--are commonenoughtodayin the mountainsand plainsof the region. Finally, both Xerxesand Odysseuswould appreciatethe warningson wind forceat Athosand Malea mentionedby any Mediterraneanpilot'sguide,and confirmedby the Greek MeteorologicalService;someof thisinformationappears in a table which accompanied my "Alexander's Communications", Archaia Makedonia II (Thessaloniki1977);seep. 303 for Kythira, lying oppositeMalea. Where thesematterscan be checked,it appearsthat the generalclimateof Greecein moderntimes is similar to what it was in classicalantiquity. 41. See Bruce-Chwatt

365-77.

42. Evidence of malaria may exist in some first-millennium BCmummies, and

splenomegalyis mentionedin Egyptianmedicaltexts(ibid.), althoughthe symptoms may reflect other diseasesas well. It would appear that malaria was more prevalentin upperEgypt than in the regionof the Nile Delta; seethe studiescitedby Bruce-Chwatt

373 and 376. One also recalls that Herodotus

commented

on the

excellenthealth of the Egyptians(2.77), evenwhile describing(2.95) the annoying biting flying insects(xta,•torr½•)--probably mosquitoes. Much of the above is confirmedby modern experience.It is curiousthat lower Egypt--in particular the vast Delta, which on ecologicalgrounds alone would appear likely to be malarious--is only mildly infected,even thoughthe countryis surroundedby regionsof intensemalaria in the easternMediterraneanand tropical Africa. None of the deadly anophelinevectorsof East Africa, the Near East or the northern

Mediterranean

littoral

inhabit

the Nile

Delta.

The

short

life of the

ubiquitousDelta mosquito,A. pharoensis,living under conditionsof extremeheat and low humidity, plus the inability of the fearsomeA. gambiae to penetrateinto Egypt from the southbecauseof the geographicalconfigurationof the upperNile valley, has made transmissionof the diseaseinto the region difficult. Only in the oasesof the Western Desert, where somehardy anophelineshave adaptedto the harsh existence,and in the pools of the upper Nile is malaria a serioushealth hazard; seeHackett, in Boyd, Malariology 794-95, and George Macdonald, The epidemiologyand control of malaria (London 1957) 51 and 75-77. 43. Evidencefor antiquity cited in Jones,Malaria and Greekhistory23-59(nonmedical writers) and 61-73 (medical writers). Also, see a selectionof passages describing malaria drawn from the Hippocratic corpus, Aristophanes, Varro, Celsus,Pliny the Elder and Martial• in R.H. Major• Classicdescriptionsof disease 2 (Springfield, Ill. and Baltimore 1939) 105-111. 44. Henry E. Sigerist, History of medicineII (cit. n. 14) 66; Bruce-Chwatt 377, whosecitation of Sigerist is incorrect. 45. See Bruce-Chwatt376, for the incidenceof malaria in Mesopotamiaasearly as ca. 2000 Bc. Sigerist,The Soc. ofMed. (cit. n. 8) 338, suggested that Greecewas savedbecauseof the outbreakof an epidemicof malaria that decimatedthe Persian armies. Sigerist, however, offered no evidence for this• and the reader is left

OBSERVATIONS

ON

MALARIA

123

wonderingwhetherthe Persianssufferedfrom an affliction brought with them or from one contracted in Greece.

46. Laderman(p. 592) is amongotherswho havesuggested a Persiansourcefor the disease.For anotherexampleof an infectedarmy spreadingmalaria, seenote8 above.

An extraordinary example of malaria being transmitted into a region by an infected person occurred in 1952-53, when a single outbreak of 35 caseswas reportedin California. All the victimswere females(mainly teenagers)who had spent 12daysat a Camp Fire Council summercampat Lake Vera, California, in the foothills of the Sierra Nevada near the Nevada border. Anopheline mosquitoes inhabit the area, but malaria was virtually unknown in the region. Intensive investigationof the outbreakrevealedthat the sourceof the infectionwasprobably a singleAmericanserviceman,recentlydischarged afterduty in Korea,whohimself had suffereda malaria attack,and who wasknownto havespenta holidayweekend at Lake Vera. The presenceof the infectedman in the area in early July, and the appearanceof symptomsamongthe girls not long afterwards,coincidewith the acceptabletiming for infection and transmissionby the mosquito vector. See Rosemary Brunetti, R.F. Fritz and A.C. Hollister, Jr., "An Outbreak of Malaria in California, 1952-53", Amer. Journ. of Trop. Med. and Hygiene 3 (1954) 779-88. The developmentof internationaltravelon a massscaleat a timewhenmalariais increasingin the tropical and subtropical regions(see note 2 above) raisesthe spectrethat the diseasemay again becomea seriousworld health problem. Indeed, during the final days of revisionof the presentpaper, the travel sectionof the Sunday edition of the New York Times (May 11, 1980) carried a bleak and ominously-entitled article, "Travelers Are Warned of Increasing Danger of Malaria."

47.

See note 34 above.

48. In Greece the Lake Kopais region was also highly malarious in modern times, yet continued to support a large, though diseased,population. For the possibilitythat the area was malarious in antiquity, see note 33 above. 49. I have not attemptedyet to read through the accountsof death and disease among the ancient Macedoniansas preservedin authors from antiquity, but attention should be called to the article of Donald Engels,"A note on Alexander's death", CP 73 (1978) 224-28, who argues that Alexander died of malaria first contractedin Cilicia. Perhaps the king was reinfectedat Babylon, the original sourceof the diseasehaving been Macedonia itself. One might also wonder whether, when in the courseof a campaignagainstthe Olynthiansin 380 Bcthe Spartan king Agesipoliscamedown with feverand died at Aphytis on the Cassandreanpeninsula,the "fever" might not have beenmalaria. SeeXen. Hell. 5.3.10; Diod. 15.22.2,23.2; Paus.3.5.9; Tod, GHI2.120. Mythanks to ProfessorFordyce Mitchel for calling this incidentto my attention. 50. History of Macedonia I (cited n. 1 above) 149 and 160. 51. Admiralty Handbook, Greece111(cited note 7 above ) 140 and plate 57. 52. The most comprehensive recent studies are by Hammond, History of Macedonia I 145-47,C.F. Edson,"Strepsa(Thucydides 1.61.4)", CP50 (1955) 176, and J. Bintliff, "The plain of Macedon and the neolithic site of Nea Nikomedia",

124

E.N.

BORZA

PPS 42 (1976) 241-62,and Hist. Geog.Balkans(citedn. 37 above),chap.3, whose reconstruction of the geologicalhistoryof the EmathianplainI follow. Apparently the observablechangesin the plain's topography are the result of a fairly wellunderstoodalluvial process,not of any tectonic or volcanic shift. 53. Seethe distributionof prehistoricsitesin D.H. French,Index of prehistoric sitesin centralMacedonia(Athens 1967),now reproducedasan end paperin R.A. Crosslandand Ann Birchall, BronzeAge migrationsin the Aegean. Archaeological and linguisticproblems in Greek prehistory (London 1973). 54. Modern maps and travellers' accounts agree that medieval and modern settlementalso was confined to the nearby higher ground. ]'he establishmentof

villagesin the plain itself had to await the twentieth-centuryprogramof reclamation.

55. This is, roughly speaking,the route of the new National Road, from which can be seen, near the village of Kleidi, the sole remnant of the Roman road, an impressivebridge arch. For background,seeHammond, History of Macedonia 1 162, and Edson 180.

56. In Colonel Leake'stime (1806) the lake of Yiannitsfi was deep and fresh enoughto provide large pike to the residentsof Naoussa.See William Martin Leake, Travels in northern Greece III (London 1835) 287.

57. We must assumethat Macedonianfarmers,like thoseof Egypt and Mesopotamiain an earlierage,werecapableof irrigatingand drainingland on a small scale.But the evidenceis lacking for any large-scalereclamation in the Emathian plain, which was not receptiveto flood-controland irrigation worksas usedin the Nile and Tigris-Euphrates valleys. Even at Babylon, flooding continued to be a problem, and we hardly need a reminder that it was in those very Babylonian marshes that Alexander's final ailment struck; see note 49 above.

58. I am indebtedto Don Brothwell, Department of Environmental Studiesof the University of London, to Alexander C. Tokarewicz, M.D., and to the library staff of the London School of Hygiene and Tropical Medicine for their assistance with someof the technicalmedicalaspectsof this paper.The Editor of and readers for this Journal have offered additional valuable criticism, which I acknowledge with gratitude. None of the above is to be held responsiblefor my views.

THE

THREE

DAUGHTERS

OF AGRIPPINA

MAIOR

The elder Agrippina, wife of GermanicusCaesarand eulogizedby Tacitus as insignifecunditate,praeclarapudicitia (Ann. 1.41.3),wasthe mother of ninechildren,six of whom survivedinfancyto play varioussignificantroles in the evolution of monarchyat Rome: her three sonsNero, Drusus, and Gaius;and her daughtersJulia Agrippina, Julia Drusilla, and Julia Livilla (Suet. Cal. 7.1). Of the birthdaysof thesenine children,only that of Gaius Caligula is firmly fixed by our sources,to 31 AugustAD 12.1For the restwe must draw what inferenceswe can from passingand often ambiguous referencesin the historiansand from fragmentaryepigraphicevidence.The resultshave not been entirely satisfactory,especiallyfor the births of the three girls, the youngestof Agrippina's children.2

Tacitus records that Julia Livilla• the last child, was delivered on Lesbosin

18, not long after her father had assumedthe consulshipfor that year on 1 January,at Nicopolis.3Julia Agrippina, generallyacceptedas the eldestof the three, had been born on 6 November at Oppidum Ubiorum on the Rhine.4 The year of her birth is nowhere stated, but both Tacitus (Ann. 1.40.2; 1.44.2) and Dio (57.5.6-7) observe that Germanicus' wife was pregnant during the legionary mutiny that followed Augustus'death on 19 August, 14. Julia Drusilla, apparently the middle daughter,was also born on the northern frontier and accordingto Tacitus accompaniedher father in histriumph at Rome on 26 May, 17.5Logically,then, we shouldplacethe girls' births within the period of somethirty-nine monthsbetweenNovember of 14--the first year of their mother'ssojournon the Rhine--and late January or February of 18. Againstthis theory we havethe testimonyof Suetonius--whoseadmirable investigationinto the birthplace of Gaius Caligula is taken asproof of his credibility in all suchmatters--that the threesisterswereborn continuo triennio (Cal. 7). Now, a discrepancyof only three monthsfrom a temporal phrasethat does not suggestan exact computation would not, when taken alone, causeus to questionthe apparentevidence,but thereare two further piecesof informationthat makethe birth of Julia Agrippinain November of 14 most improbable. First, a letter from Augustusto the elder Agrippina, written in 14 and preservedby Suetonius(Cal. 8.4)• apparently indicatesthat shecould not havejoined Germanicuson the Rhine early enoughin 14to haveconceived

125

126

J. HUMPHREY

a child to be born in November of that year. The letter has usually been taken as proof that Agrippina and young Gaius did not leaveRome until May of that year• but it doesnot in fact allow usto be sospecific.6Sincewe can assumethat Augustuswas in Rome when he wrote the letter, it follows that Agrippina very likely was not there, but was probably already on her way north to Germanicus,havingleft her youngestsonwith Augustus.Yet, unlesswe are to imagine Agrippina making the difficult journey in the winter months of early 14, she still could not have been reunited with Germanicusbefore the middle of spring, somethree monthstoo late for a child to be born in early November. Then thereis the famousscenedescribedby Tacitus in the winter camp of the mutinous legionsat Oppidum Ubiorum, a scenethat must be dated to the middle of October, 14, at the very earliest, shortly after the arrival in camp of the senatorialembassythat had been sent out from Rome in the third week of September.7 To protect his son and pregnantwife from the violence of his own soldiers,Germanicusplanned to sendthem up river to the undisturbedterritory of the Treveri. Though he soon yielded to the

pleas of the repentantmutineersand allowed young Gaius to remain in camp, reditum Agrippinae excusavit ob imminenteraparturn et hiemere (Ann. 1.44.2; cf. 1.41.2;Dio 57.5.7). After the rebellionhere at Oppidum Ubiorum was suppressed,Germanicus made his way downstream to Castra Vetera, where the fifth and twenty-first legionswere stationed,and led them on a short autumn campaign acrossthe Rhine before returning to winter quarters (Ann. 1.49.5-51.9). It is almost impossibleto imagine Agrippina returning to Oppidum Ubiorum during her husband'sabsence, or indeed before the birth of her child and within three weeks (at the most) of her original departure:after all, parts of the mutinousfirst and twentieth legionswere still in camp there.8 So the youngerAgrippina could not have beenborn in Oppidum Ubiorum in the first week of November, 14. This much seemsreasonablycertain.

Over a century ago Mommsen devised for the births of Germanicus' children a chronology that has been accepted almost universally ever since.9 Acknowledgingthe improbability of Julia Agrippina'sbirth in 14, he basedhissubsequentreasoningon two assumptions: first, that the order of the girls' names in Suetoniusand on coinage,and the arrangement of Julia Agrippina'smarriagefive yearsbefore the betrothal of her sisters, prove beyondany doubt that shewasthe eldestof the three;l0and second, that Suetonius'phrasecontinuotriennioindicatesthat the birthsmusthave occurredin three successive calendaryears.• Mommsen'sconclusions,and the principal argumentsin favour of them, were as follows:

THREE

DAUGHTERS

OF AGRIPPINA

127

1. Julia Agrippina was born on 6 November, 15. A date in the previous year must be ruled out on the basis of the evidence alreadymentioned,while November of 16 is impossiblylate since a woman cannot produce three children by three separatebirths within a spaceof fifteen months. It naturally must follow that the elder Agrippina'spregnancyin the autumn of 14 could only have produced a stillborn child.12 2. Julia Livilla, the youngest,must then have beenborn in the last months of 17. Since it is unlikely that Germanicuswould have exposedhis pregnantwife to the rigoursof a winter journey from Rome to the East, he must have sentAgrippina ahead to Lesbos beforethe end of the sailingseasonof 17.Tacitus,who postponed mention

of the birth

until

Germanicus'

own arrival

on Lesbos

early the next year, could have clarified the order of events by writing ediderat rather than edidit in Annals 2.54.1.13 3. Julia Drusilla

was born at Vicus Ambitarvius

near the end of 16.

The testimonyof Dio, who recordsapparentlyin the first half of the year 39 a posthumouscelebration of Drusilla's genesia or birthday (59.13.8),14must be rejected:the seconddaughtercould not have been born before August of 16 (that is, at least nine months after Julia Agrippina), and Drusilla's proven birth in Germany and subsequentattendanceat Germanicus'triumph in Rome make the early part of 17 very improbable. Thus, as Suetonius' words implied for Mommsen, Germanicus' three daughterswere born in three consecutivecalendaryears:Agrippina in 15, Drusilla in 16, and Livilla in 17. Yet there are difficulties in Mommsen's

arguments,difficultiesthat arisefrom his interpretationof the ambiguous phrasecontinuotriennioand from his belief in Julia Agrippina'sposition as the eldest daughter. By acceptingthese points we are compelled to questionTacitus'Latin, to rejecta clear statementin Dio, and to postulate a hitherto undisclosedbirth of a dead child that might well have been viewedasa tragicresultof the military revolt.There mustbea betterway to reconcilethe imperfect evidenceof our sources. III

Consider

first

Suetonius'

use of the word

triennium

and its relatives

biennium, quadriennium, and quinquennium, which occur twenty-seven times elsewherein the vita Caesarurn.15For sevenof these the specific periodcannotbe determined,or the chronologyis still a matter of dispute that goesbeyond Suetonian usage.16From the remaining instancesit is apparent that the words indicate an absolutelyexact period of time only when they refer to a formal legal proposalor to the duration of a public

128

J. HUMPHREY

office, or when they are expanded by a specifiednumber of monthsand days; and in at least ten of thesecasesthe time implied is not a number of calendaryears,but rather an uninterruptedperiod beginningin any of the twelve monthsand lastingvery closeto a certain number of years.17More to the point are the four passagesthat match the idiom of continuo triennio:Julius Caesarpasseda law forbiddingcertaincitizensto remain outside Italy plus triennio continuo (Jul. 42.1); Tiberius did not leave Rome biennio continuopost adepturnimperium ( Tib. 38), and after his retirementto Capreae in 26 saw his mother Livia only oncetoto quidera triennio before her death in 29 (Tib. 51.2); and Vitellius wasfirst governor of Africa and then legatusof his successorbiennio continuato ( Vit. 5). In each case the accompanyingadjective does not imply a successionof calendaryears--the conceptof succession is after all implicit in the nouns themselves--buta period of time, amountingto a certainnumberof years, in which there were no real interruptions.•8 The point, then, of Suetonius'commentthat the three daughtersof the elder Agrippina were born continuotriennio is not the coincidencethat they were delivered in three consecutivecalendar years, but that their mother, a dutiful and productiveRoman wife, within a period of about thirty-six months bore three girls in succession,with no male offspring intervening.There is no need, as Mommsen felt there was, to qualify Tacitus' reference to the birth of Julia Livilla; it can confidently be left where the historian put it, in late January or February of 18. We pass now to the seasonof Drusilla's birth. The position of Dio's passagereferred to above (59.13.8) suggeststhat in 39 the posthumous celebrationof her birthday was held in the first half of the year. In Dio's accountit precededthe reintroductionof trials for maiestas(an eventthat cannot yet be dated), the affair of the pontoon bridgeat Baiae(probably, but not necessarily,in the summer), and Gaius' hurried departure for Mogontiacum in the north, firmly fixed to the end of the first week in September.19This evidencewasdismissedby Mommsenwith thecomment that "die Ereignissesind bei Dio schwerlichin strengerZeitfolge aufgeffihrt".20This is true enough:Dio often mingledspecificincidentswith undated anecdotesand broad generalizationsthat encompassedmonths and sometimesyears. But in this case he is quite clear about the proper order of events.

At 59.13.1 Dio recordsGaius' secondconsulship,which he heldfrom 1 January, 39, for only thirty days.2l There follows a vaguedescriptionof executionsand of the deterioratingrelationshipbetweenthe Emperorand the people,all introducedby the inexactphrase"x•[ •, zz zo• •rez,.zo•" (59.13.2)andendingwith thecommentthat "xod

THREE

DAUGHTERS

OF AGRIPPINA

129

(59.13.6). Finished with his generalitiesand anecdotes,Dio now indicates carefullythat he is resuminghischronologicalaccountat the very point he left it, in Januaryof 39:"He alwaysactedthisway(-r•t6-r•t y.•,•)... and once even (x•tt rrozz) ... but at the time I was speakingof (zdzz 8•) .... -2 Angeredat this time bya publicdisturbanceat showsand games,Gaiusleft

for Campania,returning"later" (y.zz& zo6zo)to celebratetheanniversary of Drusilla's birth with two days of varied displays.Dio is quite clear that Gaius was presentfor the celebrations,which would of coursehave been impossiblehad Drusilla's birthday fallen after the beginningof September, as Mommsen would have it: the Emperor wasthen on hisway to Germany. Dio's narrativemayallow usto be still morespecificabout the date of the birthday celebrationsin 39. The referenceto "showsand games"at which the precedingpublic disturbanceoccurredcould be appliedto a numberof festivals,beginningwith the first ludi of the year, the Palatine Gamesthat beganon 17 January.23But whenwe read in Dio that, beforehisdeparture for Campania, Gaius transferred control of the games to others, it is reasonableto assumethat he wasoneof the seniormagistrateswho had the privilegeof droppinga mappaas the startingsignalfor the chariots.TM If this is so, then the ludi in questionmust be the Palatine Gamesheld during his brief consulship.Further, the Emperor'strip to Campania must havebeen a shortone, sincehe wasback in Rome on 30 Januarywhenhe resignedhis

magistracy(Dio 59.13.1), and Dio's wording (&r•x,zX0• rr?• -r&-r7• Apou,tXX• ¾zx,•,[•) suggests that his sister'sbirthdayfollowednot long after. That it did occur at any rate before the autumn is confirmed by the absenceof any record of a celebrationat the end of the previousyear, 38, although an annual commemoration had been decreed that very year, sometime between Drusilla's death on 10 June and her consecration,

perhapsas late as 23 September.25 By limiting the date of Drusilla's birth to a time betweenthe very end of January and late August, we can define as well the year in which it must have occurred. We can eliminate 17 from the beginning:shecannot have been born after 26 May of that year, when she was seen in her father's triumph, nor indeed between February and May, since her conception would then have occurredat a time during the previoussummerwhen we know that Germanicuswas on a prolongedcampaign.26The early months of 16would be possibleonly if Julia Agrippina'sbirth were movedback to 14, while a time in August of 16 is just credibleif conceptionoccurred within a month of the deliveryof Julia Agrippina the previousNovember: an unlikely but not quite impossiblecircumstance? Yet even if this were the solution, or if a premature birth were assumedin Drusilla's case,three difficultieswould still remain unresolved:Suetonius'"space of threeyears" mustbe appliedwith abnormal looseness to a periodof onlyabout twenty-

130

J. HUMPHREY

sevenmonths;Dio's accountof 39 mustbe reorganized;and we muststill createa tragic conclusionfor the elder Agrippina'spregnancyin the autumnof 14, about whichTacitusis silentand Suetonius completely ignorant.28We are left, then, with a date for Drusilla'sbirth sometimein the

sevenmonthsbetweenFebruaryand Augustof 15,and preferablyearlyin that period rather than later. Sucha date obviouslydemandsa reinterpretationof the traditionalorderof the girls'ages. IV

It hasalwaysbeenassumedthat Julia Agrippina wasthe eldestof thethree daughters.Though her seniorityis nowherestatedexplicitly,her nameis listedfirst by Suetonius,who in the samechapterrecordsGermanicus'

surviving sonsin orderof theirages(Cal. 7)•thesamedisposition of names and portraits occurson coinagefrom the reign of Gaius? and Julia Agrippinawasmarriedfirst, in 28, five yearsbeforeher two sisters. 30The evidenceis not overwhelming,and deservescloseexamination. It wasunquestionably Suetonius'habit to list at leastmale offspring from theeldestto theyoungest. 3•In thiscase,however,thereisnocertainty that the biographer knew which daughter was in fact born first. His investigationinto the birthplaceof the Emperor Gaius uncoveredsome supplementary informationabouthissisters: that theywerebornoneafter the other, two of them in the area of Gallia, and that the birth of one was

commemorated with an altar in vico,4mbitarvioamongtheTreveri(Cal. 8.1, 3). It is understandable that Suetoniusshowslittle interestin thegirls exceptwhenthey are involvedin men'saffairs:nowheredoeshe speculate on whichof them wasthe objectof the dedicationwhichPliny the Elder took to refer to Gaius, nor even on which two were born on the northern

frontier.Yet an answerto theseproblemswouldhavehelpedprovethe impossibility of Gaius'birth in castris,the wholepointof Suetonius' long excursus.His silencesurelyreflectshisignoranceof the detailsof the elder Agrippina'sGermanpuerperia? Secondly, it is always hazardous to base inferences on the artistic evidenceof coinage.The dispositionof numismaticportraits,whether labeledor not, is determinedby factorsotherthan seniority.It is truethat when two figuresare depicted,as on the reverseof a dupondiuscommemorating Gaius' brothers Nero and Drusus, the name of the elder appearson the left.33In the caseof threefigures,however,an artisticand allusive balance is effected by placing the significant or distinctive

individualin thecentre.So wefindtheheadsof GaiusandLuciusflanking that of theirmotherJulia? or, on a moreappositeexample,thefull figure of Britannicusbetweenthoseof hissisters,the youngerOctaviaon the left

THREE

DAUGHTERS

OF AGRIPPINA

131

and the older Antonia on the right.35In this instancethe namesof the outer

figuresare inscribedvertically and their headsare turned towardsthe centre,drawing attention to the prominentindividual there whosename appearshorizontallyabove.Exactlythe samepictorialarrangementwas used on Gaius' sestertiusof 37-38, in which Julia Drusilla is the dominant

central figure.36

Finally we have the evidencefrom Tacitus that Julia Agrippina was marriedfive yearsbeforeher sisters,and sowasapparentlytheeldest.37We must rememberthat girls were legally of marriageableage after their twelfth birthday;3• so even though wed in 28, Agrippina could have been born aslateas 16.Althoughthereisno evidencethat womenof the imperial family were married off in orderof their ages,it couldreasonablybe argued that, if Agrippina was not the eldest,we might expect some marital arrangement to have been made for Drusilla before 33. In fact, two betrothalsof Germanicus'daughtersseemto havebeensettledwell before Agrippina'smarriageto Domitius Ahenobarbus. When Tacitus recordsthe prematuredeath in 22 of AsiniusSaloninus, sonof AsiniusGallusand of Vipsania Agrippina, Tiberius'former wife, he callshim Caesariprogenerdestinatus(Ann. 3.75.1): the only explanation possible is that he was betrothed to a daughter of Germanicus, the Emperor's adopted son? Then, at the end of his account for 27, when Tacitus alludesto the prosecutionof Quintilius Varus, probably the sonof Augustus' unfortunate general, he describeshim as divitem et Caesari propinquum (Ann. 4.66.1). Now the known relationship of the younger Varus to Tiberius was neither direct nor close:his great-grandmotherwas probably Octavia minor, Augustus'sister;and Varus' father and Tiberius had years earlier both been sons-in-lawof Marcus Agrippa.40Either of thesedistant relationshipscould perhapsbe encompassedby propinquus, especiallywhen they relate to the imperial house? But when we find Seneca Rhetor describingthe young Varus as tunc Germanici gener ut praetextatus(Contr. 1.3.10),it is obviousthat Varus wasengagedto oneof Germanicus'daughters,though it is not clearif he evermarried her.42Since we hear nothing more about him or about any descendants,it seemsthat his intended alliance with the imperial family was abandoned, perhaps because of this prosecution, and that Varus himself, if not finally condemned,may have committedsuicide. 43 While we do not know to which of the three daughtersSaloninusand Varus were engaged, one of them could certainly have been Drusilla. Otherwisewe must assumethree successive engagementsfor Agrippina while her two sisters,close in age, were neglectedentirely; or that the youngest,Livilla, was not expectedto wait her proper turn--in which case the marital arrangements indicate nothing about seniority.44 Though

132

J. HUMPHREY

Tacitus was aware of at least one of these two earlier betrothals, he

neglectedto recordthem in their chronologicalplace,presumablybecause nothing significant resultedfrom them. Agrippina was indeedthe first to marry, but that cannot be taken by itself as proof that shewas the eldest daughter.

From this accumulatedevidence,then, it would appearthat Drusilla could havebeenGermanicus'eldestdaughter,born early in 15,and that the births of Agrippina and Livilla followed in November of 16 and early in 18.45It now remains to trace the movementsof their parents, to ensurethat such a chronology is possible. Germanicus,after holding the consulshipof 12 for the entire year (Dio 56.26.1), seemsto have lingeredin Rome for a time at the beginningof 13 (Dio 56.28.4 (Xiph.)) before setting out for his command in Germany (Suet. Cal. 8.3). His wife Agrippinafollowedhim the next year, probablyin the early springwhen the Alpine route could be travelled without serious difficulty; and his youngestson Gaius left for the Rhine later, on 18 May, with an escortsuppliedby Augustus(Suet. Cal. 8.4).46Sometimein May, not long after their reunion, the couple's seventhchild could have been conceived, so that by the time of the legionary revolt in October Agrippina's pregnancyhad advancedby more than five months.47Sent to the Treveri for her protection, sheremainedthere until after the birth of her daughter Drusilla the next year--as early as February--at Vicus Ambitarvius, where an altar was erectedto commemoratethe event (Suet. Cal. 8.1). A few weeks later Germanicusbegan an early springcampaign(Tac. Ann. 1.55.1), returning to the Rhine for a short time beforeembarking on the major summeroffensiveof 15 that would occupyhim until after the fall equinox (Tac. Ann. 1.56-57; 1.70.2). Shortly before his return Agrippina displayedher courageand resourcefulness during a frightenedattempt to destroythe bridge acrossthe Rhine, probablyat Vetera (Tac. Ann. 1.69;cf. 1.49.6: Str. 4.3.4). Femina ingensanimi, wrote Tacitus, munia ducisper eos dies induit. It could hardly have escapedthe notice of her admiring historian if at the time shehad also been sevenmonthspregnantwith her daughter Agrippina, as Mommsen'schronologywould have it. In fact their seconddaughter seemsnot to have been conceiveduntil about February of 16. Soonafter this Germanicussetout on an unusually early campaign and did not return until he led some of his troops into winter quartersat Oppidum Ubiorum (Tac. Ann. 2.5.4; 2.26.1; cf. 1.39.12). There, on 6 November of 16• Julia Agrippina was born.

THREE

DAUGHTERS

OF AGRIPPINA

133

It was there, too, that Germanicus received a series of letters from

Tiberius requestinghis return to Rome (Tac. Ann. 2.26.3-6). The correspondencebetweenEmperorand generalmusthaveoccupiedseveralweeks. Sincehiscommandhad beenan extraordinaryone to which no immediate successor wasappointed,Germanicuswasableto postponehisreturnuntil the springof 17,in part perhapsbecauseof hisnewlyborn child.The family musthave left the Rhine by the beginningof April, certainlyno later and probably somewhatearlier. Germanicuscould not have arrived in the neighbourhood of Rome long before his triumph on 26 May without enduring an enforced sojourn outside the pomerium.48 His wife's final pregnancywould have begun about this time, and was far advancedwhen sheand young Gaius set out for the East before the end of the year (Suet. Cal. 10.1). Whether or not Agrippina accompanied her husband to Nicopolisand Athens, they were undoubtedlytogetheron Lesbosfor the birth of Julia Livilla early in 18--almost exactly three years after the delivery of Drusilla in Germany. VI

It is possibleto carryour investigationonestepfurther.TheActa Fratrum Arvalium supply us with fragmentary referencesto sacrificeson the birthdaysof variousmembersof the imperial family. It is this record,for example, that preservesthe anniversaryof Julia Agrippina'sbirth for 57 and 58; but even the most determined efforts at restoration have not

produceddefinite datesfor the birthdays of her sisters.The Acta of 38, the first from Gaius' reign, list sacrificesin honour of Livia (30 January), Antonia (31 January), Germanicus(24 May), Augustus(23 September), and Tiberius (16 November); the record for the following year adds the birthday of the elderAgrippina(25/26 October);and that of 40 confirms the date of Germanicus'birth? There are in addition two fragmentary notationsfrom which the namesof thosehonouredare missing:sometime between6-12 Februaryof 39 and perhapsbetween2-4 Juneof 40 therewere imperial birthdays.50 In hisnoteto CIL 6.32346e,thefirst of thesefragments,Huelsenrejected the date in early February for the birthday of either Drusilla or Livilla, on the basis of Mommsen's argument that each was born near the end of a year, and suggestedinsteadthat the missingname might be that of Lollia Paulina, Gaius' third wife.5•A similar solutionwas offeredby Mommsen for the birthday in earlyJuneof 40, althoughhis restorationof[Caesoniae uxoris C. Caesaris] Germanic.Aug. leavesthe last two imperialnamesin the reverseof their normal order.•2 It should be noted, however,that in the extant portions of the Acta from 14 to 66 the two Augustae--Livia in 27 or

134

J. HUMPHREY

28 andJuliaAgrippinathirtyyearslater--are theonlylivingpersons (apart from the Emperors) who were awarded such sacrifices.While Gaius was

seldomrestrainedby established precedent,it would be surprisingto find eitherhiswivesor hissisterssosingularlyhonouredduringtheirlifetimes. 53 On the other hand, we shouldexpectthe Arvals to acknowledgewith a sacrificethe birthday of the Emperor'sdead, and now deified, favourite sister;and Drusilla,aswehaveseen,couldwellhavebeenbornasearlyas February.Moreover,this birthdaybetween6-12 Februarywashonoured for the first time in 39: the Arval recordsfor the previousyear, almost completefrom Januaryto May, showno sacrificein earlyFebruary,when Drusilla

of course was still alive. vii

To summarize.It is possible,and I think evenprobable,that Julia Drusilla, born in the territory of the Treveri in the first or secondweek of February, 15, was the eldestdaughterof Agrippina Major and Germanicus;and that the birth of her youngersister,Julia Agrippina, should be assignedto 6 November of the following year. This rearrangementof the traditionally acceptedorder of eventsis by no meansconclusive,and I put it forward with only as much confidenceas the evidencejustifies. It is, of course, dependenton severalassumptions--each one more likely than not--about the movements of the elder Agrippina and about the duration of her pregnancies.But more important, it rests on the supposition that the meagertestimony of our ancient sourcesisvalid and needsno correctionto have it comply with an inventedchronology.54 University of Calgary

John Humphrey NOTES

The followingspecialabbreviations are usedthroughout the Notes: BMC Emp. 1 : H. Mattingly,Coinsof the RomanEmpirein the British MuseumI: Augustusto Vitellius(London1923;rev. 1976)

EJ2 = V. Ehrenberg andA.H.M. Jones,Documents illustrating thereignsof Augustusand Tiberius,2nd ed. (Oxford 1955)

Henzen,AFA = W. Henzen,ActaFratrumArvaliumquaesupersunt (Berlin 1874)

Smallwood, Documents = E.M. Smallwood, Documents illustrating theprincipatesof GaiusClaudiusand Nero (Cambridge1967)

THREE

DAUGHTERS

OF AGRIPPINA

135

1. CIL 6.2298, 2300 = EJ2 p. 51; Suet. Cal. 8.1; cf. HA "Commodus"1.2; 10.2. 2. Julia Livilla was the youngest of the nine (Tac. Ann. 2.54.1: novissimo

pa•'tu),andweknowfromPliny,HN 7.57thatthegirlswerebornoneafterthe other.

3. Ann. 2.53.1-54.1.Visitsto Athensand EuboeaprecededGermanicus'arrival on Lesbos,probablyin late Januaryor in Februaryof 18. 4. The date is recordedtwice in the AFA, for the years57 and 58 (Henzen, AFA

pp. lxiv, lxx = Smallwood,Documentsno. 19,21; cf. ILS 229),andagainin the Fasti Antiates(CIL 12p. 249 = Inscr.It. 13.2.210= EJ2p. 54). The site,latercalled ColoniaAgrippinensis,is givenby Tacitus(Ann. 12.27.1;cf. Germ.28.5);seeC. M. Wells, The Germanpolicy of Augustus(Oxford 1972) 134-136. 5. Cf. Suet. Cal. 8.3: cure Agrippina bis in ea regione(Gallia)filias enixa sit. Tacitus(Ann. 2.41.4) mentionsthat five of Germanicus'childrenwere with him in histriumphalchariot.Thesecouldonlybe histhreesurvivingsons--Nero,Drusus, and Gaius-- and Agrippina and Drusilla, sincehis three other sonshad already died in childhood(Suet. Cal. 7; 8.2-4;cf. Tac. Ann. 2.71.6; Dio 57.18.11).The date of thetriumphgivenby Tacitusisconfirmedby thefasti(Inscr.It. 13.2.462;cf. EJ2 p. 49). 6. Suetonius'passageis worth quoting in full:

extat et Augusti epistula, ante paucosquam obiret mensesad Agrippinam neptem ita scripta de Gaio hoc (neque enim quisquam iam alius infans nomine pari tunc supererat):"puerum Gaium XV Kal. Iun. si dii volent, ut ducerentTalarius et Asillius, heri cum iis constitui. mitto praetereacum eo ex servismeis medicum, quem scripsiGermanico si vellet ut retineret. valebis, mea Agrippina, et dabis operam ut valensperveniasad Germanicumtuum." The wording of the letter (e.g., pervenias) certainly doesnot suggestthat shehad been with Germanicusall along and was at this time only temporarily separated from him. Yet it is not at all clearwhyshehad not accompaniedhim to hispostearly in 13; the reason must have been a seriousone, sinceshe was seldomwilling to be apart from him (cf. Tac. Ann. 1.40.3). It can perhapsbe inferredfrom Augustus' letter that young Gaius had not beenwell. It seemslesslikely that in 13 Agrippina had beenpregnantwith oneof the two sonswho diedwhilestill an infans(Suet. Cal. 7) and so postponedher journey, though sucha pregnancycould have developed following the birth of Gaius in August of 12. It shouldbe pointedout that we rely for the date of this letter on Suetonius'

statement

that it was written"a

few months"

before Augustus' death; the only indication in the letter itself that it was not sent earlier is a strict interpretation of the word puerum applied to Gaius. Although there is no suggestionof how long before 18 May the letter waswritten, it isdifficult to imagine Agrippina travelling north during the winter months(seebelow, note 46). 7. The arrival in the camp of the senatoriallegati(Tac. Ann. 1.39.1;Dio 57.5.45) gives us our only evidencefor dating the subsequentdeparture of Agrippina. After Augustus'death Tiberius had sent letters to the legions(Tac. Ann. 1.7.8), but it wasnot until the meetingof the Senateon 17 September,at which Augustuswas consecrated,that the senatorialembassywas dispatchedto inform Germanicusof

136

J. HUMPHREY

his receipt of proconsularimperium and to consolehim on the lossof Augustus (Ann. 1.14.4). Even if they pressedahead quickly, the legaticould not have reached Oppidum Ubiorum in much lessthan a month. Drusus,it seems,had travelledthe 460 milesto the Pannoniancampin only ninedays(from the meetingof the Senate on 17 Septemberuntil the eclipseearly on the 27th (Ann. 1.28.l)), averagingsome fifty miles a day (I cannot acceptK. Wellesley'sredating of theseeventsin JRS 57 (1967) 23-30). It is also clear from Tacitus (Hist. 1.12, 14, 18) that the imperial cursus in an extreme emergencytook only eight days to pass a messagefrom Colonia Agrippinensisto Rome, a distanceof about a thousandRoman miles.The

questionis, how wouldan embassytravel?Sinceno newsof the Rhine mutinyhad reachedRome by the time of the meetingthat had appointed them (cf. Tac. Ann. 1.46.l), there wasno needfor extraordinaryhaste:theycouldmanagetwentymiles a day or so on foot, perhaps forty miles a day by wagon (see W. Riepl, Das Nachrichtenwesen des Altertums (Berlin 1913) 129-138 (marching), 147-151 (horseback),152-157(vehicles);for the imperialpost,A.M. Ramsay,JRS 15(1925) 60-74, and C.W.J. Eliot, Phoenix 9 (1955)76-80). At theseratesthe embassycould not havearrived at Oppidum Ubiorum earlierthan the middleof October;it iseven possiblethat they were not there by the beginningof November. 8. In his account of the mutiny Dio is less clear about the movements of

Agrippina(57.5-6).The soldiers seizedherandyoungGaius"6•zx•z•zq•O•¾zo• •o• 6•6 'to5 F•pp.•v•xo5... x0d.'r•v p.•v'Ayp•votv •yx6p.ov0•oS•0•v&q•qx0•v •x6'r• 8•0•¾z•, r6¾82 8• F•tto¾xaz&o•o¾."Dio is ambiguous,but he doesnot contradict the clear statementof Tacitus that Agrippina did not in fact return to the camp. If we assumethat the embassydid arrive as early as the middle of October and that Agrippina, once sent away, did not return to Oppidum Ubiorum during Germanicus'subsequentabsence,she could not possiblyhave been there on 6 November: three weeks are hardly sufficient time for the punishment of the ringleaders,Germanicus'marchwith hissoldierssomesixtymilesto Veteraand his delay outsidethe camp there until the massacrewas over, the bridgingof the Rhine (cf. Caesar, BG 4.17-18; 6.9), cutting of roadsthrough the forest, ravagingof the countrysideand slaughteringof the population, the legions'return to the Rhine through an ambush,and finally Germanicus'trip back to Oppidum Ubiorum with part of his army. 9. "Die Familie des Germanicus", Hermes 13 (1878) 245-265: Gesammelte Schriften4 (Berlin 1906)271-290. Most later scholars,trustinglessin the accuracy of Suetonius'continuotriennio, have not acceptedMommsen'searly date for the birth of Livilla, but have adopted his chronologyin the casesof Agrippina and Drusilla. See,for example,PIR 2 164 l, 664, and 674: Henzen,AFA li n. 4 (cf. p. 53); H. Furneaux, The Annals of Tacitus,2nd edition (Oxford 1896) 169-170;F.R.D. Goodyear, The Annals of Tacitus(Cambridge 1972), on 1.44.1;J.A. Maurer, A commentaryon C. Suetonii Tranquilli Vita C. Caligulae Caesaris(Philadelphia 1949)23-24; and J.P.V.D. Balsdon,The Emperor Gaius(Oxford 1934)7 n. 2, and 15-16 (also his articlesin OCD 2 on "Agrippina (2)" and "Drusilla"). 10. Hermes 13 (1878) 252 = Ges. Schr. 4.277. I I. Hermes 13(1878) 254: Ges.Schr. 4.279: "... die drei T/Schtergeborensind in drei auf einander folgendenJahren." 12. Hermes 13 (1878) 260: Ges. Schr. 4.285: "... diese Schwangerschaftder

THREE

DAUGHTERS

OF AGRIPPINA

137

Agrippina mit einer Frtihgeburt geendethat." Mommsen wasright to rejectthe idea that this birth could be assignedto one of the two sonswho died in infancy (Suet. Cal. 7; the third early death. of a son called Gaius, must obviouslyhave occurred before 31 August, 12). To his argumentsshould be added the epitaphsof the two infantes(ILS 181a and 181b) found near the Mausoleum of Augustus,in which the words hic crematusest allow little possibilitythat either could have died on the northern frontier: if the remains had been transferred to Rome for burial, we should

expecta phraseof the type found in CIL 6.35020 (... cuiuscremati reliquiae hoc loco positae sunt). At least one of these two boys must have been born, as Mommsen conjectured, between 8 and 10, before the birth of the first son called Gaius. See above, note 6. 13. Petira inde Euboea tramisit (sc. Germanicus) Lesbum, ubi Agrippina novissimopartu Iuliam edidit. Of the tenseof the verb Mommsen observed:"Abet es ist ein leichter Irrthum, dessengleichen bei Tacitus keineswegsselten sind" (Hermes 13 (1878) 255 = Ges. Schr. 4.280). 14. Though Mommsen gave it no consideration,he was quite right in assuming that Dio was referring here to the anniversaryof Drusilla's birth. The distinction in

classicalGreekbetween'r&¾e•i0Xt0• (thebirthdayof a livingperson)and'r&¾e,•i(•t0t (the anniversaryof a person'sdeath) had becomeconfusedby Dio's time (for the classicaldistinction, see F. Jacoby, CQ 38 (1944) 65-75). Josephus,for example, twiceuses¾z,•t•to½ (sc.•z•p0•) wherewe shouldexpect¾z,•0Xto½ (AJ 12.196,215), and it occursthus in the Gospels(Ev. Matt. 14.6: Ev. Marc. 6.21). So it is not surprisingto find a fourth-century grammarian describingat length the impropriety of using genesia of the living (Ammonius, De Adfinium Vocabulorum Differentia 116). Dio was more careful than somein distinguishingbetweenthe two words. Only once does he use genethlia of a posthumousbirthday celebration (78.11.6, of

SeptimiusSeverusin 217), but he regularlyappliesthis meaning,rather than the classicalone, to genesia:see47.18.5-6, on Julius Caesar'sbirthday after 42 Bc(cf. Inscr. It. 13.2.481-482); and particularly 60.5.1. where Claudius decreescircus gamesiC -r&¾•,•kot0• of his dead parents(cf. Suet. Claud. 11.2-3).Genesiain this sense--the birthday of a personnow dead--appears also in inscriptions:in IGRR 4.353N for example,¾•,•(•t• Z•[%t(rro• is 23 September. 15. See A.A. Howard and C.N. Jackson, Index verborum C. Suetoni Tranquilli (Cambridge, Mass. 1922). 16. Jul. 24.1: 89; Aug. 8.1: Tib. 13.1: 20: 76• Cal. 25.1. 17. Jul. 42.1; Aug. 34.1; Tib. 16.1 (triennio): 38: 48.1• 51.2: Cal. 59: Vesp.2.1: Tit. 1l: Dom. 9.2: seealsoAug. 26.2 (an extraordinaryapplicationofbiennium, to which we might comparethe MS reading of Tac. Ann. 3.31.1): 32.3: Tib. 9.3:16.1 ( quinquennium); Cal. 1.l; Galba 7. l: 15.1: Vit. 5. Apart from the duration of offices entered on i January, only once does any of these words seem to refer to a successionof calendar years: compare Tib. 15.2 ( Gaio et Lucio intra triennium defunctix) with Aug. 65.1 ( C. et L. in duodevigintimensiumsparioamisit ambos): their deaths occurred on 19 September of 2 (Lucius) and on 21 February of 4 (Gaius). ! doubt that quinquennium in Aug. 65.3 should be taken as a second example of this. 18. Compare the anecdote about Biberius Caldius Mero (Tib. 42.1): Noctem

138

J. HUMPHREY

continuumquebiduum epulando potandoque consumpsit.That he indulged himselffor a periodof two successive daysis obviousfrom biduumitself;that hedid it withoutanyinterruptionsisimpliedby continuum.Anotherusefulparallelcanbe found in Livy 45.9.2: . .. cum quadriennium continuum bellaturn esset...; cf. 33.46.7.

19. The terminuspost quem for the Emperor'sdeparture is clear from Dio 59.20.2 and Suet. Cal. 26.3. On or shortlyafter 2 SeptemberGaiusremovedthe suffectconsulsfrom office, allegedlybecausethey had failed to commemoratehis birthday on 31 August, thoughthey did celebratethe anniversaryof Actium two dayslater. Therefollowedan interregnumof threedaysbeforetwo newmagistrates wereappointedby Gaius,who then musthaveleft Romealmostimmediately,since he wasto arrivewith hisentourageat Mogontiacum,1000milesnorth of thecity,by 22 Octoberat the latest.SeeJ.P.V.D. Balsdon'schronologyinJRS24 (1934) 16-17. 20. Hermes 13 (1878) 254 n. 3 = Ges. Schr. 4.279 n. 4. 21. The length of the consulshipis confirmedby Suet. Cal. 17.1. 22. Similar clearly-wordeddistinctionsbetween general behaviour, random anecdotes,and chronologicallyarranged incidents("... •z• ... (x• •:ozz...) z6'rz 8• .... ") can be found throughout the imperial books of Dio. In Book 59 alone, compare chapters5.5-6.1; 7.3; 15.6; 19.1-2; 23.6-7; and 24.5. 23. Dio 56.46.5; Suet. Cal. 56.2; Tac. Ann. 1.73.4; Jos. AJ 19.75; Inscr. It. 13.2.400401.

24. Suet. Nero 22.2; Juv. 10.3642; 11.193-195. Suet. Cal. 18.1 and Claud. 7

probably refer to the occasion mentioned by Dio (59.7.4) during Gaius' first consulship. 25. On the decreeof the annual festival, seeDio 59. l 1.3;for the date of Drusilla's death, Fasti Ostienses : Inscr. It. 13.1.190-191:

Smallwood, Documents no. 31.

The date of her consecrationon 23 September,dependenton the restoration of a fragmentary portion of the AFA by Henzen (p. xlvi), is by no meanscertain: Drusilla'sname appearstwice on the tablet (lines 15 and 16), precededthe second time by [?div]ae (or by [?Iuli]ae), with no indication that this was the day of her deification. Exactly when Gaius decreedthe annual celebrationsis also a matter of doubt: Dio (59. ! l) seemsto compressthe eventsof her death, subsequentiustitium, and final deification:he may be closerto the properduration of time than Henzen's restoration

would

allow.

26. It undoubtedlytook the family at leastsevenweeksto make thejourneyto Rome: this assumesa daily average of twenty miles, an arduous task for such a group. So any birth in Germany must have occurredin February or March, with conceptionprobably in May or June of 16. But Tacitus notes that the year's campaignin 16 began maturius (Ann. 2.5.4)--certainly well before May, if the previousyear is any guide (cf. Ann. 1.55.l)--and that Germanicuswas occupied with it until the autumn (Ann. 2.26.1). It would, I think, be quite unreasonableto claim that Agrippina might have accompaniedher husband on an active and dangerouscampaignin hostileterritory, or that Germanicushimself might have returnedto basein the midst of the campaignsimply to visit his wife. At any rate, a birth early in 17 must also assumethat Germanicusand Agrippina were able to travel over a thousand Roman miles in early spring, with a child lessthan two months old, and at a speedof twenty miles every day.

THREE

DAUGHTERS

OF AGRIPPINA

139

27. The birth of two children within ten monthsis extremelyrare. though not impossible.Postpartum ovulationalmost neveroccursbefore the sixth week, even in non-lactating mothers (cf. E.W. Page et al., Human Reproduction,2nd ed. (London 1976) 312): so the fact that Agrippina employeda wet-nursefor Drusilla and her brother Drusus (ILS 1837) is of no great significance. Any computation of the length of a pregnancymust, without certain evidenceto the contrary, assumethat the pregnancylastedcloseto the normal term of 40 weeks. A term of 28 weeksis today consideredthe normal minimum for the survival of a prematurechild. In antiquity the normal minimum must have beenconsiderably longer, despite the record of such abnormal--and, to modern obstetricians, incredible--casesas that of Vistilia, the mother of Caesonia(Gaius Caligula's fourth wife), who is said by Pliny (HN 7.39) to have producedfive children in the seventhmonth and another in the eighth. Any strictcalculationof term is somewhatcomplicatedby the ambiguityof the phrase"in the seventhmonth". The seventhmonth of a pregnancyis normally taken as beginningwith the 27th week, so Vistilia's five premature children could have been born as late as the 210th day of their mother's pregnancies.By today's standardsthis is certainly possible:some5% of total birthsdo occurprematurely, that is before the 37th week, and the chances of survival for children born between

the 28th and 36th weeksare now--with modern medical techniques--about75%: lowernearthe 28th week,highernearthe 36th,and dependingon the weightof the fetus (see V.M. Crosse,The Preterm Baby, 8th ed. (Edinburgh 1975) 261). In Vistilia'scase,too, the birth of one child in, say,the 30th weekwould increasethe possibilityof equally prematurebirthsfor subsequent children.So the evidenceof Pliny in this matter cannot be disregarded. On the other hand, as the ancients themselvesrealized (cf. Gellius, NA 3.10.8; 3.16), it is extremely unlikely that any particular child in antiquity--in our case Drusilla--was born before the 37th week of pregnancyand survived.In the first place,the remarkableaccomplishments of Vistilia, abnormal enoughto attractthe attentionof Pliny, are nowherecreditedto Agrippina Major, soit wouldbefalseto draw a comparisonbetweenthe two women.Then there is the statisticalimprobability of successful prematurebirths.Todayabout49 birthsin 1000are premature; and of these49 children, some36 will survive(Crosse,loc. cit.). Even if we ignore the undoubtedly higher perinatal mortality rate of a premature child born in a Gallic town almost two millennia ago, the chancesare less than I in 27 that Agrippina's pregnancyfailed to reach closeto full term. It is safeto concludethat Drusilla was not a premature baby. On all thesepoints, see,in addition to Pageand Crosse,P.C. MacDonald and J.A. Pritchard, Williams Obstetrics,15th ed. (New York 1976) 146 and passim; H.E. Evansand L. Glass,Perinatal Medicine (New York 1976)64-65. I owethese medicalreferencesto my friend Dr. R.S. Sauveand othercolleaguesin our Faculty of Medicine.

28. Cf. Cal. 8.3, where it would almost certainly be mentioned. 29. BMC Emp. 1, Caligula no. 36 (sestertius).See also, for this coin, U. Kahrstedt,Klio I 0 ( 1910)295; G.G. Belloni, Aufstiegund Niedergangd. tom. Welt 2.1.1045-1046.

30. Tac. Ann. 4.75.1 (Agrippina); 6.15.1-4 (Drusilla and Livilla).

140

3. HUMPHREY

31. Notice, for example, the children of Antonia and Drusus(Claud. 1.6) and thoseof the Emperor Claudius( Claud. 27.1: the date of Octavia'sbirth isa problem in itself). Augustus' grandsons by Julia and Agrippa--Gaius, Lucius, and Agrippa--are also listedaccordingto seniority(Aug. 64.1), while the relativeages of his granddaughtersJulia and the elder Agrippina mustbe deducedin part from this passagein Suetoniusand from the supposeddatesof their marriages(seePIR 2I 635; PIR l V 463; Mommsen, Hermes 13 (1878) 246 = Ges. Schr. 4.272). 32. Even greater ignoranceis perhapsindicatedby Tacitus in his introductory commentson Germanicus(Ann. 1.33.2):... Agrippinam in matrimoniopluresque ex ea liberos habebat.

33. BMC Emp. 1, Caligula no. 44. 34. BMC Emp. 1, Augustusno. 106 and 108. 35. BMC Emp. 1, Claudiusno. 242 (didrachmfrom the mint of Caesarea).The birthdatesof the two girlsare uncertain,but not their relativeages:Antonia wasthe daughterof Aelia Paetina,Octaviaof Messalina.Britannicuswasundoubtedlythe youngest of the three. 36. It is unfortunate that Agrippina's three daughters are not mentioned togetherin an inscription.I canmakeno real senseof thecuriousdedicationto four of Gaius' siblings,three of whom were dead when the inscriptionwas set up at Mytilene, sometimein the year or so betweenthe deificationof Drusilla and the

banishment of Agrippinaand Livillain 39 (ILS 8789= IGRR 4.78):N•pto,•x0•[I •6'roxp&ro•o• I F•to K0•,0•po•.HereAgrippinaismentioned beforeDrusilla,but Livilla is omitted entirely. 37. The passagesare worth quoting: Ceterum Tiberius neptem Agrippinam Germanico ortam cum coram Cn. Domitio tradidisset, in urbe celebrari nuptias iussit(Ann. 4.75.1);... diu quaesitoquosneptibussuismaritosdestinaretCaesar, postquam instabat virginum aetas, L. Cassium,M. Vinicium legit (Ann. 6.15.1). Livilla wascloseto her fifteenth birthday at the time, but it would be rashto deduce any chronologicallimits from virginum aetas. 38. Dio 54.16.7; cf. L. Friedlander, Roman life and manners (kondon 1907) 4.123-131(appendix xviii); M.K. Hopkins, "The ageof Roman girlsat marriage", Population Studies 18 (1965) 309-327. 39. Tiberius' only other granddaughter, Julia the daughter of Drusus, was already married to Germanicus'son Nero (Tac. Ann. 3.29.4). 40. Seethe fragment of the laudatiofunebrispublishedwith a commentaryby L. Koenen(ZPE 5 (1970) 217-283,particularly257-268),who arguesthat the relation-

shipbetweenthe elderVarusand Tiberiusresultedfrom the formeftsmarriageto a daughterof MarcusAgrippa'ssecondmarriage,to ClaudiaMarcella.Accordingto M. Reinhold, on the other hand (CP 67 (1972) 119-121), the elder Varus' first marriagewas probably to a sisterof Vipsania Agrippina, Agrippa'sdaughterby Pomponia. At any rate, Varus and Tiberius were for a time brothers-in-law. 41. The word propinquusis a vagueonein Tacitus.It occurstwenty-fourtimes in the first six books of the Annals (cf. A. Gerber and A. Greef, Lexicon Taciteum

(Leipzig 1903)): in sevenof theseno particularrelationshipis indicated(2.3.4; 2.34.2; 2.34.6; 2.50.4; 2.88.3; 4.8.3: 4.17.3). In the other cases,the word is distin-

THREE

DAUGHTERS

OF AGRIPPINA

141

guishedfrom adfines (2.10.1' 6.8.4); from parentes (3.43.1' 4.62.6; 6.1.5), frater (4.62.6 again), liberi (1.58.8); and, as would be expected,from amici(l.61.1' 6.19.4) and clientes(1.57.4). So the relationshipindicatedby the word is not an immediate

one(propinquussanguisin 3.12.9,thoughit m.ight be takento includea brother,is purposelyvague):this seemsclearfrom its usein 2.71.4;in 4.75.2 it refersto a greatuncle, in 3.31.5 perhapsto both an uncle and a stepfather(cf. PIR 2 C 1462, 1465, and stemma), and in 2.51.2 and 2.30.4 to intermediate relationships, almost certainlyby marriage.We are left with this referenceto Varus (4.66.1) and with the related passageat 4.52.3, where the word is applied to the elder Agrippina and Varus' mother Claudia Pulchra,also calledhersobrina,technicallya secondcousin on her mother's side (cf. Festus, de Sign. Verb., "sobrinus"). From the last two passagesit would seem that the relationship known to Tacitus was not one projected by somefuture marriage, and probably not that betweenTiberius and Varus'father,but wasdistantlyderivedthroughVarus'mother,Claudia Pulchra,in turn the daughter of Augustus'niece, Marcella Maior. 42. The word generis oftenlooselyapplied to a man betrothedto one'sdaughter. L. Silanus,for example,is thus describedby Suetoniusand Tacitus, although his marriageto Claudius'daughter Octavia was neverperformed(Claud. 27.2; Ann. 12.4.3' cf. 12.9.2); Dio, too, calls him Claudius' ¾0tbq3O6q (60.21.5). M. Winterbottom, in his translation of Seneca's Controversiae for the Loeb series, was

apparently as unaware of this not uncommonapplication ofgener as he was of the meaning of praetextatus: "Quinctilius Varus, then son-in-law of Germanicusand only a very young man .... " 43. So, for example, R. Syme, Roman revolution (Oxford 1939) 496; and R. Seaget, Tiberius (London 1972) 205. 44. According to PIR • Q 26, Varus wasprobably betrothedto Julia Livilla, but no reasonis offeredfor preferringher to either of hersisters.The samerelationship is assumedby R. Hanslik in RE 24.1.987 ("Quintilius" no. 28). 45. The absenceof a detailed study of women'snamesin the early Empire makes it impossibleat this stageto draw any defensibleinferencesfrom the fact that Julia Agrippina carriedthe name of her maternalgrandfatherthroughher mother,while Julia Drusilla's name came from her father's side of the family. It would be unreasonableto compare in this regard daughterswith sons,the eldestof whom usuallybore the cognomen of his father, the seconda cognomenfrom hismother's name (see,for example, R. Cagnat, Coursd'gpigraphielatine, 4th ed. (Paris 1914) 68-72).

Nor is it possibleto deduceanythingabout the date of Drusilla'sbirth from the fact that an Egyptianmonth wastemporarilyassignedher name. If A.E.R. Boak's analysis of the "Gaian" names in the Egyptian calendar is correct Apouc•XX7•o½", JEA 13(1927) 185-186),the monthcalledafter Drusilla wasPauni, which in Gaius' reign fell between 26 May and 24 June (cf. E.J. Bickermann, Chronology of the ancient worm (London 1968)40-41; 48-49; 151). If there wasany logic in the assignmentof names,it doesnot seemto havebeenbasedexclusivelyon birthdates:althoughFeobto•¾[xe•oq (the EgyptianPachon)includedthe birthday of Gaius' father on 24 May, 1-'0t[•qoc• or Hathur fell between28 October and 26 November: and Pauni, of course, included the date of Drusilla's death. For the

142

J. HUMPHREY

Roman attitude to renamingmonths,seeSuet. Aug. 31.2; Tib. 26.2: Dio 55.6.6-7: 57.18.2.

46. This appearsto me the only way of understandingAugustus'enigmaticletter to the elder Agrippina (seeabove, note 6). It seemsquite clearthat the messagewas written after her departure from Rome but before her reunion with Germanicus• and it is certain that mother and son were not travelling together. 47. It might be argued that this hardly justifies Tacitus' phraseob imminentera parturn (Ann. 1.44.2), but it seemsimpossibleto place conceptionany earlier: indeed. Mommsen'schronologywould reducethe stageof Agrippina'spregnancy at this time by more than a month. At any rate, it would hardly be out of character for Germanicus(or Tacitus)to exaggerateAgrippina'sdangerin the circumstances. On the location of Vicus Ambitarvius--probably near Koblenz--see E.M. Wightman, Roman Trier and the Treveri (New York 1971) 129. 48. Compare the actionsof ¾iberiusin 8-7 Bc (Dio 55.8.2). Dio--as reportedby Xiphilinus--intimates that Germanicuswasstill campaigningin 17 (57. i 8.1); but a comparisonof his accountthere with Tac. Ann. i.60-62; 2.25; and 2.41.1 indicates that Dio is recording eventsfrom the two previous years. 49. Henzen, AFA pp. xli-li: CIL 6.32346e:Smallwood, Documentsno. l-! 1. 50. On 6-12 February 39 (CIL 6.32346e- Smallwood, Documentsno. 8): [A.d. ~-

l]dus Febr. [[ [k. SalviusOtho flamen]et pro ma(g)istro c[ollegiiI fratrum Arvalium]nominenatali[---I in Capito]lioIovi O.M. bove[mmarem[immolavit]. The record for 24 June of 40 (Henzen, AFA pp. 1-1i, n. 4; cf. Smallwood,

Documents no. 10)readsin part:... obnatalem[--]1 Germanic. Aug.Iovi bovem marem,I lunonivaccam,Minervaevaccam[immolavit].Therewereat leastfour other sacrificesin Capitolio during the year 38 (11 January; and on three dates between30 May and 22 September(frg. d, lines 19-39)),but thereisnothingexcept the location to suggestthat they representedimperial birthdays. 51. It is possiblethat they were no longer married by this date in 39. Their weddingseemsto havetaken placein the summeror fall of the previousyear,not longafterthe deathandconsecration of Drusilla(Dio 59.12.1).Dio and Suetonius (Cal. 25.2) both claim that the marriagewasa very brief one,although Dio later suggeststhat the divorce occurredlate in 39• apparently while Gaius was in Germany (59.23.7). 52. The survivingtranscript of the original fragment containserrors enoughin itself to make even the boldest epigraphistwary of proposingrestorations.If anotherimperialbirthday is indicatedhere, it certainlycannotbelongto kivilla, who wasin exileat the time (Suet. Cal. 29.1;39.1:Dio 59.22.8).Yet, if the copyisat all accurate,the personhonouredmusthavebeena relativeof the Emperor,and there are not many who would qualify. 53. The literary sourcesmake no mentionof specialhonoursgrantedto any of Gaius'wives.They do acknowledgeother honours--noneof them extraordinary for membersof the imperialfamily--bestowedon Gaius'sisters:the privilegesof the Vestal Virgins, which perhapsincludedthe right to a lictor (cf. Dio 47.19.4)and certainlyspecialseatsat the theatreandgames(Dio 59.3.4):andthefavourof being includedin the annualprayersandmagisterialoaths(Dio 59.3.4;Suet.Cal. 15.3:cf. Aug. 58.2), though--most significantly--neveras individualsor by name (for

THREE

DAUGHTERS

OF AGRIPPINA

143

examples,seeSIG 797= Smallwood,Documentsno. 33; AFA for 12January,38, as restored by Mommsen (CIL 6.2028b : 32344) and Henzen (AFA p. xliii) : Smallwood, Documentsno. 2). Moreover, after the suppression in October,39, of Gaetulicus'conspiracy,in which Agrippina and Livilla were implicated,Gaius prohibitedthe future award of any honoursto membersof hisfamily (Dio 59.22.9): this presumablyincludedhis wife Caesonia. 54. Whatever value this article has is owed in large part to ProfessorE. Badian and the two anonymous refereesfor the Journal, whose careful reading and perceptivecriticismof an earlierversionrescuedmefrom someerrorsof omission and judgement.

ZUR

ZEREMONIE IN ROM

DER NAGELSCHLAGUNG UND IN ETRURIEN

t)ber den r6mischenBrauchdes clavumfigere bzw. pangereliegen zahlreiche

Nachrichten

lateinischer

Autoren

vor.

Die Problematik

des

Wertes und der BedeutungdiesesBraucheswurde von philologischerwie von religionsgeschichtlicher Seite bereitsdiskutiert und hat zu mehreren Ansichten geftihrt; daneben treten in der Literatur verschiedene Vorschl•ge rechtshistorischen Charakterstiber das Amt und die amtliche Stellung des Magistrats auf, der diesenBrauch ausgefiJhrthaben soll.•

Im folgenden sollderVersuchunternommen werden,diesenOberlieferungskomplex abermals darzulegen und durch religionsgeschichtliche Er6rterungen erneut zu beleuchten. Untersuchtman die antiken Quellen, sowerdenftir Rom zwei Riten des

Nagelschlagens tiberliefert.EinerdavonsollnachLiviusmit der Grtindung des Juppitertempelseingeftihrt worden sein. Er fand j•hrlich statt und diente zur Jahresz•hlung:Der h6chsteMagistrat schlughierftir an den Iden des Septembereinen Nagel ein, nach einem alten, an der rechten Seitenwandder aedesCapitolina angeschlagenen Gesetz.2 Nach Paulus Diaconus wurden die j•hrlichen N•gel in den W•inden von aedessacrae fixiert.3 Cicero und Plinius bringen, nach manchen Forschern, weitere einschl•gige Hinweise.4 Man mug solche Hinweise jedoch ablehnen: Cicero verwendet das Wort clavus in bezug auf den Stift eines Steckkalenders,aber nichtaufeinen Jahresnagel.5Pliniuserw•hnt nirgends,daf5

derkurulische •dil Flavius(anl•l•lichderWeihungderaedesConcordiae, welche204 Jahrenachder WeihungderaedesCapitolinaerfolgte)im Jahre 449 a.u.c. 204 N•gel in der Seitenwandder aedesCapitolinagez•hlt h•tte. 6 Die herkbmmliche Meinung rtihrt von einer ungenauenAuswertungder Pliniusstelle sowie von einem gegenseitigenAbschreiben innerhalb der modernen Forschung her.7 Es ist zwar m6glich, daf5 die von Plinius erw•hnte Datierung post Capitolinam dedicatam auf diesen Brauch zuriickgeht.8 Belegtist die Meinung in den antiken Quellenjedoch nicht. Die Stellen Ciceros und Plinius' dtirfen aufgrund des Besagtenkaum berticksichtigtwerden. Ein weiterer Ritus des Nagelschlagens wurde sporadischdurchgeftihrt und diente zum Fixieren: Negative Kr•fte wurden bei verschiedenen Anl•ssenabgewehrtund entstihnt,indemman siefestlegte.Liviusund die Fasti nennen die im Jahre 363 v.Chr. erfolgte Wahl des Diktators L. Manlius Imperiosus, welcher eine Nagelschlagungvornehmen sollte, um die in Rom wtitende Pest zu bannen: 9 Livius bezeichnet diese Zeremonie

144

NAGELSCHLAGUNG

IN ROM

UND

ETRURIEN

145

als "alten" Brauch. Es ist allerdings bemerkenswert, dag er bei der Schilderungfrtiherer Seuchemetwa der Jahre 463 und 453 v.Chr., diesen

Brauchnicht erwfihnt,wasden Gedankennahelegt,dal• zujener Zeit eine Nagelungals EntstihnunggegenPestentwedergar nicht bekanntwar oder von seinenGewfihrsmfinnernnicht aufgenommenwurde.•0 Livius berichtet auch, dalg ein Magistrat ftir eine Nagelschlagung anlfilglichder Seuchevom Jahr 313 v.Chr. gewfihltwurde;• im Jahre 310 v.Chr. wurde dieset Brauch noch ausgetibt.•2 Die Nagelschlagungals Piaculum in Verbindung mir Seuchen ist somit ftir Rom ftir die Zeit zwischen363 und 310 v.Chr. eindeutig belegt: Das Jahr 363 v.Chr. ist, Livius folgend, ein Terminus ante quem. Nagelschlagungen als magischefBann ohne Bezug auf eine Seuche werden jedoch in Zusammenhang mit abscheulichenEreignissenvon Livius erwfihnt. Er koppelt zwei Nachrichten, wobei eine als Rechtfertigungder anderenerscheint:RifmischeMatronen, im Jahre 331 v.Chr. der Giftmischereibeschuldigt,werden zum Tode verurteilt, eine Strafe, die schon"frtiher", und zwar anlfilglicheinerVolkserhebung,verhfingtworden war. •3 Man sieht keinen Grund, an dieset Tradition zu zweifeln. •4 Die

Berichte des Livius tiber Volkserhebungender vergangenenZeit und die

entsprechende 1Dberlieferung bei Dion. v. Hal. nennenjedochden Ritus nicht. •5

Nach Plinius fixiert eine Nagelschlagung,die an einem Ort erfolgt, wo ein Kranker gesttirztist, die finsterenKrfifte der Krankheit.•6 Der Nagel dient bei Plinius auch als Abwehr oder Schutz vor den schfidigenden Einfltissen

der Krankheit

und des Donners:

Er wirkt als Amulett

oder als

Talisman.•7 Ebenfalls im Sinne einer Fixierung negativer Krfifte mifchte

ichdenNageldeuten, derda• beschriftete Bleimanchef defixionum tabellae durchbohrt:•sDie (ihrem Charakter nach) negativenFltiche sowie die feindlichen Worte sollen fixiert bleiben. •9 Auf

einer

Malerei

aus Herculanum

aus dem

1. Jahrhundert

v. Chr.

besiegeItVictoria das unverfinderlicheSchicksal,indem sieeinen Nagel in ein Tropaeumvor dem siegreichenFeldherrneinschlfigt. 20In der Dichtung des Horaz trfigt NecessitasNfigel, zusammenmir Keilen, Klammern und Blei: Sie fixiert

damit

das Schicksal aller Menschen. 2• Die Hfirte und die

UnbeugsamkeitdesNagelsund desSchicksalswerdenbetont.22Ciceround Petroniusheben dieseQualitfiten besondershervor, wenn sieden Nagel in einer sprichwifrtlichenWendung nennen:23Sie drticken gleichzeitigden Gedankenaus, dalgdas, was mir dem Nagel fixiert wird, nicht mehr bewegt werden kann. Die Nagelschlagungzum Fixieren ist eine magische Handlung, die nach dem Prinzip der Analogie geschieht:Der Nagel ist primfir ein Gegenstand,der eingeschlagen wird, um andere Gegenstfinde festzuhalten.

Petronius

erwfihnt

ihn auch in diesem Sinn. TM

146

L.A.

FORESTI

Die antiken Autoren nennen nicht immer den AnlaB einer Nagelschlagung:Die Fasti vom Jahr 263 v.Chr. nennenlediglicheinendictator clavifigendi causa,Cn. Fulvius Centumalus;2•Cass.Dio, als einzigeantike Quelle, erwihnt nur eine Nagelschlagungseitensder Zensoren bei der Aufzihlung der yon Augustus an den Mars-Ultor-Tempel erteilten Vorrechte.26

Es lieBensich zwei Arten yon Nagelschlagungen for Rom nachweisen, zum Jahrzihlen und zum Festlegen(yon negativenKr/4ften,desSchicksals und yon Gegenstinden).Der yon Horaz erbrachteZusammenhangmit Gegenstinden,die zum Bauenerforderlichsind, machteswahrscheinlich, dab eine dritte, nicht expressis verbis tiberlieferte Vorstellung eines zusammenhaltendenNagels in Rom bekannt war. Ein Teil dieserBr/4uchewar Rom und Etrurien gemeinsam.Die Nagelschlagungals Jahreszihlungwurde nach Livius, der sichauf den Antiquar L. Cinciusberuft, auch im Tempelder Nortia zu Volsinii ausgeiibt:•vZur Zeit des Cincius, d.h. im 1. Jahrhundert v. Chr., konnte man in Volsinii

nochdie Nilgel sehen.28Es l[iBtsichkeinesfallsmit Sicherheitsagen,ob der Ursprung dieses Brauches in Rom oder in Etrurien zu suchen ist.29 EtruskischesKulturgut liBt sich allerdings in Rom for das sp/qte6. Jahrhundert

nachweisen. 30

Die im punischenText der aus einemTempel yon Pyrgoi stammenden Bronzeblecheerwihnten "Sterne" wurden mit den Goldnigeln gleichgesetzt,welchein einer Tonplatte desTempelssteckten,und mit den clavi annales identifiziert.

3• Ich kann mich dieser These nicht anschlieBen: Die

Zahl der Goldnigel sowieder angenommenenclavi annalesbleibt immer beschr[inktund liBt sichdaher kaum reit Sternenvergleichen,deren Zahl unendlichseinsollte,wie esder Optativ der Wunschformelzum Ausdruck bringt.32 Ein etruskischerSpiegelaus dem Ende des4. Jahrhundertsv. Chr. mit der Darstellung eines magischenRitus bezeugt eindeutig eine Nagelschlagungals Schicksalsfestigung, welcheyon der Schicksalsg6ttinAthrpa vollzogenwird.33In etruskischenGrabfreskenfinden sichbemalteT0ren zum Jenseits, die mit bemalten bullae verziert sind.34Sie sind ein Abbild der Tore heiligerund profanerBauten,aufderen FliigelnNilgel angebracht waren: Diese waren technisch bedingt oder dienten als sparsame Verzierung.35Es fragt sichjedoch, ob nicht eine urspr0nglicheVorstellung zugrundeliegt, nimlich diejenige,dab damit das Schicksalunwiderruflich gemachtwerdenkann, bzw., dab etwasNegativesverschlossen bleibensoll. Eine Verbindung zwischender r6mischenSchicksalsg6ttinFortuna und der etruskischenNortia wird yon Juvenal hergestelltund yon Martianus Capella sowiedurch weitereQuellen untermauert.36 Eine Nagelschlagungzum FestlegennegativerKr[ifte ist for Etrurien

NAGELSCHLAGUNG

IN ROM

UND

ETRURIEN

147

literarisch nicht belegt. Sie ist allerdingsm.E. anzunehmen:In Etrurien kommen,wie in Rom, tabulaedefixionumvor, die mit fixierendenN•igeln durchbohrt sind.37Ob allerdings Etrurien die Zeremonie einessollemne clavifigendigekannthat, bleibtfraglich;dennLiviuserw•ihnt,daf5Spiele zum Abwehren sch•idlicherKr•ifte apotrop•iischeMagnahmen waren,die Rom aus Etrurien einftihrte, wobei er aber eine Nagelschlagungnicht nennt.38Er zeigt auch, dag der clavuspiacularis,in Anschlugan r6mische Br•iuche,in Rom wieder aufgenommenwurde.39

Die l[lberlieferungen tiber die Nagelschlagung in Rom, die wegen mancherAspekteauch ftir Etrurien geltendgemachtwurden, bilden, nach einigen Forschern, zwei Traditionsstr•inge,die von Livius zusammengeworfen worden seien:40 Eine magische Handlung apotrop•iischen Charakters, mit welcher man das Schicksalfestigte4• oder b/3seKr•ifte vertrieb,42seij•ihrlich ausgeftihrtworden;einej•ihrlicheNagelschlagung, an derenHistorizit•itkaum gezweifeltwird,43k6nnedurchauszur Z•ihlung der Jahre gedient haben; Livius habe den Ritus rationalisiert,wenn er die Nachricht einer apotrop•iischenNagelschlagungmit derjenigen einer Jahresz•ihlungverkoppelte. lch m6chte allerdings nicht ausschliegen,dag N•igel, zumindest ursprtinglich, primfir der Jahresz•ihlungdienten.44 Diese Methode (mit

Magdelain)abzulehnen,bedeutet,die l[lberlieferung des Livius allzu modern auswertenzu wollen. Das Vorhandenseinvon Nagelurkundenmit k6niglichen lnschriften sowie Hinweise in literarischenQuellen des Alten Orients bieten m6glicherweiseein entsprechendesVergleichsmaterial. 45

Der Aufbauderlivianischen l[lberlieferung istallerdings konsequent: Der Autor rechtfertigt den clavuspiacularis aufgrund des clavus annalis• um anschliegendwieder an den clavuspiacularis anzukntipfen.46 Man wird daher zur Zeit des Livius einenuns nicht mehr ersichtlichengemeinsamen Nenner ftir beide Br•iuche angenommen haben. Der Vergleich mit zahlreichen Beispielenaus der Vflkerkunde sichert die Interpretationeiner Nagelungin Rom zum AbwehrennegativerKr•ifte durch Fixion. Eine solcheFixierung h•ilt Leidenund Krankheiten bzw. die entsprechendenD•imonen fest,47 sie fesselt Ungltick und Schmerzen in Steine, Mauern und Ttirpfosten, sie stecktSeuched•imonenin L6cher, und sieh•ilt Tote zurtick:48Sie isteineursprtinglicheForm von Zauberei, durch Jahrtausendeweltweit verbreitet und bis zur heutigenZeit belegt.49 Einige Gegebenheitenlassenjedoch daran zweifeln, dag der clavus annalisin Rom denselbenapotrop•iischenCharakter wie der clavuspiacularis gehabt habe. lch weise auf folgende Beobachtungenhin: Die Nagelungals Jahresz•ihlungh•ittean einemTempel stattgefunden? 0 die in bezug auf die Pest jedoch wahrscheinlichnicht.5• Der Umstand, dag ein Ritus ordentlich-periodisch,der andere augerordentlich-sporadisch war,

148

L.A. FORESTI

1/igt auf verschiedenen Charakter der beiden Riten schliegen. Die (•ber-

lieferungdes clayuspiacularis bezieht sichauf eine sp•itereZeit als die des clayus annalis, was ein spriteresAufkommen des letzteren Brauches bedeuten kann. Eine apotrop•iischeZeremonie zur Vertreibung sch•idlicher Kr•ifte l•igt sichnicht ohne weiteresan einem Tempel vollziehen.Ich m6chte daher das Problem der Bedeutung dieses rituellen Aktes noch einmal aufwerfen,wasauf die Frage hinausl•iuft,ob andereVorstellungen bei der Bildung des r6mischen Brauchesder Jahresnagelungeine Rolle gespielthaben. Bei den Hethitern tritt die Vorstellungauf, der Nagel strahleeinepositive Kraft aus, die durch das Einschlagenst•indigwirksam werden und etwas festlegensoil: Eine magischeNagelschlagungim Ritual sieht vor, dab auf belden Seiten des Tempeltores ein Nagel eingeschlagenwerde.52 Diese Nagelschlagungist nicht sosehr mit dem Gedanken einer sich wiederholenden apotrop•iischen Handlung, als vielmehr mit dem eines einmaligenZu-sich-Rufensund einer Beibehaltungder g6ttlichenKraft bei einer Dedikationszeremonie

zu verbinden.

In Rom steht der clayus annalis

mit einer Tempelgrtindungszeremoniein Zusammenhang, welcher kosmischeBedeutung zukam.53 Eine positive magischeMitwirkung der N•igel am Zusammenhalt des Weltgeftigestritt im EposRagnarO'kauf; in einertechnomorphenSternenerkl•irung ist der Polarstern der Nagel, welcher den Himmel an seinem Platz h•ilt. Diese Art von Nagelung verbindet sich hier auch mit dem Gedanken einer magischenAnteilnahme an der kosmischenOrdnung.54 Wir finden bei einem lateinischenAutor die Gestalt einer Nageltr•gerin, n•imlich der Necessitas, deren Ger•ite zum Fixieren des Schicksals und wahrscheinlich

auch zum Zusammenhalten

dienen. Auch hier kann man

von einer positivenmagischenMitwirkung der N•igelsprechen,sowievon einer magischenAnteilnahme an der Ordnung des Mikrokosmos. Das Modell von Sternenn•igelnals Zierde tritt in den kosmogonischenVorstellungen des Anaximenes auf.55 Im finnischen Kalewala-Epos wird das Himmelsgew61bevom Schmied Ilmarinen gemacht und mit silbernen Sternen ausgestattet. 56 In bezug auf Rom fragt es sich, ob nicht die Nagelk6pfeder claviannales,die im mythischenDenkenmetaphorisch die Vorstellung von Sternen hervorrufen, als Zierde des Himmels und daher als positiv empfundeneJahresn•igelzu betrachtensind, da sie mit dem Juppitertempel,dem kosmischeBedeutungzukommt, zusammenh•ingen. N•igel als Talismane, d.h. positiv wirkende Gegenst•inde,werden von Plinius erw•ihnt und sind noch in der heutigenZeit bezeugt.57Die Feststellung,dab N•igelauch als positivwirksamempfundenwurdenund dal• sich eine Nagelung an einem Tempel nicht so sehr mit der Bannung negativer,als vielmehrmit demFesthaltenund Herbeirufenpositiver,etwa

NAGELSCHLAGUNG

IN ROM

UND

ETRURIEN

149

gbttlicher Kr/ifte verbinden l/ifgt, weisen m.E. darauf hin, dafg die Zeremonie

des clayus

annalis

in Rom

nicht

eine Abwehr

sch/idlicher

Kr/ifte, sondern die Best/itigung, die Fixierung und gleichsam die Erneuerung und Bewahrung einer wirksamen giJttlichen Kraft bedeutet hat.s8Der Ritus erfolgtnichtan irgendeinemTempel, sondernam Haupttempel Roms, welcherder Inbegriff desStaatesund seinerHeilskraft ist, wie die in Rom mit diesem Ritus betraute Gestalt despraetor maximus beweist,und ist mit dem BestehendiesesStaatesverknfipft. Trifft das zu, dann folgt, dal• Livius nichtsmehr yon einerNagelungin Zusammenhangmit positiven Kr/iften gewufgtund daher die yon seinen Gew/ihrsm/innernfiberlieferteNachricht einer Jahresz/ihlungan der aedes Capitolina mir der Nachricht eines noch in sp/iterer Zeit in Rom stattfindenenRitus apotrop/iischenCharaktersverknfipft hat: SeineInterpretation ware sozusageneine spatereVerbindung und Deutung nicht mehr

durchsichtiger Riten,wobeieinet]berlieferung fiberdieVertreibung yon Pestd/imonennicht dagegengesprochenh/itte, denn die Erneuerungder Kraft w/ire auch bei Pest sinnvoll gewesen.Die Wiederaufnahme des

Brauches zur Zeit desAugustus und seinel•bergabean einenTempel (zweifellos ein Privileg!) zeigt, dafgzur Zeit des Augustusnur mehr eine Bedeutung des Brauches empfunden wurde, n/imlich diejenige als Stihnemafgnahme.

Auch ffir das an Quellen weit sp/irlichere Etrurien w/ire eine solche Vorstellungdenkbar: Ist doch die Wichtigkeit desNortiakultes zu Volsinii durch mehrere Inschriften belegt und Nortia wie auch Athrpa als (unbeugsame) SchicksalsgiJttinbekannt.S9Man k6nnte sich auch noch fragen, ob nicht den im punischenText yon Pyrgoi erw/ihnten"Sternen",60 die, wie wir schon sahen, aufgrund ihrer beschr/inktenZahl nichts mit schonvorhandenenclavi annaleszu tun haben,die Vorstellungyon positiv empfundenen N/igeln zugrunde liegt: lhre Nennung wtirde somit den Wunsch ausdrficken,"solangedieseN/igel (zum Zusammenhaltder Welt) eingeschlagenwerden", d.h. wohl, solangeder Staat besteht. Die Vorstellung eines sich in Rom und in Etrurien wiederholenden Tempelritus, dem kosmische Bedeutung zukommt, scheint mir diesem

komplexenl•berlieferungsstrang deutlicherRechnungzu tragen.Diese Interpretation liJstjedoch keinesfallsdie vielenProbleme,die mit der Stelle Liv. VII 3,3 ff. verbunden sind: Sie bleiben bestehen,denn die Beschaffenheit der Quellen gestattet es nicht, ein vollst/indigesBild zu gewinnen. Gleichwohlfragt es sich,ob der bier dargelegteVorschlagnicht als Basis weiterer Untersuchungendienen kiJnnte.6• Universit/it Graz (Austria)

Luciana Aigner Foresti

150

L.A.

FORESTI

ANMERKUNGEN

1. Th. Mommsen,Die r6mischeChronologie(21859) 176-180;O. Leuze, Die r6mischeJahrziihlung.Ein Versuchihregeschichtliche Entwicklungzu ermitteln (1909) 159-163;G. Favaro,I1clavusannalise il Dictatorclavifigendicausa,in Atti del I. Congresso nazionaledi studiromani II (1929)223-229;A. Momigliano, RicerchesullemagistratureromaneI: I1 Dictator clavi figendicausa,in BCAR LVIII ( 1931)29-55;idem, Praetor maximuse questioniaffini, in Studi in onoredi G. GrossoI (1968) 161-175(beide in Quarto contributoalia storia deglistudi classici e delmondoantico(1969)273-283und403-417neugedruckt;im folgenden beziehensichdie Zitate auf diesenBand);K. Hanell, Das altr6mischeeponyme Amt (1946) 118-144; idem• Probleme der r6mischen Fasti, in Entretiens sur

l'Antiquitdclassique XIII (1967) 177-191;D. Cohen,The Originof RomanDictatorship,in MnemosyneX (1957) 300-318;R. Werner,Der Beginnder r6mischen Republik. Historisch-chronologische Untersuchungen fiber die Anfangszeitder liberarespublica(1963)26-35;J. Heurgon,L. Cinciuset la 1oidu ClavusAnnalis,in Athenaeum N.S. XLII (1964) 432-437; A. Alf61di, Early Rome and the Latins (1965) 47, 78 f., 325; idem, R6mischeFriihgeschichte. Kritik und Forschungseit 1964 (1976) 97-110; M.J. Pena, La "Lex de clavo pangendo",in HAnt VI (1976)

239-265.•ltere Bibliographie tiberdieses ThemabeiE. Meyer,R6mischer Staat und Staatsgedanke(1948) 158-160.

2. Liv. VII 3,5: Lex vetustaest,priscislitterisverbisque scripta,ut quipraetor maximussit idibusSeptembribus clavumpangat;fixa fuit dextro lateri aedisIovis optimi maximi ex quaparte Minervae ternplumest.eum clavum,quia raraeper ea temporalitterae erant, notam humeri annorumfuisseferunt eoqueMinervae templo dicatam legem quia humerus Minervae inventurnsit. Wie aus Livius hervorgeht,wurde das Gesetzund nicht die N•igel an die rechte Seitenzella angeschlagen (so bei Ch. Daremberg-Edm.Saglio,DictionnairedesAntiquit•s grecqueset romainesI 2 (1877) 1241;G. Colonna,I1 santuariodi Pyrgi alla luce delle recenti scoperte,in SE XXXIII (1965) 219). 3. PaulusDiac. Exc. p. 56, 10M (=Festus,design.verb.p. 49 L): clavusannalis appellabatur,quifigebaturin parietibussacrarumaediumper annossingulos,ut per eos numerus colligeretur annorum.

4. Favaro, op. cit. 229; Momigliano, op. cit. 282. 5. ad Att. V 15: Laodicem veni pridie Kal. Sext. ex hoc die clavum anni movebis.So auch bei Hanell, Das altrO'm.epon.Amt 137f. Eine Parallelstellebei Cicero selbst ersetzt n•imlich den lateinischen Ausdruck clavum anni durch den

griechischen (ad Att. V 14:in provincia meafore meputabamKal. Sext.ex ea die ... rr0tp0[rr•yy.0t •,•0t6(•m,• commoveto). Die Bedeutung der von Cicero verwendetenLokution clavum movere deckt sich mit derjenigenyon Livius (clavum pangere) nicht ganz; die Autoren beziehen sich auf zwei verschiedene Kontexte, da die yon ihnen erw/:ihntenZeiten ftir die Ausftihrungdes Ritus verschieden sind:Liv. VII 3,5 beziehtsichaufdie Idus Septembris,Cic.adair. V 15 auf die KalendasSextiles.In diesemSinn auch Pena, op. cit. 260. 6. Plin. n.h. XXXIII 19: Flavius vovit aedem Concordiae... inciditquein

NAGELSCHLAGUNG

IN ROM

UND

ETRURIEN

151

tabella aerea.factameam aedem CCIIII annispost Capitolinamdedicatam.ida. CCCCXXXXVIIII a condita urbe gestum est. 7. Die Behauptungfindet sichbereitsin der filterenLiteratur:cf. Leuze,op. cit. 159if. und J. Beloch,R6mische Geschichte(1926) 40. Ftir die neuereForschungcf. Werner, op. cit. 27 if.; Heurgon,op. cit. 433:"sipeutomb•e en d•su•tudequ'en304 encorel'•dile curule Flavius, comptant 204 clous ..."; Alf61di, R6m. Friihg. 99: (Flavius) "hat also die Jahresnfigelam Kapitol einwandfrei abzfihlenk6nnen und hat esauch zweifellosgetan":Alf61di,Early Rome 323:"the204 annualnailsdriven in the wall of the cella of Juppiter, which were countedin 304 a.c."; A.J. Pfiffig,

ReligioEtrusca(1975) 61: "... der kurulische,•,dil Flavius204 Nilgel zfihlte". Dagegen:Hanell in Entretiens187:"Cn. Flavius als Nagelzfihlerist nattirlicheine sehrhypothetischeErscheinung";Pena,op. cit. 264;E. Gjerstadin Entretiens192f. 8. In diesem Sinn Hanell

in Entretiens

195.

9. Liv. VII 3,3: Itaque Cn. Genucio L. Aemilio Mamerco irerum consulibus, cum piaculorum magis conquisitio animos quam corpora morbi adficerent, repetitum ex seniorummemoria dicitur pestilentiam quondam clavo ab dictatore

fixo sedatam.ea religioneadductussenatusdictatoremclavifigendi causadici iussit; dictus L. Manlius Imperiosus L. Pinarium magistrum equitum dixit. Die Wahl desImperiosuswird yon der Eintragungin dieFasticons.Cap.vomJahr 363 v.Chr. bestfitigt. A. Degrassi, Inscr. Italiae XIII 1,32: [L. Manlius A.f A.n. Capitolin(us)] Imperiossus dict(ator) [L. Pinarius -f -n.] Natta mag(ister) eq(uitum) clavifig(end0 caussa. 10. Ftir das Jahr 463 v. Chr. cf. Liv. III 6-7: Die Pest wtitet in Rom, der Senat weil• keine menschlicheHilfe mehr und befiehlt lediglich, die G6tter anzuflehen (Liv. III 7,7: inopsqueSenatusauxilii humani ad deospopulum ac vota vertit ). Ftir das Jahr 453 v.Chr. cf. Liv. III 32,1 if. mit der kurzen Beschreibungeiner Seuche, die Rom und das umgebende Land verheerte, jedoch ohne Hinweise auf Sondermal•nahmen.

11. Liv. IX 28,6: pestilentia orta clavi figendi causadictatorem dictum. Livius erwfihnt nicht nut die Seuche, sondernauch die Aufgabe des Diktators. Das

bedeutet,dal• nicht die ganze1Dberlieferung tiber den Auftragdes Diktators "schwankt",wie K. Latte behauptet(R6mische Religionsgeschichte, in Handbuch der Altertumswissenschaft V 4 (1960) 154) sondernnur die Fasti cons.Cap. (Inscr. haliae XIII 1,36: C. PoeteliusC.f C.n. Libo Visolusdict(ator) C. Junius C.f C.n. Bubulcus Brutus mag(ister) eq(uitum) rei gerund(ae) caussa). 12. Liv. IX 34,12: quem clavifigendi aut ludorum causadictatorem audacter crees?Er erw•ihnt nicht ausdrticklicheinen clavuspiacularissonderneinendictator clavifigendi. Die Erw•ihnungvon ludi im selbenKontext (cf. auch Liv. VII 2,3) l•il•t auf Stihnemal•nahmen

schliel•en.

13. Liv. VIII 18,12: itaque memoria ex annalibus repetita in secessionibus quondam plebis clavum ab dictatore fixurn alienatas[que] discordia mentes horninure eo piaculo compotessui fecisse, dictatorera clavi figendi causacreari placuit. creatus Cn. Quinctilius magistrum equitum L. l/alerium dixit, qui fixo clavo magistratuse abdicaverunt.Die Wahl ist in den Fasti cons.Cap. vom Jahr 331 v.Chr. vermerkt.Inscr. Italiae XIII 1,34:Cn. Q[u]inctius T.f. T.n. Capitolinus

1.52

L.A. FORESTI

dict(ator) C. ValeriusL.f. L.n. Potitus,postea quam co(n)s(ulatu)abiit mag(ister) eq(uitum) clavifig(endi) c[aussa]. 14. Hanell, Das altrOm. epon. Amt 135, bezweifelt die Glaubwtirdigkeitder

livianischen[lberlieferung,ohne jedoch verbindlicheArgumentedagegenzu bringen. 15. Liv. II 32-33; III 52-54; Dion. v. Hal. VI 46-49; IX 67-69; X 53. 16. n.h. XXVIII 63: clavumferreum defigerein quo loco primurn caputfixerit corruens morbo

comitiali

absolutorium

eius mall dicitur.

17. n.h. XXVII146: iidem in quartanisfragmentum clavia cruce. . . liberatoque condunt cavernaquam sol non attingat; X 152: remediumcontra tonitrus clavus ferreus sub stramineovorumpositusaut terra ex aratro. J. Haekel, "Religion", in Lehrbuch der VOlkerkunde(41971) 118-120. 18. S. Audollent, Defixionum tabellae( 1904);R. Wiinsch,Inscr. Graecae(1897) III 3 (Attica); idem, Antike Fluchtafeln (21912). 19. Pfiffig, Rel. Etr. 363 f. interpretiert einen solchenNagel als "eine Form magischenT6tens", was mir wenigersinnvoll erscheint. 20. L. Barr6-H. Roux, Herculanum und Pompeft (1841) II, Tar. 67. 21. Carre. 135,17-20:te semperanteit saevaNecessitas / clavostrabaleset cuneos manu/ gestansaena necseverus/uncusabestliquidumqueplumbum. II124,5-8: si

figit adamantinos/summisverticibusdira Necessitas/ clavos,nonanitaurnmetu/ non mortislaqueisexpediescaput. Bei der Stelle135,17findet sichauchdie Lesung servanebender Lesungsaeva.Mit saevaist eineParallelstellezu dira von II124,6 gegeben.NachKiessling-Heinze writeesunm6glich,dasg6ttlicheWesenNecessitas als Dienerin zu bezeichnen:Q. Horatius Flaccus, Oden und Epoden (1964) 148;O. Jahn, in Ber. der Siichs. Gesellsch.d. Wiss.7 (1855) 106 fl. 22. cf. llI 24,5: adamantinos.

23. Cic. in Verr. V 21,53: ut hoc beneficium, quemadmodum dicitur, trabali clavofigeret; Petr. Sat. 75: quod semeldestinaviclavo trabalifixum est. 24. Sat. 135,4:turn clavum,qui detrahentemsecutuscurecamellaligneafuerat, fumoso parieti reddidit. Daremberg-Saglio,op. cit. (Anm. 2) 1240 und K.O. Miiller-W. Deecke, Die Etrusker (1877) 308 nennen irrtiimlich diese Stelle als Beweiseiner Nagelschlagungals magischeHandlung. 25. Fiir die Fasti cons.Cap. desJabres263v.Chr. cf. Inscr.Italiae XIIl 1,40:Cn. Fulvius Cn.f. Cn.n. Maxim(us) Centumalus dict(ator) Q. Marcius Q.f. Q.n. Philippusmag(ister)eq(uitum) clavifig(endi) caussa.

26. Cass.Dio LV 10,4:•X6• •r½•G,.&6rr6•r&•,[F•,½u,0t•,(o•n:poom:•)wu•00tt. Pena(op.cit. 259f.) zweifeltan derGlaubwiirdigkeit desCass.Dio: Er beweise keinesfalls,dab z.Z. des Augustusein clavusannalis von den Zensoreneinge-

schlagen wurde,da Dio dieeinzigeantikeQuelleist,welchebeimAufz•ihlender Privilegien, dieAugustus demMars-Ultor-Tempel zukommen liel•,dieseEinzelheit nennt. Der Einwand von Pena reicht m.E. nicht aus, um Cass. Dio zu widerlegen. Pena erw•ihntauch den HinweisdesSueton(Claud. 16,1),dab zwischen22 v.Chr. und 48 n.Chr. keine Zensurals ordentlicheMagistratur ausgetibtwurde (gessitet

censuramintermissamdiu post Plancum Paulumque).Sueton bezieht sich allerdings auf dieZensur,nichtaberauf denZensus (Suet.Aug.27,5:censum...

populiteregit;cf.auchResgestae 8:censum populi. . . egi). Cass.Diokannsich

NAGELSCHLAGUNG

IN ROM

UND

ETRURIEN

153

auf die Beamtenbezogenhaben,die zwar keineZensorenwaren,den Zensusjedoch austibten,soda13der Einwandyon Penam.E. nichtstichhaltigist.Wie Favaro, op. cit. 225, Hanell, Dasahr6m. epon. Amt 138, Momigliano, op. cit. 282 und Werner, op.

cit.30, m6chteichdaherdieOberlieferung desCass.Dio geltenlassen, ohnemich jedoch vorl/:iufigauf den Inhalt der Nagelschlagungzu fixieren (cf. oben S. 149). 27. Liv. Vll 3.7: Volsiniisquoqueclavosindiceshumeriannorumfixos in templo Nortiae, Etruscae deae, conparere diligens taffum monumentorum Ct.nct.us adfirmat. Einige Beispiele yon in verschiedenenOrten Etruriens belegten etruskischen Onomastika stehen der lateinischen Form Cincius nahe, so da13eine

etruskischeHerkunft fiir Cincius wahrscheinlichgemacht wurde, was fiir die Verl/:iglichkeitder Nachricht spr/:iche:Heurgon, op. cit. 436, Anm. 1. Zu Nortia cf. E. Bernert,in RE XVII (1936) Col. 1043ff. s.v.Nortia; Pfiffig, Rel. Etr. 258 f.; G. Wissowa,Religionund Kultus der R6mer, in Handbuchd. Altertumswissenschaft IV 5 (1912) 288. 28. Liv. VII 3,7:-conparere... Ct.nciusadfirmat. Nach Heurgon, in Entretiens 193, bedeutetder Hinweisdes Livius, da13die Nagelschlagung z.Z. desCinciusin Volsinii noch iiblich war, was mir jedoch eine Verallgemeinerungder Liviusstelle erscheint.

29. Nach Alf61di, Rdm. Frt•hg. 98, ist die r6mischeJahrz/:ihlung"etruskisch geworden";nach Pfiffig, Rel. Etr. 63 war der Brauchurspriinglichetruskisch: cf. auch Latte (Anm. 11) 154. Eine dritte M6glichkeit, n•imlichdie einer parallelen Entstehung,f•illt m.E. aufgrund der NachbarschaftzwischenEtrurien und Rom yon vornherein weg. 30. In unseremZusammenhangist besondersjene Nachricht zu erw/:ihnen, wonachdie Z/:ihlungder Jahre in Rom mit dem Etrusker Caecinain Verbindung gebrachtwurde (Schol. Ver. ad Aen. X 200: (Caect.na)constitut.tannum). Auch wurde die aedes Capitolt.na im Rahmen einer von Etrurien beeinflu13ten Baut•itigkeiterrichtet;auchetruskischeSeherwurdenzum Rat gezogen(Liv. ! 55,6:

t.dqueita cecinerevatesquiquein urbeerant quosquead eamreinconsuhandam ex Etruria acciverant). Das Kultbild wurde vom EtruskerVulca angefertigt:Plin. n.h. XXXV 157: Vulcam Veisaccitum,cui locaret Tarquint'usPriscusIovis effigiemin Capt.toliodicandam. 31. G. Colonna, G. Garbini, M. Pallottino, L. Vlad Borrelli, Scavi nel santuario etruscodi Pyrgi, Relazionepreliminaredella settimacampagna,1964,e scopertadi tre lamine d'oro iscrittein etruscoe punico,in ArchClassXVI (1964) 97, 104;G. Garbini, Considerazionisull'iscrizionepunica di Pyrgi, in 0,4 !V (1965)35 ff.; M. Durante, Le formule conclusivedei testietruschidi Pyrgi, in RAL 8 XX (1965) 311 ff.; G. Colonna, M. Cristofani, G. Garbini, Bibliografia delle pubblicazioni recenti sulle scoperte di Pyrgi, in ArchClass XVIII (1966) 279-282. Weitere einschl/:igige Literatur bei R. Werner, Die phoinikisch-etruskischen Inschriftenyon Pyrgoiund die r6mischeGeschichte im 4. Jh. v.Chr., in GBI (1973)241 if.; cf.auch J. Ferron, in Aufstieg und Niedergangder rb'm. Welt I 1 (1972) 189-216.

32. Cf. denpunischen Text, in ArchClassXVI (1964)66-76:w3ntlm • 'lm bbty 3nt km hkkbm 'l ("und die Jahre der Statue der Gottheit in ihrem Tempel (seien soviel)Jahre wie dieseSternehier"). Weitere Bemerkungenin diesemSinn bei G. Colonna, La donazionepyrgensedi Thefarie Velianas,in Arch ClassXVII (1965)

154

L.A.

FORESTI

286-292; idem, op. cit. (Anm. 2) 201-209,219; A.J. Pfiffig, "Wie dieseSternehier" (kin hkkbm 7, LP I, 10f.). Der Sternschleierder Astarte?in Hommagesa M. Renard III (Coll. Latomus 103) (1969) 461-473. 33. E. Gerhard, G. K6rte, Etruskische Spiegel l-V (1884-1897) II 176; J.D. Beazley,The world of the Etruscanmirror, in JHS LXIX (1949 [1950]) 11 if.; Pfiffig, Rel. Etr., Abb.14: Athrpa als gefitigelteG6ttin ist dabei, in Anwesenheit von Turan und Atunis sowiedesPaaresMeliacr und Atlenta einenNagelmit einem Hammer in die Wand einzuschlagen; dasSchicksaldesMeliacr ist somitbesiegeIt. Nortia als Schicksalsg6ttin: Waagnerund Deecke,in W. H. Roscher,,4usfiihrliches Lexikon der griechischenund r6mischenMythologie (1897-1902) III 456 f. s.v. Nortia.

34. M. Pallottino, Etruscan Painting (1952), Taf. 37; M. Moretti, New monuments of Etruscan painting (1970), Abb. 100, 103, 139, 301; idem, Etruskische Malerei in Tarquinia (1974), Abb. 12. 35. Colonna, op. cit. (Anm. 2) 204; Pfiffig, op. cit. (Anm. 32) 465. 36. Juv. Sat. X 74 f.: si Nortia Tuscofavisset;Mart. Cap. 1,88:quam aliisortem asserunt,Nemesimquenonnulli, Tychenquequam plures aut Nortiara. Ftir die Wichtigkeit des Kultes der Fortuna yon Praenestecf. Otto, in RE XIII A (1910), Col. 12 ff. s.v. Fortuna. Der Fortunakult soll yon dem angeblich aus Etrurien stammendenServius Tullius nach Rom eingeftihrt worden sein: Liv. X 46,14: reliquo aere aedem Fortis Fortunae de manubiisfaciendam locavit prope aedem eius deae ab rege Ser. Tullio dedicatam. Die r6mische Fortuna wird yon C.O. Thulin mit den etruskischenVerhtillten G6ttern in Zusammenhanggebracht:Cf. Die G6tter desMartianus Capellaund die Bronzeleberyon Piacenza,in Religionsgesch. Versucheu. Vorarbeiten lll 1 (1906) 37 ff. Dagegen abet Otto, op. cit. 37. Pfiffig. Rel. Err. 363 if. 38. Die Darbietungen yon anlfi131ich einer hartnfickigenSeucheim Jahre 364 v.Chr. nach Rom geholtenetruskischenSchauspielernsollenden g6ttlichen Fluch bannen: Liv. VII 2,1. 39. Liv. VII 3,3 (Anm. 9).

40. Momigliano, op. cit. 280 ff.; Hanell, Das altr6m. epon.,4mr 126;Heurgon, op. cit. 433; Alf61di, R6m. Fr,•hg. 98. 41. So Alf61di, R6m. Friihg. 98. 42. Nach Latte, op. cit. (Anm. 1l) 154, Heurgon, op. cit. 433 und Pfiffig, Rel. Etr. 61 ff. ist die Jahreszfihlungeine rationalisierendeUmdeutung dutch Livius einer Nagelschlagungals magischefRitus; Hanell, Das altr6m. epon. ,4rot 116, Werner, op. cit. 28 f. und Pena, op. cit. 254 sehenim clavus annalis einen apotropfiischen Neujahrsritus,im clavuspiaculariseinerituelleS•ihnemal3nahme. 43. Die Historizitfit der lex wird yon A. Magdelain geleugnet(Praetor maximus et comitiatus maximus, in Jura XX (1969) 257-286). Der Verf. nimmt an, dal3der Bericht•iberdenclavusannalisaufgrundeinersp•terenNagelschlagung magischen Charakters zusammengestelltwurde. Dagegen:Alf/51di,R6m. Frt2hg. 100. 44. Eine Nagelschlagungzu kalendarischenZwecken, allerdings nicht als Jahres-sondernalsS•ikularrechnung, findet sichbereitsbeiTh. Mommsen,op.cit. 176if., Miiller-Deecke,op.cit. (Anm. 24) 307if. undG. Wissowa,op. cit. (Anm. 27) 430.

NAGELSCHLAGUNG

IN

ROM

UND

ETRURIEN

155

45. E. Unger, in Reallex. der Vorgesch. VIII 422 s.v. Nagelurkunde. In literarischen Quellen kommt das Wort pal• m6glicherweise mit der Bedeutung Nagel vor: W. v. Soden, Akkadisches Handw6rterbuch (1958- ) II 817 s.v. palfi(m); A. Boissier,Mat6riaux pour l'•tude de la religionassyro-babylonienne, in Proc. $oc. Bibl. Arch. XXIV (1902) 220 if.

46. Liv. VII 3,8: a consulibusposteaad dictatores,quia maiusimperiumerat, sollemneclavi figendi translatum est. intermissodeinde more digna etiam per se visa res propter quam dictator crearetur. 47. Kuhnert, in RE IV (1901), Col. 2373-2377 s.v. defixio. J.G. Frazer, The Golden Bough (31912) VI, 59 if. mit Beispielenaus der ganzen Welt und mit Erw/ihnung der Nagelschlagung in Rom; A. Wuttke, Handwdirterbuch des DeutschenAberglaubens(1927) u.a. VI, 815, 819 in bezugauf die Festhaltungeines Totem und u.a. 1957, 1293,1337in bezugauf Krankheiten.Kieglingbeschreibtden Brauchdes"Stock im Eisen"in Wien, mit dem man eine Krankheit einnagelte(cf. Die Stephanskirche,der Stock im Eisenund der rote Turm in Wien(1920) 30 if.); R. Lasch, Die Verbleibsorte der Seelen der im Wochenbett Gestorbenen, in Globus LXXX (1901) 111if. mit mehrerenBeispielenausder Insel Seranglao(lndonesien); E. Mangin, Les Mossi, in Anthropos IX (1914) 733: Hier wird erw/ihnt, dag die Mossi (ein Volk der Sudan-Neger) die Augen der H/4uptlingemit N/4geln durchbohren. H. Jungwirth, Die vergleichendeVolkskunde im altsprachlichen Unterricht,in Mitteilungendes VereinesKlassischer Philologenin WienIV (1927) 9 ff.; E. Stemplinger,Arttike und moderne Volksmedizin(1925) 72 f. 48. Frazer, op. cit. III 233 if.; Kuhnert, op. cit., mit zahlreichenBeispielemG. Fogolari,//Museo Nazionale Atestino in Este (31967)29 und 62 mit einem Bericht tiberN/4gelausdem Reitiaheiligtumvon Este;Daremberg-Saglio,op. cit. (Anm. 2);

Pfiffig,Rel. Etr. 363ff.; R. Wtinsch,AntikesZauberger/itaus.Pergamon,in JDAL Erg/inzungsheftVI (1905) 43 f. 49. Frazer, op. cit. (Anm. 47), Index s.v. nails. 50. Das wird zwar yon Liviusin bezugauf Rom nichtexplicitegesagt(cf. Anm. 2), 1/41gt sichaber aufgrundder Stellebei PaulusDiaconusund desVergleichesmit Etrurien

mit h6chster Wahrscheinlichkeit

annehmen.

51. Anderer Meinung ist u.a.A. Momigliano (op. cit. 403), obwohl aus Plinius (Anm. 16) hervorgeht, dag eine Bannung negativer Kr/4fte dort erfolgte, wo die D/4monen weilen. Die Parallelstellen aus der V61kerkunde und Arch/4ologie best/4tigen es:Die D/4monenwerdenin Bgumen(Frazer, op. cit. (Anm. 47) II 36: VI 202), in Gr•ibern (Fogolari, op. cit. (Anm. 48) 62), in Menschen(Lasch, op. cit. (Anm. 47) l 1l), in T•ifelchen(Pfiffig, Rel. Err. 363 if.) fixiert. Weitere Beispielebei Kuhnert, op. cit. (Anm. 47), Col. 2374. 52. H. Ehelolf, Keilschrifturkundenaus Boghazk6i XXIX (1938), Nr. 4, Col. I, 34-36: "Dann nehmcnsie zwei N/qgelaus Bronze und schlagcnsie in desTempeIs Hof-Tor hinein, den einen auf der einen Seite, den anderen auf der anderen Seite." Cf. H. Kronasser, Die Umsiedlungder schwarzenGottheir, in SA WW CCXLI 3 (1963). Cf. auch A.J. Pfiffig, Uni-Hera-Astarte. Studien zu den Goldblechenyon S.

Severa/Pyrgimit etruskischerund punischerlnschrift,in Denkschr.d. Osterr. Akad. d. Wiss.,phiL-hist. Kl. LXXXVIII 2 (1965) 48. 53. Aus Plutarch, Popl. 14 geht hervor, daf5 die lden des September der

1.56

L.A.

FORESTI

Grfindungstag der aedes Capitolina waren: EtSo•g o•v Z•rcz•tlBp{ottg... 54. A. Olrik, Ragnardk. Die Sagenvom Wehuntergang(1922)400 f., 403 f., 424. 55. Diels-Kranz, Die Fragmenteder Vorsokratiker(•1964) I 13 A Anaximenes, 14: 'A. •Xto• 8tx• x•z•z•y•t z& •½zp• z&t xpu•z•XXoztSz•;E. Topitsch, Mythos-Philosophie-Politik.Zur Naturgeschichteder Illusionen (1969) 47. 56. W. Gundel, Sterne und Sternbilder im Glauben des Altertums

und der

Neuzeit (1922) 1 if. 57. M. Pittam La Sardegna nuragica (1977) 181. 58. Frazer, op. cit. (Anm. 47) 66: "their prudent ancestors appear to have determinedthat so salutarya measureshould... regularlydiffuse its benefitsover the communityby anticipatingand, as it were, nipping in the bud evilswhich, left unchecked,might grow to dangerousproportions." 59. Berve, in RE XXVI A (1927), Col. 2040-2059 s.v. lustrum; R.M. Ogilvie. "kustrum condere", in JRS LI (1961) 31-39. Wird die Nagelschlagungmir der ebenfallsvon den Zensorenvollzogenenlustratio verbunden,sobeziehtsiesichauf einen Reinigungsritus:Boehm, in RE zit. Col. 2029-2039. 60. Cf. StellenbeiJuvenalund MartianusCapella(Anm. 36) sowieTac. Ann. IV 1 und Tertull. apoZ 24, ad nat. II 8, welche Nortia mir Volsinii verbinden. Inschriften aus r6mischer Zeit nennen ausdrficklich eine G6ttin Nortia (CIL XI 7287:c[urator t]emplideae•ort]ia[e] ; CIL V1537,4: Nortia, te venerorLari cretus

Vulsiniensi); Weihinschriften for NortiaausVolsinii:CIL XI 2685f. 0her eine m6glicheIdentifizierungeineseinzelligenTempelsauf dem PoggioCasettavon Bolsenamir dem Heiligtum der Nortia, cf. R. Bloch, D•couverte d'un habitat •trusque archaiquesur le territoire volsinien,in MEFR LXVII (1955) 80. 61. Cf. Colonna und Pfiffig, op. cit. (Anm. 32). FOr weftvoile Hinweise und manche Anregungen sei Herrn ProfessorG. Dobesch bestensgedankt.

ADDITIONAL

NOTE

In an articleentitled,'Thucydidesand IIoXt•r•p0•¾y.o(•6¾•l', in Vol. IV 10-22 of thisjournal, I discussed the significance of the termr•oXt•r•p0•¾y.o(•6¾• l in the light of its soleoccurrencein Thucydides(V187.3). I concludedthat the word is not a key political catchword in the fifth century and that its connotationswereso negativeasto precludeits useby either detractorsor proponentsof Athenian policy. I further stated(incorrectly)that the word was found in only one other fifth-century author: Aristophanes, Acharnians833. Through an oversightI failed to includethe occurrencein Pseudo-Xenophon,Athenaion Politeia II 18:2

But anyonewho wishes[to attack] private personsthey [the demos] encourage,knowing that the personso abusedin comedydoesnot generallycomefrom the demosand the majority, but is one marked by wealth, high birth or power.Somefew poor and averagecitizens, however, are mocked, but even they only on account of their

meddlingnatures(x0t[068'o•zot k&¾ y.• 8t&rroXurrp0tyy. o(•6¾•¾) anda desirefor greaterinfluenceandadvantagethan thedemos,sothat the demos is not annoyed either when such people are ridiculed in comedy.

The point of thispassageisthat the demosexertsa definiteinfluenceover comicwriters.2In additionthe authordistinguishes the subjectssuitablefor comicmockery:the demosencourages ridiculeof the individualpossessed of wealth, nobility, or power; thus it avoidscriticismof itself;the average citizenonlybecomesthebutt of thehumoronaccountof r•oXt•r•0•¾y.o(•6¾• which must denote 'officious meddling'. The useof the word here harbors the same sort of sarcasmevident in the passagesof Aristophanesand Thucydides;thus,thisthird occurrenceof the term in the fifth century--to my knowledgethere are no others--reinforcesmy conclusionthat r•p0•¾?.o(•6¾• was neither a widespreadnotion in the fifth centurynor reflectiveof any seriouspolitical position. June W. Allison

The Ohio State University NOTES

1. In checkingthe usageof the fifth- and fourth-centuryauthorsI inadvertently left the occurrencein the Pseudo-Xenophonwith the figures for Xenophon.

157

158

J.W. ALLISON

Xenophon does not usethe noun (seenote 30 of the article). There is no one now who would date the Pseudo-Xenophonlater than the lastdecadeof the fifth century and the middle of the century seems most probable. See G.W. Bowersock, Constitution of the Athenians, in Xenophon, Scripta Minora, Loeb Classical Library (1968) 461-65, and 'Pseudo-Xenophon',HSCP71 (1966) 33-8; and J.M. Moore, Aristotle and Xenophon on democracy and oligarchy (1975) 20, for discussionsof the problem. 2. Bowersockrightly notes(Constitution of the Athenians 496 n. 2) that the passage'has nothingto do with the known banson comedyin 440/39-437/6 or in 415'.

THE

DUBIOUS

ORIGINS

OF THE

'MARATHON'

During the frenziedpreparationsfor the first modernOlympicgames,held at Athensin 1896,Baron Pierrede Coubertinacceptedthe suggestionof his friend, the noted philologistMichel Br•al, that he add a 'marathon race'to the event being planned. Br•al even offered to furnish a silvercup as the winner'strophy. This proposal,called a 'curious,bogusHellenic intrusion' by a recent historian of the first modern games,1may very well have been inspiredby the widespreadpopularity of Robert Browning'spoem 'Pheidippides', published in 1879 in the poet's Dramatic Idylls. The poem describedthe runner'strip to Sparta to ask for aid in resistingthe Persians. Like most literary Englishmenof his day, Browninghad a thorougheducation in the classicsand knew his Herodotus.But healsouseda greatdeal of poetic licensewhen he brought Pheidippidesback to fight at Marathon and then run back to Athens to report the victory: Till in he broke: 'Rejoice, we conquer!' Like wine through clay, Joy in his blood bursting his heart, He died--the

bliss!

A noted writer on running matters has recently claimed that the whole story of Pheidippidesis highly improbable.2 Although his article was met by a storm of indignant lettersto the editor protestingthe application of hochlcritil•to athletic legends,his suspicionswere justified, for the Pheidippidesstory can be shown to be basedon at least two other, different storiesand a manuscriptvariant. Nevertheless,for the historiographerthe evolution of the tale is not without interest,both as a generalexample of how suchthingsget startedand asa specificinstanceof agglutinationand contamination.

The germ of the story is in Herodotus,who describedthe journey of a certain h•merodromos to Sparta to ask for aid (6.105-106). The first problem is the name of the runner. The best class of Herodotus' manu-

scriptshas the reading Pheidippidesand this usedto appearregularlyin older editionsof the historianand in nineteenth-century generalhistories of Greece(e.g., Grote, Curtius,etc.). But the weightof ancienttestimonyis certainlyon the sideof the readingPhilippides.Almost everywriter who mentionedthe run to Sparta seemsto have read Philippidesin his text of Herodotus (although there are a few manuscriptvariants).3 Second,the name Philippidesis frequentlyattestedin Athens, whereasPheidippides

159

160

F.J. FROST

appearsonly as the comic name of Strepsiades'son, addictedto horse

racing, intheClouds (invv.59-6•thename isexplained asa compromise betweenCallipidesand Pheidonides).Finally, the readingPhilippidesis found in four other manuscriptsof Herodotus.Pheidippides,therefore,is probably to be seenas a copyist'serror, althoughone can neverbe certain in these instances. 4

Nor is there any evidencefor a run from Marathon to Athens in the rest of Herodotus' descriptionof the battle and its aftermath. We are told that upon boarding their shipsand receivinga traitor's signal, the Persians

attemptedto doubleSounionandreachAthensbeforethe armycould return (6.115-116). But the troops marchedback 'as fast as possible'and arrived well ahead of the foe.5The silenceof Herodotus doesnot, of course,

prove that no one ran from Marathon back to Athens.On the contrary, plaincommonsensewoulddictatea greatraceof observers, beginningeven before the battle was over, eachman wishingto be the first to report the amazing victory. Unfortunately for the legendsof long distancerunners, someonein one of the many villagesalong the route undoubtedlyjumped on a horse and swiftly outdistancedthose on foot. At any rate, for one interestedin running records, the journey of Philippidesto Spartais by far the greaterachievement.The distanceby any reckoningis somethingmorethan 240 km and PhilippidesreachedSparta the day after he started. This is quite a respectablerun even for a professionalcourier. The Plataeanshad a tradition about an evenmore difficult run from the Persian War era, and one element of the story may have been borrowed and attached to the spurious Marathon tradition much later. After the victory at Plataea, the Greeksweretold by an oracle to extinguishall fires throughoutthe land and to rekindlethem with fire takenfrom the koine hestiaat Delphi. A PlataeanrunnernamedEuchidas ran from Plataea to Delphi and back, all in the sameday, handed over the torch and expired. The distanceon the modern road is 214 km; Plutarch, who tells the story (Aristeides20.4-6), gave the distanceas 1000 stades, which converted to the Delphic stadion of 177.55 meters adds up to a shorterdistanceof about 178 km, or somethingover a hundred miles--a staggeringachievement,consideringthat the outward leg is all uphill? Plutarch sayshe saw an epitaph to Euchidasin the sanctuaryof Artemis Eucleia at Plataea, where the runner was buried:

Euchidas running to Pytho came back here the same day. There is no reason to doubt that this is what Plutarch himself saw;? the

questionof authenticityrestson the integrityof the Plataeansof the Pax Rornana:hadtheycarefullypreserved a monumentof PersianWar date•or

ORIGINS

OF THE

'MARATHON'

161

had they cooked up a batch of antiquities based on highly embroidered traditionsin orderto draw tourists? 8If the storyof Euchidaswasearly,and well known, it may haveinspiredthe creationof similarstoriesabout other notable

battles.

The first extant evidencefor sucha storyabout Marathon is the report of HeracleidesPonticus07. 340 BC),who claimedthat the newsof the victory had beenbrought to Athens by one Thersippusof Erchia.9 In the passage where Heracleidesis cited, however, Plutarch goes on to say: ß. . but most writers saythat it was Eucleswho ran in his armor, hot from the battle, and burstingin through the doorsof the Po'taneis(?) said only, 'rejoice' and, 'we have won(?)' and then immediately expired.•0 The identity of the writers mentionedby Plutarch is lost, unfortunately; we know only that thistradition existedalreadyby mid-fourth century.We may further speculatethat the tale was createdsimply to give the Plataean runner Euchidas a Marathonian parallel. For all we know, every great battle in antiquity eventuallyattractedanecdotalembroiderylike this, with a runner arriving at the gatesof the victoriouscity, gaspingout the good newsand breathinghis last. Even Xanthippus'dog was supposedto have expired after swimming behind the boat taking his master from Athens to Salamis. •

Later writers on marvelous exploits were more interestedin absolute distance.In a sectionon great feats of running, the elder Pliny (NH 7.84) catalogued the run of Philippides to Sparta, that of two of Alexander's couriersfrom Sicyonto Elis (241 km) in one day, and eventhe recordof a Roman boy of eight, who ran 75,000 passus (c. 120 km) 'meridie ad vesperam'. But he either had not heard of the Marathon run or saw no reason to include such a relatively minor stroll. The evolution of the legendis completewith Lucian, in the late second century AD, who is the first extant writer to take the variously named runner from Marathon, give him the name of Herodotus' h•merodromos Philippides,and make him expire after giving his glad message: Philippidesthe h•merodromos,reportingthe victoryfrom Marathon to the archons,who were seatedanxiously awaiting the resultof the battle, said, 'rejoice, we have won', and sayingthis, died at the same time as his report, expiring with the salutation. (Pro lapsu inter salutandum 3) Whether this was the popular tradition by Lucian'sday, or whetherthe

162

F.J. FROST

writer, in his usual manner, was making fun of the communisopinio by gettingit slightlywrong,we do not know. But we can sumup the developmentof the traditionin a chronologicaltableandthusseewith a simplified perspectivehow the tradition finally reachedthe nineteenthcentury. 490 BC 479 c. 424 c. 340

by 1st cent. BC C. AD 180

the battle of Marathon.

the battle of Plataea and Euchidas' run (?). Herodotus'history mentionsPhilippides'run to Sparta. Heracleidesnames Thersippus:other (perhaps later) writers say it was Eucles who ran from Marathon and died after giving his message.

the variant Pheidippidesappearsin editionsof Herodotus (thus Nepos: Phydippumque). Lucian says it was Philippideswho ran from Marathon

1879 1896

and died.

Robert Browningmixesthe wholelot togetherin his 'Pheidippides'. Spyros Louis wins the first modern marathon with a time of 2:58:50.

There is thusabsolutelyno historicalsupportfora marathonrace.But it might be addedthat thiswhollymoderntwenty-sixmile run (nowadaysrun by hundredsat a rate of onemileeveryfive minutes)needsno Pheidippides; it is quite enoughof an achievementto stand by itself. Frank

University of California Santa

J. Frost

Barbara

NOTES

1. Richard D. Mandell, Thefirst modern Olympics(Berkeley and Los Angeles 1976) 89; seeV.J. Matthews, "The Hemerodromoi", CW68 (Nov. 1974) 161-169.

Hugh M. Lee has kindly let me seehis manuscript,"Modern longdistancetheory and Philippides'run from Athensto Sparta", which washeardat the APA annual meeting, Vancouver 1978. 2. JamesFixx, Sports Illustrated, Dec. 25, 1978, 60-66. 3. Pliny, NH 7.84; Plutarch, De Hdt. mal. 862AB; Pausanias1.28.4, 8.54.6: Clem. Alex. Protr. 3.44.3; Lucian, Pro lapsu 3; Pollux 3.148; Suda s.v. Hippias. Nepos, Milt. 4, however, has Phydippumque(sic), showingthat the variant 'Pheidippides'had appearedsometimebefore the later first century BC. 4. See the discussionby R. Renehan, Greek textual criticism (Cambridge, Mass. 1969)68 f. Renehancautiouslyprefersthe readingPhilippides,becausethe

THE

NAME

OF THE

RUNNER

163

other writers cite it and becauseit is a common Attic name. Pheidippidesas the name of a real personcan be found only twice, on decreesfrom Thera and Eretria (cited by K.J. Dover, Aristophanes, Clouds xxv-xxvi). 5. As conclusivelyshown by A.T. Hodge, JHS 95 (1975) 169-71. 6. But by no meansincredible:seethe long distancerunning recordsin The Guinnessbook of records(1979 ed.) 659 if. 7. An abbreviated version of the story in schol. Aristides, Panath. Ill 200 should not be regarded as independentconfirmation of the story, as it is based either on the tradition followed by Plutarch or on Plutarch himself. 8. The tradition could go back to the third century •c when the citizensof the recentlyrebuilt town boastedof the battle to all travelers:HeracleidesCreticus 1.11, ed. F. Pfister (Wien 1951): see my comment in C&M 22 (1961) 189. 9. Plut. De glor. Ath. 347C; Heracleides, fr. 156 (F. Wehrli, Schule des AristotelesVII). The meaninglessEroeusof the MSS wasemendedto the demotic Eroieus by Xylander, to Erchius by Wilamowit7. 10. The text is quite corrupt. Prytaneisisa conjecturefor rrpd)vto¾ and thewords of the dying runnerare usuallyemendedfrom 7,0•{?g-rg...7,0d.?o•.gv to 7,0d.?gw... vtx•)•.gvon the basisof Lucian, Pro lapsu 3 (q.v., infra). 11. Philochorus and an unknown work of Aristotle, cited in Aelian, NA 12.35

(FGrH 328 F 116), report more than one dog. Plutarch, Them. 10.10saidthe tomb of the dog was preserved in his day and was known as the "dog-monument"-kynoss•ma.

THE

NAME

OF THE

RUNNER

A Summary of the Evidence

ProfessorFrosthaskindlygivenmepermission to setouta summaryof the evidence, as I see it, on the runner's name. Since I do not share his

preference for'Philippides',anddid notsucceed in persuading himeitherto abandonit or to setout theevidencetully enoughfor the readertojudge,I gladly avail myselfof that permission.My summarywill be brief, but, togetherwith ProfessorFrost'snotes,shouldenableanyoneinterestedto pursue the matter.

The main referencesare cited in ProfessorFrost's Notes 3 and 4 (with

text): we are agreedon the factsanddisagreemerelyon howoneoughtto look at them.The bestclassof HerodotusMSS (a; alsoParisinus)haswhat I shallcallthed form;theinferiorclass(d) hasthel form. We do notagain come acrossthe name in any author for four centuries.The next time we

meetit isin Latin,in Nepos'biography of Miltiades(4,3).Heclearlygotthe actual name wrong:yet the divisionin his manuscriptsparallels,indeed exaggerates, that in Herodotus:the best codicesread Phydippumor Phidippum:onlywhatonemightcalldeterrimihavetheI form:they(and

164

E. BADIAN

consequentlyit) will not evenbe found in Marshall's recentTeubner text. It

thereforedoesnot seema fair summaryto saythat 'the variant "Pheidippides" had appeared sometimebefore the later first century BC'(Frost, Note 3). The fact is that, to judge by the manuscriptevidence,the only two writers beforeour era who giveusthe name, in different languagesand four centuriesapart, both seemto have used the d form. Whether Neposread Herodotusis uncertain:he doesnot cite him, and it seemsmore likely than not that he did not read eventhe Classicalauthors whom he does cite, but got his material from Hellenistic intermediaries. This is generallyacceptedby scholars.•But if so, it would follow that the author who transmitted the story to him also usedthe d form, i.e., that he read it in Herodotus.(IfNepos got it out of Herodotusfor himself,it would even follow that it was still the prevailingform in Herodotustexts in the first century BC;for Nepos cannot lightly be creditedwith criticaljudgment. But that is perhaps unlikely anyway.) The first author (in chronologicalsequence)who, on the basisof our manuscripttradition, can be said to haveactuallyusedthe l form comesa centurylater still: it is the Elder Pliny. What his sourcewas,we cannottell: it could even be a bad text of Nepos himself,or somewriter of the age of Augustus.From that point, in our tradition, thereis no turning back.The d form does not appear again: not even in the learned Plutarch who, among all the authors who tell the story, is perhaps the only one who may confidentlybe assumedto have read Herodotus himself.By about AD 100, it was evidentlyno longer in the current texts of Herodotus. That confusionbetweenA and A is palaeographicallytrivial is obvious and need not be stressed.Nor can one say,apriori, that it is more likely to go one way than the other. We mustarguefrom the factsof the manuscript tradition and from generalprobability• asthey apply to eachcase.We have already looked at the facts of the tradition: on the manuscriptevidence alone (I repeat), no author writing before the first century AD can be proved, or even reasonablyassumed,to haveusedthe l form. What about generalprobability? Professor Frost points out (as others have) the rarity of the name 'Pheidippides'.Apart from this instance,it appearsin the Attic recordonly in Aristophanes,whereit seemsto be usedfor a particularpunningeffect.2 What do we make of this?The fact that Aristophanesmade a pun on a name is neither here nor there. He does it dozens of times with real names;

on the other hand, he will readily invent a name (like Philocleon or

Bdelycleon)if it suitshispurpose) The uniqueness of the namein itselfis more important. It follows that, accordingto the old and very reasonable principle of the difficilior lectio, we ought to prefer it. Given the easy corruption of one of thesetwo lettersinto the other, it would be very easy

THE

NAME

OF THE

RUNNER

165

for a unique name to be miscopiedas a common one; but why should

anyonedo the opposite? The'copyist's error'assumed byProfessor Frostis exceedinglydifficult to explain, if it went the way he wouldwant it to: the copyist would have to be a literary man, who for some reason was reminded,while copyingHerodotus'story of the h•rnerodrornos, of the characterin Aristophanes'Clouds.Nothing shouldbejudgedimpossible, in copyistsof manuscripts;but this is certainlya little far-fetched.The principleof the difficilior leerio,in this instance,is difficult to counter. Should we be worried by the rarity of the name as such?Is it likely that there were namesin Athens at the time that, to our knowledge,are unique in the Attic record?The answer must be: yes, very likely indeed. A really surprisingnumberof earlynamessimplygo out of fashion,we do not know why. I have not checkeda recent index of Attic names on this, but Kirchner'sProsopographiaAttica may still be said to givea good general idea of the state of Attic nomenclature, even if the detailed statisticshave

changedin manycases.A goodway to geta feelfor the fate of upper-class nameswill be to look at the archon-listof the period. Between511 and 491 BCwe know the names of 14 archons. Of thesenames,no fewer thanfour

are unique:Harpactides,Hermocreon,Hybrilidesand Smyrosneveragain occur in Kirchner's collection. In addition, one or two othersget pretty closeto it.4 For reasonsthat we cannot fully understand(though they are no doubt closelyconnectedwith the immensesocialand political changes that came over Athens in the fifth century), it seemsto be far from unusual--in fact, quite the opposite--for names known before or even slightly after 500 to disappearfrom the record. Unique as an Athenian name, 'Pheidippides'(if that was the man'sname) would by no meansbe unique in its history. These are the principal reasonswhy, despite'the weight of ancient testimony' (which, in the aggregate, cannot be denied), I must part company with Professor Frost on this one point in his delightful and instructive

article. 5 E. Badian Editor

NOTES

1. See, e.g., Schanz-Hosius4 I 359 and, more recently, Edna M. Jenkinsonin T.A. Dorey (ed.), Latin biography(1967) 6. The importanceof Nepos'evidenceis sometimes overlookedby Greekpalaeographers. Failureto appreciateit flaws,e.g., the usefuldiscussion in R. Renehan,Greektextual criticism(1968) 68 f. 2. Scholarsbeingas they are, it couldhardlyfail to happenthat this shouldbe

166

THE

NAME

OF THE

RUNNER

suggestedfor the Herodotean characteras well: he 'sparedhorses'becausehe could run! This atrocious idea seemsto be due to Salmasius, to whom I found it ascribed in the Bipontine edition of Lucian (III 586). 1 did not look any further. 3. Renehan (loc. cit.) rightly stressesthat Aristophaneswould not be above making fun of a well-known character connectedwith Marathon. 4. Acestorides:two archons(504/3 and 474/3) and a kalosinscriptionon a redfigure vase, probably referring to the younger of them. Lysagoras:an archon (509/8), then only one more entry, a name on a fourthcentury tombstone. Scamandrius:an archon (510/09), then nothing until around 100 Bc• when we get the name twice (in 107/6 and in 97/6 Bc). Out of thesefourteen names,therefore,four are neverfound again and three more are (to say the least) very rare• one of them unknown after 474/3. 5. I should add that James Fixx, in the article cited by Professor Frost, had obviously profited by a telephoneconversationwith me, but used my name contrary to my expressedwish and (inevitably) misrepresented what he had been told.

REVIEW-DISCUSSION

Ronald Syme,Roman Papers,editedby E. Badian.2 volumes,Oxford 1979.

Severalyearsof hardwork werenecessary beforethelong-awaited collection of the Roman Papersof Sir Ronald Symecouldbe published-unfortunately, but unavoidably, withoutan index,but with helpfulbibliographical addenda. I The two handyvolumes, carefullyeditedby oneof the mostprominentpupilsof the author, mark an importanteventin contemporary research in Romanhistory:fifty-ninepapersof anoutstand-

inghistorian, including somebasicreviews, uptothepresent dispersed ina longseries ofvariousperiodicals, monographs andotherpublications, now appeartogether--onthe one hand,by their homogeneityof subjectand method,servingasan exampleof concentrated scholarship, on the other handdemonstrating the path followedby a greatscholarin the courseof forty years,from hisearliestincludedpaper,'The imperialfinancesunder Domitian, Nerva and Trajan', originallypublishedin 1930,down to the latestreprintedstudy,'The conquestof north-westSpain',whichfirst appearedin 1970.Nevertheless, the two volumesnecessarily constitute onlya torsoof Syme'swork;theycanonlybeusedto full advantage, and valuedin anadequate manner,together withhisvariousfurtherpapers, in part re-editedin othercollections of hisessays, and, of course,together with his books. 2

The subjectof Roman Papersis, with few exceptions,the sameasthat of thewholework of our author:it is Romanhistory.3Or, to beprecise,it is Roman historyfrom the Late Republicdownto the turn of the fourth and fifth centuries AD; andif we wantto be evenmorerigorous,it is Roman historywith specialattentionto the periodfrom Caesarto Hadrian or AntoninusPius,withimportantbackwardlooksasfar astheageof Marius

andSullaor earlier,andwith morethana glancethroughlaterperiods, down to the age in which the Historia Augustawas written. But such

precision wouldprobablybelongtothatkindofrigourwhich,according to our author, is a disease. 4 The mostimportantpoint in the definitionof Syme'scontributionto Romanhistoryor evento historyas a wholeis certainlynot thequestion of chronological limits;it concerns hisconception of history.

167

168

G. ALFC)LDY

This conceptioncan, I think, be clearly defined on the basisof a wellknown passagein The Roman revolution:'In all ages,whateverthe form and name of government, be it monarchy, republic, or democracy, an oligarchy lurks behind the faqade; and Roman history, Republican or Imperial, is the historyof the governingclass.'5Thus, the main subjectof Syme'swork, briefly expressed,is preciselyhis famous'lurking oligarchy': the socialand political conditionsfor the power of aristocraticgroups;and at the sametime the studyof how this powerwasexercisedby them,and of the mentalityand the behaviourof their members,asrevealedbythe fate of individual representativesof the aristocracy,and by their noblestintellectual occupation--literature. It is, therefore, not possibleto simplify Syme's approach to history, in fashionableantithetic terms, either as 'structuralhistory' or as a 'history of events'.This kind of history is at one and the sametime an analysisof both socialand political structures,and of eventsand individual vicissitudes; in its complexityand in its limitationsit is somethingabsolutely original and inimitable. II

The main patternsof this kind of history,the cardinaltopicsof Syme's historical research,are clear and on the whole surprisinglysimple. First, 'MissingPersons'.This phraseis borrowedfrom the commontitle of three papers6 supplyingaddendaand corrigendato the prosopography of Roman senatorsand knightstreated in somevolumesof the Realencyclopiidieder classischen Altertumswissenschaft, and treatedin an unsatisfactorymanner,althoughfor a betterresult'modestyand accuracywould have been enough? I would like to use this phrase to indicate Syme's efforts and achievementsin filling gapsin the prosopographyof the Late Republicand of the Empire.They wouldsufficeto givehim a rank among top scholars.There is an immensenumberof Roman aristocrats--magistrates, provincial governors,other functionaries,or simply personsof rank--whom we know solely, or at least better than before, thanks to Syme'saccuracyin collectingsources,correctingnames,completingfragmentary evidence,putting together manifold data, and dating careers, tenuresof office, events.The caseof'the wrong Marcius Turbo', a modern conflationsplit up by Symeinto two real persons,the Prefectof the Guard Q. Marcius Turbo and the procurator T. Flavius Gallonius Fronto Q. MarciusTurboS--is only oneof innumerableexamplesof the resultsof this prosopographicalGrundlagenforschung, while the long list of the condemned'bogusnames'in the Historia Augusta9 liberatesthe historyof the Roman Empire from an unedifying company of wraiths invented by a 'genialimpostor'.•0And if theauthorputsthequestion,'Who wasDecidius

RONAED

SYME, ROMAN

PAPERS

169

Saxa?' or 'Who was Vedius Pollio?', • his answer does not contain mere arid biographical details but reveals--without over-interpretation of the sources,merely with 'modesty and accuracy'--full-blooded men of character. Such is Domitius Corbulo, who 'had acquiredglory and prestigetoo great for a senatorand a subject';•2 or M. Aemilius kepidus, the capax imperil: 'Birth, the dynastic connection, and virtus, such was Marcus kepidus.'•3 Second, the origo of Roman aristocrats. 'kocal origins or alliances sometimesfurnish a clue'14--for the explanation of social background, education, mentality, interests, political aims or career. Nobody has emphasizedthis branchof prosopographicalresearchin a more consistent and productive manner than Syme. His article entitled 'Senators, tribes and towns' demonstratesthe possibilitiesof detecting the origo of distinguishedmen, on the basis of nomenclatureand tribe, in a long seriesof individual cases,treated in an exemplary way;•5and his studyconcerning the origin of Domitius Corbulo, from Peltuinum in the Abruzzi, basedon the splendidinterpretation of a misunderstoodinscriptionfrom Bominaco and on a combination of further epigraphicand onomasticevidence,offers a masterpiecenot only of method but of intuition. •6Of course,researchof this kind, includingother techniquesas well, suchas, in severalcases,the attention paid to the specialinterest of an author in a region or a town, cannot arrive at more than a hypothesis.But even Syme's hypothetical suggestionsfor the patria of senatorsor authors lack neither method nor intuition--as in the caseof Tacitus, who, accordingto Syme, may have been a Narbonensian,•7 or of Juvenal, who may have been from Africa, perhaps from Thabraca.•8 But that is not all. What is most important in Syme's endeavoursto reveal the patria of Roman aristocrats is his interest in the established regionalgroupsof senatorsfrom the peripheryof the Empire, in 'colonial •lites'--such assenatorsfrom theTranspadana(whichaccordingto him, to turn a phraseof the Elder Pliny on Narbonensis,isprovincia veriusquam Italia), from SouthernGaul, perhapsmost of all from Spain,but alsofrom other partsof the Roman Empire.19'The strengthand vitality of an empire is frequentlydue to the newaristocracyfrom the periphery'? and (e.g.) the 'resplendentsuccess'of the Spanishsenators'provesthem eager,ambitious and innovatory'.21To be sure,'It is personalquality that counts,not raceor origin.'22 But Syme points out clearly, with a deep feeling both for sociologyand for psychology, how much the development of personal quality could depend on milieu, group spirit and a 'lobby', on the other hand on the eagernessand ambition of the rising'new man'. In the Roman aristocracyit was alwaysthe dynamictype of the homo novus--under the Caesarsabove all the 'new man' from the risingupperclassesof'provincial'

170

G. ALFOLDY

Northern Italy and of some provinces--who devoted himselfwith extreme vigour and engagementto the governmentof the Empire and at the same time to the preservationof the mos maiorum. Even if all this, like so much in Syme'swork, may to someextent be basedon personalexperience(and why not?), it is none the lessfundamentalfor the understandingof Roman history. Third, inseparablefrom the studyof originsand localgroups,socialand political connexions.For the composition,regenerationand activity of the Roman aristocracythey were of major importance. They often clearly resultedfrom community of patria, as (e.g.) in the caseof the powerful Spanisharistocratsat the end of the first centuryAD.23But commonorigin was by no meansa guaranteefor common political viewsand interests--as revealed,not least, preciselyby the fate of someSpanish senatorsin that very period.24The Roman aristocracy was an imperial one, with broad cross-connexionsbetweendifferent personsand local groups on the basis of family ties, marriages, adoptions, common study, common service during the postsof the cursushonorum, friendship, patronage,etc. The

documentationand evaluationof socialand politicalrelationsof this kind for the history of the Republic, already practised with successby F. Miinzer and others,is oneof the centralthemesin Syme'swork from the outset, but above all since The Roman revolution. It is his particular merit to have pointed out the importance of these relations for social and political history under the changed circumstancesof the Empire. As to family ties:'Familiesin their riseand duration are a themethat cannotfail to charm and detain.' 'Their vicissitudescan also be exploitedfor political history.'25What stemmatamay presentfor history is demonstrated(e.g.) by Syme'spaperson Antonine relatives,on the Sentii Saturnini, or on the Ummidii.:6 The role of friendshipand patronagewill be clear from the influence of somefriends of the Caesars,from Ovid's contacts, or from the endeavours of the Younger Pliny to support his lesssuccessfulfriends? Fourth, fasti. Lists of Roman officials, above all the fasti of consulsand

provincialgovernors,serveas a frameworkfor political historyand offer cluesfor social history too. They illuminate not only the structureand developmentof the Roman Imperial administration,but alsothe chronological settingand the historicalcontext of political eventsand of a good deal of the policyof the Caesars.At the sametime, the studyof the fasti-together with that of the senatorialand equestriancursushonorum-revealsthe changesin the compositionof the Roman 'governingclass'and demonstratesalike the privilegesof birth and rank and the eagernessand trustworthinessof the hominesnovi in the serviceof emperorsand Empire. This fresh and dynamicconsiderationof Roman fasti from the historical point of view,insteadof thecompositionof merechronologicallistswith at

RONALD

SYME, ROMAN

PAPERS

171

bestsomeresultsfor the study of imperial administration,which wasusual before, is one of Syme'smost important contributionsto the investigation of Roman history. His Roman Paperscontain a remarkable setof studies of different fasti and posts,above all his important review of A. Degrassi's Fasti consolari dell' Impero romano, various articles on provincial governors, and also the fundamental analysis of consulatesheld by imperial legatesof praetorian provincesin absentia.28But his entire work presents considerably more of this kind of research, thus his studiesof details of consularfasti, of governorsof variousprovinces,as (e.g.) Moesia, Dacia, Pannonia, Dalmatia and (recently)onceagainAfrica proconsularis29--not to forget the abundance of his efforts to establishdetailed consular and provincialfasti in The Roman revolution and in the monumental Tacitus.Dø Fifth, power, and strugglefor power--politics, conflicts,wars. From all that has been said, it is obvious that in Syme'sview power in Rome was always wielded by an oligarchy,at first split up into independentfactions, later controlled and also representedby the emperor. Therefore, just as his Tacitus 'did not want his Roman annalsto degenerateinto a sequenceof imperial biographies',31 so our author describesthe history of Roman politics during the Empire by no means only from the point of view of imperial power. Like Tacitus, Syme takesas a starting-pointfor political history'the relationsbetweenthe Caesarsand the senatorialaristocracy'? He has, of course, a sure sensefor the power and policy of Roman emperors,as it wasrepresented,e.g., by the foreign policy of Augustus?-•or as it was expressedin a sophisticatedmanner by the nomenclature of the first princeps, 'Imperator Caesar'.34 He has also given us fascinating portraitsof emperors:of Augustus,a man who'eludesgraspanddefinition' and remindsone of the sphinx on his signet-ring?5 or of Hadrian the IntellectuaP6--who, as such, may among all Roman emperorsbe one of the figuresfor whom Symefeelsthe greatestsympathy.But as men of power and men who have a historical role there appear again and again magistrates, governors, and above all army commanders, suchas Mucianus, the generalof Vespasian? or the famousgovernor of Syria in AD 97 who gave rise to far-reaching political rumours?8 and the clash--or the balance-betweenemperorsand 6lite is always one of Syme'scentral themes.Wars and other problems of military history have frequently attracted his interest,thusin particularthe wars on the northernfrontiersof the Empire and in Spain (treated not leaston the basisof a splendidknowledgeof the geographyof those areas)? But among the main themesof this military historythereisalsothe prosopographyof armycommandersand legionary legates,aswell asthe importanceof warsand troop movementsfor changes in the power structuresof the 'governingclass'.4ø Finally, somethingquitedifferentfrom politicsand wars,but, like them,

172

G. ALFOLDY

extremelyrevealingfor the Roman aristocracy:literature. Syme, himselfa brilliant author, likes literature (prose more than poetry, it seems),and in almost all his works he devotesspecialattention to the literary sourcesof his history. That, of course,means, first of all, the historiographyof the Late Republic and of the Empire--Sallust, Tacitus, and the Historia Augusta. But there are alsoauthorsoutsidethe group of historians,suchas Ovid, Seneca the Elder and the Younger Pliny;41furthermore, historians outside the circle of Roman authors, such as Thucydides or Edward Gibbon,4: who havereceiveda vivid profile from Syme'spen. His interests in literature, asdemonstratedmainly by hisstudieson Sallust,Tacitus,and the Historia Augusta, include a long series of themes which normally belongto the central concernsof philologicalresearch:the sourcesusedby the author, his erudition, his reasonsfor his choice of subject,his techniques in describingthe topics selected.his talent in composition, his languageand style, and, last but not least, his humour. 'That is not all. Literature and history converge.,43The reason for this convergencecan be establishedon the one hand by the themesof the literary sources,on the other by the personalityof the authors.Both, for Syme,reflectthe samething,namelyhistoryas seenor realisedby an dlite. Thus, literature is of interest,in his view, as the most important part of historicaltradition. Therefore, the philologicalstudy of ancientliterature in Syme'swork is never pursuedfor its own sake, but--and in this it is treatedlike epigraphyor prosopography--itis a part of historicalmethod. As to the subjectsof literature: if relevantfor Syme,literature is not only the sequenceof reported facts and events,but also a report on ideasand behaviour.Ovid'slovepoemsare a matter of importanceto our author, not merelyfor their referenceto politicsand wars, but as an expressionof the moral climate and mentality of the Augustanaristocracy,and becauseof the fate of the poet who did not take up the opportunityof a political career.44Whether it is the speeches in the Annals of Tacitusor wit, humour and forgery in the Historia Augusta--even when he is dealing with the compositionof the speeches or with the techniqueof forgery, they representfor him, in thelastanalysis,historicalmatters.The speechof Claudius, arguingfor the extensionof the iushonorurnto the aristocracyof Gaul, in theAnnals,i.e. in Tacitus'version,possesses a morecoherentandtherefore more convincingarchitecturethan in the originalof theemperor,known in good part from the famous tabula of Lyon. But the whole matter is ultimately of interestbecauseof the idea there expressedconcerningthe need to revitalise the Roman aristocracy with hornines novi from the

provinces,an ideadoubtlessbetterarguedby Tacitusthan by hisimperial predecessor. 45And if we learnabout the author of the Historia Augusta, from a long seriesof detectedfrauds, that 'the fellow is alert as well as

RONALD

SYME,

ROMAN

PAPERS

173

carelessand cynical',46thenthissituateshim preciselyasa contemporaryof 'an age of dishonesty'. 47 Nevertheless,the age in which the Historia Augusta was written was 'also an age of fanaticism and murder. Novel therefore and welcome that the voice of a humorist

should be heard on the

side of frivolity and tolerance while others were demolishingtemplesand burning synagogues,or inditing treatisesin justification.TM As already indicated by the fate of Ovid, the main point where history and literature convergeis the biographyof authorsand, first of all, that of historians. 'History among the Romans took its origin in the governing class,and it tended for a long time to be the monopoly of senators.They wrote after experienceof affairs, somein retreator in revulsion,frustrated, angry, and censorious. '49The making of the historian:that is oneof Syme's favourite interests. It is clearly expressed,for example, in his study 'How Tacitus came to history'? But it is demonstrated also in a set of other studies,whose subjectis not necessarilythe making of the Roman historian. His paper entitled 'Thucydides'could equally well have beenentitled 'How Thucydides came to history'--compare his study, 'How Gibbon came to history'.S•These three articles, very similar to one another, also offer an abundanceof autobiography.This will be clear from reading the answersto the questionof what qualities the historian needs.'First of all, erudition.'S2Then, as both Gibbon and Syme say,'diligenceand accuracy'. 'But to write history,you needsomethingmore:the thing mustbe readable, it must have a certain unity of style and design;'it must have 'balance, volume, antithesis'. However, 'accuracy and style are not quite enough; surely one needsstructureand architecture.'S3Nevertheless,beyondthese literary virtues the historian also needsexperienceof the matterswhich he has to describe. Familiarity with politics, and with human nature as a whole, can resultfrom exile;this 'may be the making of an historian'.s4So it was in the case of Thucydides or Gibbon, but not in that of Tacitus or Syme. 'Foreign travel', therefore,S5if there is no coercionto get to know foreign countriesand peoplesby exile, is recommendedby our author-himselfa well-knowntraveller, but alsoa man with somedirectexperience of political affairs. The Roman senatoras a historian, of course,at this point had an enormousadvantage:men like Sallust or Tacitus, by magistracies,missionsand governorships,had the occasionboth for travel and for the understandingof political vicissitudes,and often also for that of military problemsand wars.Thus, historyin Rome was'a prolongationof public life'.56 To sum up. Prosopography,politics and literature, developed in a fascinating combination into a coherent social and political history and completedby a historyof ideasand a historyof the historiographyitself-and all this usedto revealthe role of the •lite in the Roman Empire:thus

174

G. ALFOLDY

may we characterizethe main subjectof Syme'swriting, as shownboth in Roman Papersand in his entire work. A phraseon Edward Gibbon comes to the reviewer'smind: 'He was fortunate in discoveringa historical theme of high import that was congenialto his tastesand not beyondhis talent. III

So far, so good. But some questionsarise. They concern above all the breadthand limits of the subjectthat hasbeendescribed,and the methodof history. And then, the personality of the author. To beginwith the method.I think that, for the subjectchosen,Syme's method, with its emphasison prosopographicaland philologicalfoundations, could not be improved upon. This may be so clear that it doesnot need further explanation. None the less,I would like to deal with some aspectsof this methodwhich are not self-evident,as we often seein the work of other scholarswho also deal with Roman history on the basisof prosopographyand literature. First of all, this is a method without a 'methodology'. I mean by this the talent of our author in developinga complex method with the help of the sourcesand the facts alone, without long discussionsabout the method itself, and without a 'theory' from which the method must be deduced.'To insiston "die gesundeMethode"istedious.And it maybe superfluous.Nor is the cult of methodologyalwaysa signof strength.'58There is no paperof our author devotedto prosopographyor philologyas method. In hiseyes, methodis in the first place,ashesaysof Thucydides,simply'hard work and accuracy'? i.e. consistentanalysis of sourcesand facts. Furthermore, 'theory',whichis so popularin our time, is not neededfor Syme'swriting. 'It is a question,how far a composerof historyhasto bea powerful"mind" or an original thinker. Too much of thought might sometimesbe a detrimental infusion', as in the caseof a work 'overloaded... with many

philosophical disquisitions'. © Thisisan attitudewhichlargelycorresponds to that of the Romans, who 'had an extreme distrust of abstract specula-

tion, especiallyif it touchedstateand society'. 61Syme,of course,like his predecessors--such as Thucydides, Sallust,Tacitusor Gibbon--isfull of ideas.'History is not the mere collectingof facts;the expositionmustbe built up on someleadingidea, or indeedon several,and be interpretedin their light.'62But this is not the sameas abstract'theory',deducedfrom philosophy,sociology,economics,or whatever.History itself is good enough.In Syme'sview,historyisrationalandintelligible.Or better,it can be maderationaland intelligible--andexactlythisis 'method',asexemplified by Thucydides. 63What one needsfor this, accordingto Syme,is nothingbut the qualitiesthat keepcomingup, andareexemplified,in his own writing:education,hard work and accuracy,compositionand style,

RONALD

SYME, ROMAN

PAPERS

175

familiarity with politics and human nature. All that, together, may be summed up as intelligence. 'Intelligence is the supreme virtue. TM From this it follows that historycan be written without beingunnecessarily complicated:briefly and simply, as in fact Syme doeswrite. This is not only a matter of compositionand style,but first of all a consequence of the conviction that history is rational and therefore explicableby itself. Thus, 'the historian reprobatesthe practice of discoursingat inordinate length among those who understand.TM E.g., on the Spanish wars of Augustus,or on the political ideasof Tacitus, somescholarshave written books;for Syme, relatively short paperssufficed.In thesethe readerfinds all that is really necessaryand important enough to be worth saying.66 In the same way, Syme's manner of argumentation is brief and clear. Arguing for an account, or for a hypothesis,he has no need for monumental constructions.One argument or one proof can be sufficient. 'As with illegitimate infants, enoughfor conviction.'67See, for example, the discussionof the famous Titulus Tiburtinus,an inscriptionwhich hasgiven occasionto so many discussions on the birth-date of JesusChrist: Syme's statementis attractiveas well as simple,basedreallyonan approachwhich is 'direct, without prepossessions, without the invoking of dubious or deviousargumentation'. 68Of course,if one usesprosopographicalarguments, even very minor details ought to be taken into consideration,for they may be extremely helpful. He who doesnot want to appreciatethis kind of researchreceivesa rebuff. 'Some deprecate.For various reasons. Among them (one surmises),distastefor erudition on a narrow front, to the neglectof broad aspectsand the "higher things".Which may cheerfullybe conceded. One useswhat one has, and there is work to be done. If there is a

placefor censure,it is bettervisitedupon the ignorantand incompetent.,69 Even discussionsof a hypotheticalcharacter,if they are well founded, representindispensableelementsin our effortsat investigation.'In matters of literary and historical appraisement,one cannot operate with the methods of a laboratory or furnish the proof to be demandedin a court of law. The bestis only the probable.Any who raisecomplainthavean easy remedy: to offer something better, somethingcoherent and constructive.,70 Let me statethat evenwhen someof Syme'shypotheticalaccountshave not convincedall scholars,they havealwaysbeenstimulating.Thus, his study

'Not Marius Maximus', arguingfor an 'Ignotus,the good biographer'as the main sourcefor the Hauptvitenin the HistoriaAugusta,hasbeenmuch discussed,but it is undoubtedlyan important step forward, both in the classificationand delimitationof the 'early Vitae' and in the estimationof Marius Maximus asan author who, at the sametime, 'musthaveconveyed much valuable information'

and 'was none the less an arrant scandal-

monger'.71

Clarity of argumentation and simplicity in explaining history in a

176

G. ALFOLDY

rational manner do not mean, however,that the literary techniqueof the historian should not be a matter of importance. On the contrary. Literary finesseand taste, compositionand above all style,are for Syme a part of method--as they were for hispredecessors. Tacitus'hasa supremegift for arrangementand architectonics.He seesthefield in front ofhirn and knows how to prepare his effectsin advance.TMLikewise our author, whether composinga book or a paper, as can be seenin his very first article, on 'Rhine and Danube legionsunder Dornitian'.73The appendices,too, as above all in the monumental Tacitus,TMthough not at first sight organic parts of the structureand architecture,have their well-foundedfunction-and, not least, they have somethingin common with the excursusesin ancient historiography. Concerninglanguageand style,it may be hopelessfor a reviewerwith a modestknowledgeof Englishto estimatethem in an adequatemannerasa part of the historicalmethodof a brilliant author. But someof thefinesses are really obviousto anyonereadingthe Roman Papersor indeedany of Syrne'swritings. Here again, Tacitus is one of his predecessors: 'The languageis admonitory,violent,emotional.Languageof that kind canbea valuable clue.'75 It is. Allusions, pithy phrases, provocative sentences, ironicalremarksprepare,completeor terminatearguments;theystimulate the reader to agree with the author--or to disagree,until an apparent contradiction is resolvedby an unexpectedturn; they settlea controversy, minimisingby wit and humour a problem wrongly blown up in earlier research.But first and foremost:they imitate the styleof the ancientliterary sourceswhich he uses,and thus they reflect from the first somethingof the mentality of that {5litein which Syme is interested. Someexamplesmaybesufficient.Allusions:'Tacituswascomposing the annals of decline and fall--not merely a dynasty, but an aristocracy brought to ruin.TM You feel the analogy with, and perhaps also the continuity down to, Gibbon's Decline and Fall. Provocation: after the proscriptions,the future Augustus'wasno longera rashyouth but a chill and mature terrorist'? You do not agree,perhaps,but you are stimulated to follow Syme'sview that the developmentof this man from a civil-war generalto the leader of a 'national party' or even to a pater patriae is as enigmaticas his whole personality.Or: 'Galba was a dull disciplinarian, rigid yet liableto be swayedby hisentourage.When he madea decision,it proved to be calamitous.TMThis may seemexaggerated,but it clearly correspondsto Tacitus' famous statement concerningthis emperor, omniurnconsensucapax imperii nisi imperasset(Hist. I 49). A 'problem', dealt with by wit and humour:posterityknowsthe latestwork of Tacitus under the title Annales. 'Why Annales?Or, let it be asked,why not?'79And more humour: the author of the Historia Augusta pretendsto be asexact as

RONALD

SYME, ROMAN

PAPERS

177

an annalist, and producesas proof of this a false 'document' (Aur. 17,1). 'That is the peak of impudence',Syme comments,in accordancewith his thesisconcerningthe Historia Augusta as 'cleverand elusive.... a constant baffiement'.80

Thus, in the senseintended by Syme, his method is surelyperfect.But what about his subject, and this also means his conception of history focusingon the 'lurking oligarchy'?Without doubt Syme'sview of Roman history is as well-foundedas it is innovative. He pays attention to a complex of matters which is really fundamental for understandingthe history of Rome, and in treating this complex, he has gone much further than any other scholar. Nevertheless,it is obvious that Syme's Roman history is not identical with the whole history of Rome. This is not a matter of chronological

limits, but of the problemof what 'history'comprises.It would perhapsbe superfluous to enumerate all possible historical themes for which our author, apparently, has only limited interest or in which he is clearly uninterested.This list might begin,for example, with the economy.8• and could be continuedwith the role of lower socialgroupsor with the system and developmentof Roman law? and go on to religion, or the languageof symbolsand art.a3But it may be instructiveto seeby someexamplesthe limits of the chosenhistoricalsubject,and thus also thoseof the method. The fall of the Republicand the foundingof the Principateis described,

in Syme'swork, from the perspectiveof the strugglebetweenoligarchic groups.Let us grant that this is the main aspect.But the history of the societybelow the 'governingclass'could probablyalsohavebeenput into sharperfocus.For example,Roman societyin the turbulent periodduring the rise of the youngCaesarin 44 and 43 BC.Admittedly,'party' politics and proscriptionsaffectedaboveall the 'governingclass'.But thisclassalso had its large clientelae,which (e.g.) accordingto Appian (B.C. IV 29) demonstratedagainst the proscriptions.And above all, the lower classes werethe massbasisfor oligarchicpolitics.Not very relevant,in the view of our author. We do hear, of course,in the context,about 'the proletariatof Italy', about freedmen and slaves.None the less,their role is limited. 'The proletariat of Italy' meansthe soldiers,who, 'long exploitedand thwarted, seized what they regarded as their just portion'? As to the liberti: 'Freedmen, as usual, battened upon the blood of citizens.TMAnd slaves: during the proscriptions,thereweresome'shiningexamplesof courageor

defiance,of loyalwivesandfaithfulslaves'. 86That in thepoliticalhistoryof Rome the lower classesdid not play the samerole as the governing6lite does not need discussion.But what about the importanceof the lower classesfor social history?

Again: political and socialsystemsin whichthe predominanceof the

178

G. ALFOLDY

'lurking oligarchy'was not asevidentas in the historyof Rome duringthe Late Republic and under the Principate. For example, Athenian democracy is not an attractive subjectfor our author. Writing on Thucydides,he refusesto concentratehis intereston the Funeral Speechof Periclesasan example of the mentalityof Athenian democracy,as severalscholarshave done. The subjectis the makingof this oligarchichistorian,beginningwith an accountof hisdistinguishedancestryand his aristocraticconceptionof history. Democracy evokesa very brief statement:'Thucydidesdeclares that Athens under the rule of Pericleswas a democracyonly in name.'87 That this wassaidby the historianis certainlytrue• but it might perhapsbe instructiveto ask whether he was right or not. Of course,Thucydides'has no liking for the rule of the people'? A historianin a certainharmonywith our author, it seems. But neither is an autocratic kind of monarchy an attractive theme. Thus, to return to Roman history, the monarchy under the 'dominate' of the Later Roman Empire. Whereas Syme has much

interest--and sympathy--for emperorssuchasTiberius,a tragicfigure,or Hadrian, an intellectual,he givesConstantine,Valentinian, or Theodosius a wide berth. The fictitious ancestryof Constantine,to be sure:that may be a fascinatingmatter:89but hardly his role as the thirteenth apostleof the Church. Even if the Later Roman Empire had its own variety of 'lurking oligarchy', Syme'sinterestsand sympathiesare devotedmuch more to the survivingaristocracyof the old type•and, aboveall, to the intellectuallife, under the Christian dominate, of this aristocracy and other educated pagans,as expressedby the Historia Augusta. And as to the Church in the Later Roman Empire, if we are lookingfor someoneamongits representativesfor whom Symehasany sympathy,then it is certainlynot Athanasius or Ambrosius, but the cultivated Jerome.

Even the history of the 'governingclass'of the Late Republic and of the Principate does not appear in all its possibleaspects.The economic backgroundto the power of the Roman aristocracyistakenasself-evident, and, apart from someremarks, particularly in a recentstudy on the wealth of the aristocracyfrom Baetica,90the topic fails to receivea treatment as full and innovative as is that of origins, alliances, offices. Perhaps the picture of women in the Roman aristocracyis also lessthan complete.In the account of social ties, local relations, the nomenclature of their families,

they play an important role. And we hear, in the manner of Tacitus, of women who 'emulatedmen in arroganceand licence',or that 'sharingthe ambition of the husband(and by no meansunconcernedin the making of his career), the wife sharedthe hazardsalso, going with him into exile or constant in the face of death'.9• But all this, it seems, concerns less the

history of Roman ladiesthan the fate of their fathers,husbands,brothers and sons. Yet, at least for social history, there may be other possible

RONALD

SYME,

ROMAN

PAPERS

179

approaches.Further, we only seldom hear of the religion of the Roman 6lite.92And philosophy,highly relevant for severalRoman senatorsand even for emperors,is not a significantmatter for Syme. As for Tacitus, so for him, in dealingwith a man suchas Seneca'it is not the philosopherthat exciteshis interest and sympathy•but the statesman'.93What we have in Syme's work is the history of the Roman aristocracy during the kate Republicand the Empire--concentratingon political historyas explained by several patterns of social history and extended as far as the most attractive intellectual activity of this •lite, namely literature; and it is a history full of refreshingelementsand ideas,describedwith an admirable method, with mastery of composition and style. I hope that this evaluation will not be misunderstood.A historywith the limits I have indicated is not a 'complete' history. Such a history has probably never been written, not even by Theodor Mommsen. Even a marvellous garden, lacking the beauty of desert and sea, cannot be a substitutefor the fulnessof Nature herself.Nevertheless,as a gardenit can be perfect, it can be a garden of delights.We owe to our author a Roman history which is, in its own way, perfect. Of course,it doesnot correspond to the usual present-daywriting of history,either as Geistesgeschichte or from a rigid structuralistor sociologicalpoint of view. It offersa historyof men, rather than one of ideology or institutions--quite like ancient historiography.Here oncemoreSymeon hispredecessor, Edward Gibbon, is relevant:

'He is one of the successors of the humanist

historians

of the

katin Renaissance,with all their preoccupationwith styleand the models of antiquity. But further, might wenot regardhim asthe lastof the classical historians

themselves? TM IV

At this point, the questionmay emerge, whether we can detecta development concerningsubjectand method during an activity of more than fifty years, from the author's first paper on 'Rhine and Danube legionsunder Domitian'9S--a developmentsuchas can be demonstratedfor someother classicalauthors, e.g. Tacitus. I think that as regards the main subject, whichisalwaysthe6lite of Rome,and thusalsoin the method,thechangeis on the whole lessimportant than the continuity.To be sure, there is an obvious 'development', beginning with interests focused on the kate Republicand the Early Empire, and expandedgraduallyto very different matters, such as Tacitus, Sallust, the Historia Augusta and Eater Antiquity. Without any doubt, as BadJansuggests,'few scholars,past or present, have been so successfulat constantly extending their fields of interest rather than leaving one for another'.06But the constant elements,

180

G. ALFOLDY

and thusalso homogeneity,prevail in Syme'swriting. Even his non-Roman themes--Thucydides, Gibbon? other English historians, Renaissance politics,colonialdites in the Spanishand in the BritishEmpires98--seem to presenta marginalinterest,a supplementto the historyof the Roman dite and of Roman literature. If in hiswork a real changecan be detected,then it may perhapsbe expressedby the phrase'lesswar and more humour', in that respectalso partly like Tacitus,who 'gainedin toleranceashe went on, and even in humour'. 99 Before the Second World War, until his thirties, Symehad obviouslymuchinterestin legions,frontiers,wars.•00Afterwards there are only a few studiesdevoted to suchmatters,•oland the Younger Pliny, in whose correspondenceTrajan's Dacian Wars appear only as a possible subject for poetry,•ø2may have become a subject of greater sympathyin his eyesthan the famousviri militates.Dealingwith Tacitus, and later, above all, with the Historia Augusta, more and more humour, wit, and also tolerance have emerged. This may be seen also in Syme's activity as a reviewer,which constitutesan important part of hiswriting. In earlier decades,a long seriesof reviewswas produced:they were always fair, but none the lesssometimeshard on someauthors, who, perhapsas youngscholarsin the difficult field of prosopography,did not measureup to Syme'sstandard. In the last two decadesthere have been only a few reviews. Instead, of course, we have 'A Call for Clarity', a whole book answering critics of an earlier book written by our author. Tolerant, indeed,towardsother positions:'Bad causesoftenhavethe bestadvocates. They need them.TM

Thus,we havearrivedat a lastquestionwhichariseswhendealingwith the work of everygreathistorianor writer. What about his personality,in so far as it is necessary to ask about thisfor the understanding of his work? And that meansfirst of all the question:whydoeshewrite, whetherhistory or works of literature? In the case of an author who likes to insert

autobiographical allusionsin hiswritingasmuchas Symedoes,welearna gooddealabouthispersonality andhisaims.I think thathebelieves more or less the same as, according to him, did one of his predecessors: 'Thucydidesis not advertisinga beliefin prophecy,nor doeshefancythat youcanlearnaboutwarandpoliticssolelyfromthebooks.Nonethelesshe would be appealingto the men of understanding, and, thoughhe hasnot said it, he might have hopedthat someof them might learn as well as approve.'•ø4 Butwearewarned:'It ishazardous to equatethewriterandthe man.TM A closeacquaintancewith Sir Ronald over 15 years--grande mortalisaevispatium--induces thereviewerto describethe writerandthe

RONALD

SYME, ROMAN

PAPERS

181

man by his own words, referring once more to one of his predecessors: 'Tacitus is a subtleand sophisticatedwriter, heir to a long tradition, and writing for men of understanding.The situationshe describesare permeatedwith all the ambiguitiesof high politics--and of human nature--in any age. His manner is majestic and reticent. Perhaps in himself a complicated character, perhaps not. Who can say?Remote, austere and enigmatic, on a surfaceview, yet perhaps in no way a problem in his comportment towards Rome and the Caesars.TM To conclude.

The two ancient authors who most attract

the interest of

Sir Ronald Syme are obviously Tacitus and the author of the Historia Augusta. Not adventitious, I think. Both are extremely revealingfor his personality. His Tacitus is an honestaristocrat with the dynamic of the homo novus, who, superiorto most of his contemporaries,comprehends humannatureand fascinatesby the sharpness of hisintellect,by thecritical distanceof his attitude to events,and by his unsurpassable art of writing. And the author of the Historia Augusta, albeit unlike Tacitus and, with his unscrupulousmanner, totally unlike Syme, appearsto Sir Ronald as an intellectualevoking sympathy,for whom history is not only a matter of gravity, but also a pleasure.The former is 'a poet and a dramatist, not different in that from other historians (such as deservethe name), but better';•07and the work of the latter is 'a garden of delights'.•08These last two judgementsprobablytell usmore about Sir Ronald Symeand hiswork than a long appreciation. And albeit investigationmay be a 'wearisome journey through a "selva selvaggiaed aspra e forte"', with 'thickets and pitfalls, marshy tracts and arid zones',1ø9the study of this work is like roaming in a garden of delights. University of Heidelberg

G•za Alf61dy NOTES

1. R. Syme, Roman Papers, Vol. I-II, edited by E. Badian (Oxford, at the Clarendon Press, 1979). [Henceforth RP.] The bibliographicaladdendaare the work of the editor. I am grateful to ProfessorBadian for doing me the honour of inviting me to write this review as a generalevaluation of Sir Ronald Syme'soverall contribution,and for helpfulcommentson a first draft. To ProfessorA.R. BirleyI am greatly indebted for revising my English and for severalsuggestions. 2. Booksand further collectionsof papers: The Roman revolution(Oxford 1939) [henceforth RR]; Colonial t•lites:Rome. Spain, and the Americas(Oxford 1958) [henceforthElites]; Tacitus,Vol. I-II (Oxford 1958);Sallust(Berkeley-Los Angeles-London 1964); Ammianus and the Historia Augusta (Oxford 1968) [henceforth Ammianus]; Ten studies in Tacitus (Oxford 1970) [henceforth Studies]; The Historia Augusta. A callfor clarity (Bonn 1971;the sub-title'A call of

182

G. ALFOLDY

clarity' was an unfortunate misprint) [henceforth The HA]; Emperors and Biography. Studies in the Historia Augusta (Oxford 1971) [henceforth Emperors]; Danubianpapers(Bucharest1971) [henceforthDP]; Histor,v in Ovid (Oxford 1978) [henceforth Ovid]. For a full bibliography of the author up to 1970seeRP 855 if. 3. For exceptionssee below, with notes 42 and 98. 4. Cf.

The HA

35.

5. RR 7: cf. also Elites 3.

6. RP 315 if., 455 if. and 530 if.; cf. also the study 'Missing Senators', ibid. 271 if. 7. RP 315. 8. RP 541 if.

9. Emperors 1 if.; cf. also AtomJanus 165 ff. I0. For this characterisationof the author of the Historia Augusta, see,e.g., Emperors 261. 11. RP31 if. and 518ff. 12. RP 824. 13. Studies 35. 14. Studies

114.

15. RP 582 ff.

16. RP 805 if., esp. 817 if., mainly devotedto CIL IX 3426. On a summerday of 1978, I drove some500 km to revisethis inscriptionwhich had beentreated by Sir Ronald Syme only on the basisof a photo. Apart from minor itemsconcerningthe restoration of the lost part of the text, I can only shareSyme'sviewsand admire his talent.

17. Tacitus 622 f. and Studies 29.

18. CP74(1979) I ff.,esp. 12if. 19. See, e.g., Elites 1 if.: Tacitus 585 fl.: DP 110 if.: RP 47 if. 20. 21. 22. 23.

Elites 4. Elites 18. Elites 17. Cf. on this Elites 6 if.: Tacitus 598 if.

24. See on this below. note 38. 25. RP 339 and 491.

26. RP 325 if.: 605 if.: 659 fl. 27. RP 292 if.: Ovid 72 if.: RP 477 if.

28. RP 231 if. (and seeesp. 255 if.); 378 if.: 461 ff.; 629 ff.: 774 if.; etc. 29. See. e.g., DP 160 if.: 177 if.: 192 if.: 213 if.: 225 if.: ZPE 37 (1980) 1 if. 30. See. e.g., RR 387 if.; Tacitus 637 ff. 31. Studies

4.

32. Studies 20: cf. also ibid. 120.

33. Cf. esp. CAH X (Cambridge 1934) 340 if. 34. RP361 if. 35. RR 113.

36. Lesempereursromainsd'Espagne(Paris 1965)243 if. Somemay regretthat this important study has been omitted from the presentcollection. 37. Antichthon 11 (1977)78 if.

RONALD

SYME, ROMAN

PAPERS

183

38. Philologus91 (1936) 238 ff.; seealso Tacitus631 f. I havearguedthat this governor, apparently a rival of Trajan, was the Spanish senator M. Cornelius Nigrinus Curiatius Maternus: see Chiron 3 (1973) 331 if. (with H. Halfmann). According to Sir Ronald, this senator could equally well have governed the provinceof Syria c. 89-91, and he perhapsdied in hisprovincefrom diseaseat this time:seeAkten des VI. lnternationalenKongresses ffir Griechische und Lateinische Epigraphik Mfinchen 1972 (Miinchen 1973) 451. At this point I would like to dissent from Sir Ronald's opinion. At all events, the inscription of the abovementionedsenatorwith hisfull career,found at kiria, in hispatria, musthavebeen setup after the assassination of Domitian in 96: thenameof that emperor,who was the donator of the dona militaria of Nigrinus, has been suppressed,whereasthe namesof Vespasianand Titus are mentioned.Evenif it werea matter of a funerary inscription,it would not be plausibleto argue that it was dedicatedto the senator five or more yearsafter he had died. Thus it may be conjecturedthat in 96 Nigrinus was still alive.

39. Esp. JRS 18 (1928) 41 if.; ibid. 23 (1933) 14if.; AJPh 55 (1934) 293 if.; CAH X 340 if., 781 ff. and 803 if.: ibid. XI (Cambridge 1936) 131if.; DP 13if., 84 if., 135 if. and 152if.; RP218 if. and 825 if.; cf. also Tacitus 157if.; Sallust 138if.; Ovid48 if.

40. See, e.g., DP 84 if. 41. Ovid, passim; Menschen die Geschichteroachten I (Wien 1931) 175 if. (Seneca);RP 742 if. (the Elder Pliny); ibid. 477 if. and 694 if.; also DP245 if. (the Younger Pliny). 42. On ThucydidesseePBA 48 (1962) 39 if.; on Gibbon especiallythe studyin Gibbon et Rome • la lumi&re de l'historiographie moderne (Gen•ve 1977) 47 if. 43.

The HA

6.

44. Ovid, esp. 199 if. on morality; and see 72 ff. on connexions. 45. Tacitus 317 if.; Studies 26 f. 46.

The HA

48.

47. The HA 89. 48. Ibid. 49. RP 470. 50. Studies 11 if.

51. See the bibliographicalreferencesabove, n. 42. 52. Gibbon et Rome (cited above, n. 42) 52. 53. Ibid.

52 f.

54. PBA 48 (1962) 40. 55. Gibbon et Rome 53; see also Studies 13. 56. Sallust 43. 57. Gibbon et Rome 58.

The HA

48.

5.

59. PBA 48 (1962) 41. 60. Ibid. 39. 61. Studies 119. 62.

RP

55.

63. PBA 48 (1962) 41.

184

G. ALFC)LDY

64. Ibid.

56.

65. Ibid. (the last sentenceof the paper). 66. On the Spanishwars seeAJPh 55 (1934) 293 if. and RP 825 if. On the political ideas of Tacitus, seebriefly Studies 119 if. 67. The HA 48.

68. Akten (cit. above, n. 38) 585 if. (at p. 598). 69. RP 711. 70. The HA 75. 71. RP 650 if. 72. Studies 21.

73. JRS 18(1928) 41 if. 74. Tacitus 625 if. 75. Studies 136.

76. Ibid. 105; see also 136.

77. RR 191.Of course,the word 'terrorist',when Sir Ronald wrote the phrase, did not mean quite the sameas in our day. 78. Studies 134; cf. Tacitus 182. 79. Studies 1. 80. The HA 29 and 54.

81. But see the study on the imperial finances under Domitian, Nerva and Trajan, RP 1 if. 82. Of course,our author is interested,if not in law itself,at leastin the persons of jurists:seeRP 790 if. and Proc. Amer. Philos. Soc. 116 (1972) 406 if. 83. A brief but highly interestingtreatment of religion: RP 313 f. See also below, n. 92. 84. RR 194. 85. Ibid. 195.

86. Ibid.

190.

87. PBA 48 (1962) 50. 88. Ibid.

89. Ammianus 116; Emperors 204 f.; and especially The HA 57 if. 90. Ktema 2 (1977) 373 if. 91.

Tacitus 534f.

92. But see,e.g., the brief but important remarksin Tacitus467 f.; Ovid 103if. 93. Studies

94. 95. 96. 97.

138.

Gibbon et Rome (cit. above, n. 42) 55. JRS18(1928) 41 if. Introduction to RP, p. xii. See the bibliographical referencesabove, note 42.

98. Emory UniversityQuarterly 18 (1962) i 29 if. (on Gibbon, Macaulayand Toynbee);RP 470 if. (Roman historiansand Renaissance politics);Elites 24 if. (SpanishAmerica and English America). 99. Studies 106; see also ibid. 137 and further Tacitus 476.

100. Cf. the bibliographicalreferencesabove, n. 39. 101. Thus esp. RP 825 if. 102. DP 245 if.

RONALD

SYME, ROMAN

103. The HA 83.

104. PBA 48 (1962) 43. 105. Studies 106; see also ibid. 10. 106. 107. 108. 109.

Ibid. 131. Tacitus 546. AtomJanus 4. The HA Ill.

PAPERS

185

THE

FAMILY OF PHILAGROS THE CULT OF ASKLEPIOS

ERCHIEUS AND AT ATHENS

The starting-pointof this discussionis the career of Paramythosson of Philagros Erchieus, who served as Secretary to the Athenian Boule in 387/6 BC(PA 11629;seeII 2 28 line 1).2 In order to have beenof an age to servein the Bouleof 387/6, Paramythosmusthave beenborn not later than 417/6 BC; probably earlier, in fact, since the post of Secretary was a distinguishedone.2 His father, PhilagrosErchieus,may be identifiedasthe mover of a decreefor reinstatementof a proxeny, passedin 403/2 BC(PA 14203=14209?See II 2 2 line 8). Two other men named Philagros are known in the fifth century: one, from Halai (PA 14207),is clearlynot the father of Paramythos,thoughhe could, of course,be the orator of 403/2 BC;nothing is known about the other, except that he died late in the fifth or early in the fourth century (Agora XVII no. 999): he could be the father of Paramythos, or the man from

Halai.

In the fourth century the name Philagros is found in severalfamilies, includingthat from Halai. One example that lacksattribution to a demeis Philagros son of Asklepiodoros(PA 2625a), who died at or before the middle of the century (PA 14205a;seeII 2 12888):unlesshe died during childhood, this man is likely to have been born soonafter 400 BCand was thus of an ageto be the grandsonof one of the fifth-century Philagroi, all of whom would have been born around the middle of that century. The name Asklepiodorosis rare at Athens and is not attestedbefore the fourth century:three examplesare known for that century:one isthe father of Philagros;another(PA 2627a) is the father of a certainAsklepiadesand musthavebeenborn during the secondhalf of the fourth century,sincehis son set up a dedicationto Zeus Meilichios in Peiraieusearly in the third century(PA 2589a;for the date, seeII 24619).3 The name Asklepiodorosis slightlymore commonduring the third century.Five instancesare known, apart from the father of Asklepiades:of these,two havedemoticsand three do not; none of them can be linked definitely with the deme Erchia.4 After the third century the name occursmore frequently, if sporadically,but is never common.

5

Early in the secondcenturythere occurredthe death of Asklepiodoros (PA 2629) sonof AsklepiadesErchieus(PA 2602;seeII 26102for the date): the name Asklepiodoroscan thus be shownto have emergedin the deme Erchia during the secondhalf of the third centuryand, more significantly for my argument,to have beenpart of a generation-sequence in which the namesAsklepiodorosand Asklepiadesalternated. 186