"Leitourgia" and Related Terms Lewis, Naphtali Greek, Roman and Byzantine Studies; Fall 1960; 3, 4; ProQuest pg. 175 Leitourgia and Related T etms N aphtali Lewis THE IDEA THAT THE MAN OF MEANS is obligated to expend some of his wealth and time in service to the community is common in almost all periods of history. The services performed vary with era and area. So does the nature of the compulsion: sometimes it is a statutory obligation) sometimes a less formal (though not therefore necessarily less effective) social requirement or expectation. In the ancient world it was the Greek city-state which extended the concept of compulsory public service beyond military duty and menial labor. The Roman Empire, in its turn, perpetuated this Greek institution, expanding it into a full-fledged system of local administration. The generic Greek term for such compulsory public service was AEtTOVPyLU. Beginning in the fourth century B.C. (as we shall see below) the term was occasionally applied to religious service, and it is in that sense that the word survives in the modern languages of the Western world. Leitourgia as an institution is discussed in the appropriate places in all the principal handbooks and encyclopedias of classical antiq­ uity. But oddly enough, none of these customary reference works goes into very much detail on the semantic history of the term AEtTOVPYLU and its compounds and derivatives. The only recent study of that kind - and this must surely come as a surprise to most classicists and ancient historians - is that by H. Strathmann (with a section on rabbinical literature by R. Meyer) in G. Kittel's Theo­ logisches Worterbuch zum Neuen Testament IV, 4 (1938) 221-38. While this excellent presentation includes representative citations from Greek literature, inscriptions, and papyri, its emphasis is natu­ rally on the Bible and related writings. With Strathmann's analysis as a base, the present article will examine the forms and semantic 175 176 NAPHTALI LEWIS GRBS 3] history of the lez"tourgz"a word-group as it appears in secular Greek literature, inscriptions, and papyri of all periods; and will conclude with an inventory of all occurrences in the writers and inscriptions of the classical period, during which the semantic evolution began and was in essence completed. I. The Leitourgia Word Group The following terms are found: '\ ' aI\EI:TOVPY"f/TO~ AEf,TOVPYI.W and compounds with avn-, a7To-, KaTa-, 7TPOU'-, U'VA-, and perhaps EKI AEf,TOlJpYTJILa A€f,TOvpY"f/U'La and a compound with a-. AEf,TOVP'Y'r}U'LILO~ AELTovpYLa and a compound with a-. A€f,TOvpyf,aK6~ AELTOVPY"K6~ AELTOVPYtOV A€"TOVpy6~ and a compound with V7To-. II. Etymology The earliest pertinent text is Herodotus 7.197: A-r}"TOV 8E KaAl.oV(Tf, TO 7TpVTav-r}tOv 0;' 'AXatOL.2 There is nothing further on the subject in the ancient writers till we come to Plutarch, who con­ nects Latin lz"ctor with Greek A€LTOVpy6~ (an erroneous but at the time apparently popular view), remarking OTt yap AVTOV tixp" vVv ,c;:., , \ \... ... ..7]00S oVOP.a.rOVtTLV. [1960 LEITOURG IA AND RELATED TERMS 177 anus, the third-century commentator on Demosthenes4 - and it is recorded in the Byzantine lexicons,/) one of which incorporates in its definition also the term 7TPVTUV€l,OV, culled presumably from the Herodotus text cited above.6 In short, it was the established view in antiquity that the words of the A€"TOVP'YtU group were compounded of the elements "pub­ lic" + "work", to signify "work for the people", hence "service to the state". Modern scholars find nothing to quarrel with in this. Most present-day etymologists, however, regard the origin of A7JO~ (AUOS-) itself as undetermined. T 4Ammoniu5, nEpl lJ!-,olw1I Ka~ ~LarpopWII 7I.1i~ewp, $.V.: )..7JTOvp'Yeill 6La TOU 1], Iud )..LTOVP'YeiP /5ul ToD L, /5LarpepELP rpTjul dlavp.os ~P il'TrO!-,PTIIJ.aTL /5wTlipas (I. -lip,!,?) 'L\LaaO!. TO p.ep 'Yap h7JTOVP'Y€i", TO T~ aTJp.,!, V7rTjpeTe'ip' MiTOP 'Yap cpaUL TO li1]P.OuLOP. lI.'l/iTor ap.rpe7rfPOPTO [this is obviously a quotation from epic poetry - cf. Homer Odysuy 15.467]. TO of hLTOVP'Yeip, KaKa 7I.f'YfLP (XLT- is here not the iotacized spelling of lI.eLT-, but a differ­ ent word: d. haop'Yos, hEOVP-YOS (hew-), hLTO( v )p-yos, hLWP-YOS, on which see esp. Hesychius, s.vv.) Moeris, Lexie. s.v.: XTjTOVp-yEIP OLa TOU 'l7 'ATTLKoi, liLa lie TijS eL oup()oyyov uEhA'l7PEf A'l/LTOP -yap TO 07Jf.'OfTLOP· Ulpianus, Schol. in Demosth. 20, p. 512 Dindorf: AeLToP li€ ~KaAovP 01 7raAaLol TlI a7Jp.ouLoP, Mep XELTOUP-YEip TO Els TO liT]P.OULOP ~p-yatEU()aL lAE"/O,,. 5Hesychius, $.VV.: II.7JLTovp'Ye£P· AELTovp,,/e'ip' A'l/LTOII "lap /57Jp.OULor. II.fiTOP· li7JJ.tC)(]'LOP. Zonaras, S.V.: hELTOVP'Yeip T6 V7r7JPETELV, 7rapa TO XTJLTOP, is €fTTL TO li7J/-toULOP. &71"6 oop TOU lp-yop Kal X'I/'iT01l h7JLTOVP'Yf(ip, Kal KaTa UVfTTOh~P TOU 7J els e XELTOVp'YEip, Kal KaTa CTvpalpEULP Toil E Kal , AELTOvp'YEip. Tittman (Leipzig 1808) brackets the last phrase: d. n. 6. Suidas: s.v.: AELTOVP'Yia' Kvplws 7J lJ7Jp.ofTia V7I"7JpEu[a' 7rapa TO A'I/'iTov Kat TO ~P'YOP. Thomas Magister p. 227, 17 Ritschl: oi p.ep h'l7(L)TOVP-YOS lJUl. ToD 'l7, 7/TOI <> T YJ > .ft.9 In the light of epigraphical and papyrological evidence available to modern scholars, the processes of the vocalic and orthographic changes involved can be reconstructed more completely, viz.: pre~ sumably original disyllabic Yj£ > diphthong Yjt, which came to be pronounced ~, then written Yj or Et, later pronounced ~ and some~ times written (, (iotacism). Before dealing specifically with the AEtTOvpyta word~group, it will be useful to summarize the evidence on these vowel changes in general.lO Yjt > Yj Inscriptional evidence for the silencing of the iota in YJt begins, at least in Ionic, as early as the sixth century B.C.ll The change is in evidence in Attica in the fifth century, and in Thessaly in the fourth century. Subsequently it is visible also in inscriptions from Pergamum and Magnesia, and in papyri from Ptolemaic Egypt. Yj£ > E£ Inscriptions from Athens and elsewhere (e.g. Amphipolis, Eretria, Olynthus, Oropus) show signs of this change from ca. 400 sef. supra, nn. 5 and 6. There was also minor dialectal variation in the vowel of the second syllable. IG VII, 3083 (Boeotia, third century B.C.) has w. IG JI2, 1140 and 1147 (Attica, fourth century B.C.) have 0, but omicron frequently represented ou in inscriptions of that period. sPor a brief summary on these transformations of 7)/, cf. M. Lejeune, Traite de phonbique grecque2 , (Paris 1955) 196. The essential evidence is cited below. lOBibliography: E. Schwyzer, Griechische Grammatik (Handbuch der Altertumswis­ senschaft 11.1) 1 (Munich 1939) 193, 200-2; E. Mayser, Grammatik der griechischen Papyri aus der Ptolemiierzeit 1 (Berlin and Leipzig 1923) 74-80, 87-92, 122-28; L. R. Palmer, A Grammar of Post-Ptolemaic Papyri I (London 1946) 1-3. These standard works give summaries of the relevant data and bibliographies of earlier specialized studies; to the latter should be added B. Bondesson, De sonis et form is titulorum Milesiorum Didymaeorumque (Lund 1936), esp. pp. 55-59. llLinguistic evidence points to the disappearance of the iota in "Inlaut" already in prehistoric times: d. Schwyzer. op. cit. 200. [1960 LEITOURGIA AND RELATED TERMS 179 B.C. It is impossible to tell how much earlier the change actually began, because in the earlier period E represented both E and 7J in most Greek inscriptions. The earliest clear case is IG II2, 1414, which has KA7JIS in line 44 and KAEIS in line 47. 7J=Et. Inscriptions and papyri show the equivalence of these two sounds and the interchange of these spellings from the third century B.C. on. EL = L Argive and Boeotian inscriptions show signs of iotacism as early as the fifth century B.C. Attic inscriptions and papyri show the tendency fully established in the third century B.C., as do in­ scriptions of the second century B.C. from Delphi, Magnesia, and Pergamum. The above changes are well exemplified by the leitourgia word­ group, as can be seen at a glance from the following chart: SPELLING >"."t7- >"."T- EARLIEST OCCURRENCE LATEST OCCURRENCE DOCUMENT PLACE AND DATE DOCUMENT IG IP, 1140 Athens, 386 B.C. PSI 435 SIGa,344 = Teos, 303 B.C. Sardis VII, 2 Welles, Royal Corr.3 IG IF, 305 Athens, before 336 B.C. PPetrie II, 4(9) Egypt, 255 B.C. = III, 42C(2) PLACE AND DATE Egypt, 258/7 B.C. Sardis, ca. 225/175 B.C. •• .. These spellings continued through antiquity to modern times. In order to illustrate the spelling variations which took place in the living language, the chart above is compiled exclusively from contemporary documents - i.e. inscriptions and papyri. It is clear from these that AELT- was the standard spelling from about 300 B.C. on, and that A'YJLT- and A'YJT- disappeared from ordinary usage in the next hundred years or so. But copyists continued for a few centuries longer to reproduce the older spelling in works of literature. The latest such survivals12 are found in the papyrus MS of Aristotle's 12Not counting, of course, the Byzantine lexicographers, who record these long-vanished forms as a matter of philological history. 180 NAPHT ALI LEWIS GRBS 3] Constitution of Athens, a copy made ca. 100 A.D., which has ~'YJ"T~ in all occurrences (once, in § 56, corrected from A€LT~); and in British Museum Papyrus 132, a first~entury copy of Isocrates Or. 8, which has A'YJT-/3 as well as A€LTOVPYLWV corrected from ALTOVPJI€LWV, and probably AtT-.14 A fragment of Hyperides written in the late second or early third century shows only A€LT-/f> and an early-third-century copyist of Demosthenes Or. 21 consistently writes ALT_.16 An unfortunate by-product of recent times is worth noting. Ever since it became apparent that the older spelling persisted in Attic inscriptions through most of the fourth century,17 modern editors of classical authors earlier than Aristotle have generally printed AYJT- where the codices, being of late date, have A€LT-. Whether this is necessary or even justifiable in all cases is, to begin with, questionable, since (as we have seen) in the same body of Attic inscriptions the spelling A€LT- appears already in the mid-fourth century and other instances of 'YJ" > EL go back at least to ca. 400 B.C. But, what is worse, some editors have not always been consistent in emending to AYJT-. The result - most apparent in the Attic orators, who use the leitourgia terminology most frequently - has been to deposit a legacy of enormous confusion in the editions of the last seventy-five years. To cite but a few examples among many: In his Andocides, Blass prints AYJTOVpJI€'iv in 1.132 (codd. ~€tT-), but leaves ~€LTOVPJlLat in 4.42. In his Isocrates he has changed the spelling in some places and left AEtT- in others, without explanation; and Preuss' Index Isocrateus faithfully reproduces Blass' spellings. In his Demosthenes Blass consistently prints AYJT- (but KAELr;, not KAr1~!)' but in only five of the more than one hundred occurrences does he indicate in the apparatus that the codices have ~€LT-; the unwary or unwarned reader is thus left with the false impression 13Similarly, h'r]CT7" a,s in a first· century copy of Hyperides (British Museum Papyrus 108 + 115), but h'r]'CT7"OU in the second-century Didymus scholia (PBerol. 9780). Cf. the first- and second-century authors quoted in note 4, and Quintilian 1.7.17: [iota] quibusdam etiam interponunt, ut in .AHI~THI, quia etymologia ex divisione in tris syllahes facta desideret eam litteram. l'*lournal of Phi!ology 30 (1907) 6, 11, 74. 15POxy 1607, line 20. 16POxy 1378, Jines 8, 18, 19. l7The latest is IG II2, 417, which dates from some time after 330 B.C. [1960 LEITOURGIA AND RELATED TERMS 181 that in the other instances the codices have AVT-. In the Bude edition of Isocrates, Mathieu and Bremond mostly printed AVT- in Volume I (1928), but in Volumes II (1938) and III (1942) they abandoned this practice, apparently deeming it unnecessary, and reverted to the AEl.T- of the codices. IV. Semantics When the term leitourgia first appears in Greek literature it refers to specific state services required of wealthy citizens and residents. At Athens, to which most of the evidence of the classical period pertains, the major liturgies were equipping and manning a trireme of the fleet (trierarchia), and providing a chorus for a dramatic festival (choregia); less important - and less expensive - were a number of other functions, mainly in connection with religious ceremonies (e.g. gymnasiarchia, lampadarchia). From this original use as a political terminus technicus (a),t8 the term began, toward the end of the fifth century and increasingly in the fourth,19 to be used in the broadened sense (b) of any serv­ ice to the community. The line of distinction between meanings a and b is sometimes difficult to discern, especially in the rhetorical language of the orators; examples will be noted below. From meanings a and b leitourgia began in the fourth century to be used in a still more generalized sense (c), viz. to designate a service of any kind, for any beneficiary, not necessarily for the benefit of the community.20 Within the generalized sense c, there developed two important specializations of meaning: d - cultic service to divinity. This meaning, which also makes l8The designations a-d used in this section correspond to those of Strathmann. loc. cit. 223-25. 19Strathmann's generally excellent analysis is defective on this point. Finding meaning b "ganz besonders ... in den Papyri" (p. 224). he cites examples only from papyri. none earlier than the second century :B.C. 2()As Strathmann expresses it (ibid.), the "ATiiTO! dement becomes "v611ig verblasst" in popular usage. The verb 'XOP"J,,"(EIIJ underwent similar generalization: cf. Liddell-Scott-Jones, S.tI. 182 NAPHT ALI LEWIS GRBS 3] its first appearance in fourth-century literature,21 was carried through the Septuagint into Christian usage, where it still remains today. e22 - engineer or sapper service in the military forces. The earliest extant examples of this meaning are in PCairo Zenon 59015 verso ( = Sammelbuch 6782) and PHibeh 96, both of 259/8 B.C. The following list shows the distribution of the different mean· ings in the extant occurrences of the fifth and fourth centuries B.C.23 Under a and b the texts refer to Athens unless otherwise noted; instances of general language, specific to no particular city, are indicated by the symbol *. Passages which illustrate the broad. ening in sense from a to b are identified thus: t - The passage uses the term in sense a with an overtone of b, or VIce versa. :I: - The same passage uses the term once In sense a and again in sense b. S~ns~ a FIFTH CENTURY Antiphon 5.77t (Mytilene and Athens) Isocrates 18.58, 60, 64 Lysias 20.23; 21.5, 12, 13, 16, 19+,23; 32.24 FOURTH CENTURY Aeschines l.97, 101 Andocides 1.132t; 4.42 , . Aristotle, Ath. Pol. 27.3, 56.3; Oee. 1347 a 12, 14, 1352 a 4(Caria); Pol.* 1309 a 18, 1320 b 4t; Rhet. 1399 a 3424 21Here again as in b (cf. supra n. 19), Strathmann cites no occurrences earlier than the second century B.C. He appears to regard Aristotle, Politics 7, 1330a8-13 as a mixture of meanings a and d. He ignores Demosthenes 21.56. There is also a clearcut case in PSI 435 of 258/7 B.C., on which cf. A. Deissmann, Light from the Ancient East (London 1927) 154-158 = Licht t/om Osten4 , pp. 123-125. 22Strathmann does not make a separate category of this meaning, though he remarks (p. 224) that it, like d, is a "Spezial£all" of c. 23The material of the Hellenistic and Roman periods, already collected, will be the subject of a separate article. In searching the sources I had the welcome assistance of Dr. M. Reinhold for the literature and Dr. iR. K. Sherk for the inscriptions. Their assistance was made possible by a grant from the American Philosophical Society, which is here acknowledged with gratitude. 24Cf. the anonymous commentator, Berlin ed. Vol. XXI (2), pp. 323 §15, 325 §17. [1960 LEITOURGIA AND RELATED TERMS 183 Demosthenes 5.8; 18.91 (Byzantium, Perinthus), 104, 108,267; 19.282; 20.1, 8, 18-23, 2825, 40, 126-130, 151; 21.11, 14, 5626, 61, 108, 126, 151-155, 158, 165, 167, 169, 171, 189, 225; 25.29, 76, 78; 27.64; 28.3, 17, 19, 24; 29.24; 36.39-40, 42; 38.25-26; 39.9; 42.3, 21-23, 25; 45.66, 78; 47.48, 54; 50.9,21,31,39-40,58,66; 51.7, 17; 52.26; 59.117; prooem. 48.327 Isaeus 3.80; 4.27; 5.29, 36, 39, 45; 6.38, 60-61, 64; 75, 38, 40, 42; 11.40, 48-50; fragg. 22 Thalheim (= 1 Sauppe), 29 Thalheim (= 34 Forster = 130 Sauppe) Isocrates 8.13t, 20t, 128; 12.145; 15.5, 145-146, 150, 154, 158; 16.32, 35; 19.36 (Siphnus) Lycurgus, Leoer. 139 Lysias 3,47t; 7.31; 18.7+; 19.58; 25.12; 26.3-4; 29.4 Theophrastus, Char. 23.7; 26.6 Xenophon, Mem. 2.7.6 IG 112, 30528, 417, 1140, 1147 Sense b FIFTH CENTURY Lysias 21.19+; 31.12, 15 FOURTH CENTUR~ Aeneas Tacticus 13- Aristotle, Ath. Pol. 29.5t; Eth. Nicom. 1167 b 12; Etk. Eudem. 1242 b 30; Pol. 1272 a 203() (Crete); Pol.· 1278 a 1231, 1279 a 11, 1291 a 33-3832, 1305 a 532, 1314 b 1432, 1321 a 33 Aristoxenus, frag. 35 Wehrli Demosthenes 1O.28t; 22.65; 24.172; prooem. 14.2 Isaeus 4.2932 ; frag. 30 (= 35 Forster = 131 Sauppe) Isocrates 356; 7.25 Lysias 18.7+ Plato, Laws 12, 949C· 2liIn some codices. 26Double entendre with sense d: cf. Pollux, Onom. 3.143, which refers to this passage: 17l'l TW/I f./,OVUtKW/I (sc. a"Yw/lw/I) Uj'O/l, IS ¢71Ut il71p.oufU/l7]f, TO AEtTOVP"YE'i/l T~ 8E~. 27The terms leitourgia, etc., are commented on frequently :by the scholiasts to Demos· thenes, viz.: Schol. ed. Dindorf, pp. 72, 108-10, 155-56, 233, 458t·59, 462, 466-69, 471, 473-75, 480-81, 483-84, 488, 512-14, 579-80, 627-29, 631 (line 14), 636, 645; Sebol. Patmiaka, in BeH 1 (1877) 147; also hypotheses oratt. 20 and 42. 28This inscriptiflll is too mutilated for a definite determination, but since the stone contains a state decree the sense is probably a rather than h. 29Cf. also Scho1. Demosth. ed. Dindorf, p. 591, lines 10-12 (T1,/I ~7T'tfJ&'}"AoVUa/l 7I'E/l7JU' AEtTOVP'Yla/l €7I'A1}PWU€' AEtTOVP"Yla o€ 7I'/€/l7]TOS .q Ote}; TOV uwp.aTof du¢opa) , and p. 631, lines 17ft 30possibly sense d here: d. 1330 a 12. 31Sense c followed by h. 3lilErroneously listed by Strathmann under sense a. 184 NAPHT ALI LEWIS GRBS 3] Sense c FOURTH CENTURY Aristotle,33 Animal. incessu 711 b 30; Eth. Nicom. 1163 a 29; Eth. Eudem. 1242 b 17; ]Utl. et senect. 469 a 3; Oec. 1343 b 20; Part. animal. 650 a 9, 674 b 9,20,689 b 29; Pol. 1278 a 1231, 1335 b 28 [Demosthenes] 503534 Hyperides, Lycophr. App. £rag. 1 Colin = POxy 1607, lines 20-21 Isocrates 15.156 FOURTH CENTURY Aristotle, Pol. 1330 a 134t85 Demosthenes 21.5636 BROOKLYN CoLLEGE September 19M Sense d 3SCf. also the anonymous commentator, Berlin ed. Vol. XXI (2), p.19. 34Denouncing a well-paid and well-treated trireme crew as el",O/)s ••• clTeAelas /llye,,, 'T&;" "op.,rop.~"",,, ill 'TV Jl7]2 Ae'TovP'Y'&;'" the language (note especially the underlined words) has all the solemnity of sense a, but it is used here perhaps in sense b, more likdy in sense c with possibly an overtone of b. 85C£. note 30. Bec£. note 26.