STOP Early Journal Content on JSTOR, Free to Anyone in the World This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world by JSTOR. Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries. We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes. Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- journal-content . JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact support@jstor.org. STUDIES IN GREEK NOUN-FORMATION Based in part upon material collected by the late A. W. Stratton. 1 DENTAL TERMINATIONS II. 1 Wobds in -as, avros 2 By Carl D. Buck Words in -as, -avros 2 comprise a small group of verbal adjectives formed with -vt- or -r-, a few substantives of miscellaneous char- acter, and a considerable number of proper names, personal, geo- graphical, and ethnic. 1. Words formed with -vt-. — The clearest examples are from dissyllabic bases, either in a dissyllabic form, as avdpoda/ias, etc., from -dana-vr (cf. khapaoa, trav5api,arcop) , aKap.as from -Kap,a-VT- (cf. Kanaros), raXas from ra\a-vr- 3 (cf. eraXaxraa, rakaepybs), or in a monosyllabic form, as iroKvrkas, arhas, from -rka-vr-, this from r\a- (cf. hr\r)v) with the regular vowel shortening before vt. The suffix is that which has its principal seat in the active participles, but which also occurs, in Greek and elsewhere, in words detached from the regular verbal system, as 66011$, dddvros, Lat. dens, dentis, or yipuv, Skt. jarantr. The passive force which must be assumed for aSa/xas as applied to a hard metal ('unconquered'), and likewise for the (earlier) proper name 'ASdjuas, is noteworthy. But the active force would be less insistently felt in this type than in the regular participles, and ananas, properly intransitive 'untiring' but equivalent to 'unwearied,' may have been a special factor in 1 Cf . Introductory Note, CP. 5. 323 ff. 2 Exclusive of aorist and other participles belonging to the regular verb-system, and also of contracted forms of -oe«, -aeroj. 3 The stem toXovt-, usually replaced by roXoy-, is attested by Choeroboscus 1 . 268. 33, quoting from Hipponax (riXacTi) and Callimachus (afooT&XavTa) . Horn. T&XavTa 'scales' is a neuter plural of this stem, to which was formed later the o-stem singular raKavrov. Cf. Kretschmer, Glotta 3. 266 ff., Solmsen IF. 31. 497 ff. An uncertain example of this type is tbpvubas Hesych., beside tbpvicbuv. If correct, it presumably has jr-inflection and points to -KOfa-vr-. Cf. «<6a/MS Hesych. [Classical Philology XVI, January, 1921] 51 52 Cakl D. Buck the use of ad anas. Note also the coexistence of active and passive meaning in the verbal adjectives formed with -r-, as -/JAtjs (CP. 12. 174). The pronoun was, stem wavr-, is, according to its most probable explanation (Brugmann, Totalitat 61), a similar formation from IE. Hwd- 'swell' (Skt. gvd-, Gr. kxko), the semantic development being from 'swelling, swollen' to 'full, whole.' Besides the well- known intensive compounds airas, avnwas, irpowas, we find also Boeot. fyiras, IG. 7. 2712. 19, and Cret. ^TriTras, SGDI. 4983, 5026, 5039. With the latter compare also avairas, i.e. ava-iras, Anth. P. 7. 343. 2. Words formed with -r-. — The stem of iiirepiciidas is to be analyzed as Kvdav-r-, formed from the verb-stem of icvdaivu) with the same -r- which is in Greek most commonly added to root-forms ending in d, rj, or co, as in irpoff\i)s, etc. (CP. 12. 174). Words of this type may have active or passive force, and inrepnvdas, in the passages of Homer and Hesiod where it occurs, is not passive ('exceeding renowned,' L. & S.), but active 'boasting,' cf. Brugmann, IF. 11. 291 ff. The same stem appears in Kvdas, Cret. Kbdavs, and examples of the same formation are seen in other proper names, as KipSas (nepSaivw), '0v6p.as (bvopaivo)), Ilepi^aj (4>alvu), etc. So possibly Tlyas, rarely appellative 7170$, from a *yi,yaivco related to yiyvopcu, but the name may well be of non-Greek origin. yiyYpas, if it has rr-inflection, 1 is from yiyypaivu implied by yiyypavra. 3. Miscellaneous substantives. — Under this head are grouped a series of words the analysis of which requires discussion for each individual case, though some of them are attributable to one of the two foregoing types. i/i&s is related to OE. slma 'band, rope,' Skt. slman- 'part in the hair, boundary,' for which *lp.o}v (cf. l/xovia) is the equivalent to be expected in Greek. Brugmann, IF. 11. 293 ff., assumes a derivative *lp,alvu) 'provide with a band, fasten,' whence lp,av-T- (like mbav-T- from Kvdaivco), which then like many other agent- nouns was applied to a material object (cf. Eng. fastener), so that 1 The evidence rests solely on the passage of Amphis quoted in Ath. 175, where Kaibel now reads t&v yiyypav ye (MS yiyypai/Tt), while Meineke, Fr. Com. Gr., fol- lowing Dindorf, reads rbv yiyypavra. Studies in Greek Noun-Fobmation 53 it came to have the same meaning as, and displaced the original noun from which the assumed *lfialvw was formed. Instead of following this available but circuitous route, one must prefer, if possible, to see in l/xavr- a simple extension of the original n-stem noun (cf. x«P tT_ beside x^P'"; Skt. harit- beside hari-, and slmanta- beside siman-). The obvious difficulty, namely that from an inherited by-form with ^-extension one would get *lp.ovr- or *i/*ar-, may be overcome by assuming a prehistoric but specifically Greek extension on the basis of an antevocalic weak form *ip,av- (e.g., gen. sg. *lp,avos). avSpias, obviously derived from avfip and no doubt first applied only to figures of men, is explained by Brugmann, loc. tit., through the medium of an *avhpiaivo> 'have the characteristics of man/ whence avSpiav-r- 'that which represents man.' In the absence of any trace of such a verb or of any n-stem to make its existence especially probable (though of course verbs in -aivu from other than n-stems are plentiful enough), one can have no confidence that this is the particular missing link needed. What one looks for, but fails to find in the extant material, is some noun in -(t)as, -(t) euros containing the notion of 'shape, figure,' after the analogy of which avSpias could be formed directly from kvT)p. 6/cpl/3as, KiXXtjSas, Xu/cd/?as, &\t/3as. — These rare and difficult words have the appearance of belonging together, and for the first three derivation from paiva) is commonly assumed. 6icpl(las is a technical term for (1) a platform from which actors declaimed, 1 (2) the buskin with very thick soles worn by actors to increase the stature, 2 (3) the painter's easel, 3 (4) some part of a chariot, 4 and other kinds of supports or "risers." 5 For all these uses the old derivation from o/cpis and flaivco is satisfactory. Only bnpi- represents an adjective, originally 'sharp, pointed' (cf. cucpos, Lat. 1 Plat. Symp. 1946: kvafiaivovros kiri rbv btcpifiavra juerd t&p inroKpir&v. Schol. iKpipavra' to \oyelov k4> °& ol rpayipBoi fuoivl^ovTo. rivte Si KiWipavra TpuriceSrj a' oi Xaravro oJ inroKpiraC. 2 Philostr. mi. Apoll. 5. 9; 6. 10; nit. Sophist. 1. 9. Themist. Oral. 316d; Et. Mag. s.v.; Hesyoh. bicplftavTas' ip.0a.Tas' 3 PoU. 7. 129; 10. 163. 4 Phot., Suid., Hesych.: duplfias' axy/ia (or 5xiC«) rjridxov. Obscure. 6 Cf. Suid. kclI to. inrepeio-paTa t&v fyXLvcav deaTpoiv. 54 Carl D. Buck deer, etc.), whence 'projecting, high, raised,' and the question arises whether -fiavr- is to be analyzed as -fta-vr- from j3a- of eflriv, or as -fiav-T- from (lav- of fiaipu. The latter is to be preferred, since the suffix t occurs in nouns of action (cf. Sals, Skt. stut- 'praise') as well as in those of agency, and the transfer from action to place would be parallel to that seen in (3rj(i,a and fiaais. taXXifias is likewise a technical term for various kinds of stands and supports, e.g., a stand for shields, 1 table-base or three-legged table, 2 part of a chariot-frame, 3 supports for a platform, 4 part of an engine of war, 6 and a painter's easel. 6 It is derived from kiXXos 'ass,' 7 which might without change in form be used for a stand or support, like Eng. horse, easel = Ger. Esel, Ger. Bock, or, in a diminu- tive form, Fr. chevalet. From both the formal and semantic point of view, KiXXijSas is best understood, not as the result of independent composition, but as formed from kiXXos, in its transferred sense, on the analogy of dicpifias, which is so closely allied in meaning. Xvk&/Sos in Horn. Od. 14. 161 = 19. 306 denotes a fixed period of time, presumably 'year,' 8 as it was certainly understood by the late writers 9 who used the word in imitation of Homer. The most reasonable analysis, despite objections and rival suggestions, 10 is \vKa-fiavT- 'light-course.' The second element is -/3cw-r as a noun of action (see above), and the first contains the weak grade of •Aristoph. Ach. 1122, taXKlffavra imridos. 2 Schol. Aristoph., loc. cit.; Hesych. 8 Poll. 1. 143. 4 Athen. 208c. 6 Bito de Mach. 110, 111 ( = Wescher, Poliorcitique dea Greca, 58, 59). 6 Poll. 7. 129; 10.163. 7 Hesych. s.v. Poll. 7. 56. 8 Stengel, Hermes 18. 304 ff., argues that the meaning in Homer is not 'year' but 'month.' 9 Ap. Rh., Bion, late metrical inscriptions, and coins. See L. & S. 10 Johansson, Beitr. s. gr. Sprachkunde 16 ff., objects that -fiavr- "kann unmoglich 'gang' bedeuten" and could only have participial force (but see above), and analyzes \vK-a-0-airr- with a succession of suffixes. Fick, Gott. Gel. Am. 1894, 240, followed by Beehtel, Lexilogus 218, analyzes Xvic-4J?as, referring to Hesychius' gloss ft/3a - rpoxfo (tp6xos). But if Fick's further identification of ft/3o with late Lac. yl}' hoi/jo), in which (3 and y stand for f, is accepted, it can have nothing to do with a Homeric Xw4/3as. Studies in Greek Noun-Formation 55 Xewc- in X««6s, etc. (cf. &/*$tX&i«7, etc.), with &, which is paralleled in 'AX/ca-0oos, Horn, dvpaupds. &Xl|8as is a rare expression for 'corpse,' also used for 'vinegar' ('dead wine'). 1 The popular etymology of the ancients (cf. the quotations from Plutarch) is hardly the true one, and takes no account of the pr-inflection. The probable analysis is aXt-/3cu>T-, even though the first element may remain obscure. The supposed evidence for initial long vowel (L. & S., cf. Lobeck, Prol. 289) is illusory (cf. Schneider, Callim. 2. 253). Without such specific evidence, or any reason to suppose the word is not Attic-Ionic, short a is more probable, and this is implied by the ancient popu- lar etymology. The latter also supports the traditional smooth breathing, although a rival etymology is given in Et. Mag.: dXi/Sdwras - tovs h> flaXiWjj rekevrricravTas. On the whole perhaps the most plausible connection is with ^Xl/Saros in the sense of 'deep, abysmal' (&vtp<# h> ^Xi/3aT, etc.), as suggested by Adam, » Republic of Plato 1. 132. Only we should not assume ciXi-=ijXi-, but dXt- beside ^Xi-, with vowel gradation. eX6£as, in Homer, Hesiod, Pindar used only in the sense of 'ivory,' from Herodotus on for the 'elephant,' is believed to contain an element which is seen also in Latin ebur and which in both cases is borrowed from Egyptian or Semitic (cf. Egypt, ab 'ivory,' etc. See Schrader Reallexicon 180, and for the preceding e\- the sug- gestion of Osthoff , Parerga 281) . There is nothing in the foreign origin to explain the stem in -avr-, which must be due to the analogy of some Greek word, very likely a8&p,as, which the ivory rivaled in hardness. The Sicilian coin-names Stffis, rpias, rpi|Sj, rerpas, O-as, 2 are formed on the model of Italic denominations of the type represented 1 Plat. Rep. 387c- obKovv in Kai rd wepi ravra opdpara k&vto. to. 5«vd re ko.1 tJM^epi. 6,*of3\TiTea., Kiciarrofe re xal arbyas xal tvipovs xal iMfSavras, xai SXXa &ra Toiirov tov rinrov ovop.a^bp.eva (pplrrtiv Sij woiei Trdiras robs d/tofocras- Plut. Mor. 736a; 6 8' dX{(3as nal 8 6i>Tas 'iXlfiavras' KaXoOo-ic dXifSavra. Trlvovrts (=oi dXJ/Savra wivovres. Cf Schneider, CaUim. 2. 253). Orion 30. 15 (says word occurs in Hipponax); Suidas; Et. Mag.; Eust 1237. 22, 1679. 33; Schol. Aristoph. Frogs 194; Hdn. 2. 656. 5. 8 Poll. 4. 174-5 and 9. 81, in both cases quoting from Aristotle. 4. 174-5. ol ZuceXturai robs iiiv Sbo xaXKoCs Sit-avra (MSS 5if apra, Stjayra, i^avra. Bethe adopts 56 Carl D. Buck by Lat. quadrans, sextans, etc. This is so clearly in accord with other evidence of the early introduction in Sicily of an Italic coinage system (cf. vovixpos, ovyda, and Xirpet from Hlpra, whence Lat. libra), that independent derivation of rerpfis from *n.rpa-Fivr- (Brugmann-Thumb 235) is most unlikely. ireXeicSj 'woodpecker' has jr-inflection, after the analogy of participles, in Aristoph. Birds (irtktKavTL 889, ■wekeK.a.VTts 1155; cf. partic. TrekenavTwv 1157). The word belongs properly with other bird-names, eXeas, ^XcwSs, /3ao-/cas, etc. (dat. sg. eXep, etc., Birds 885), and so to the colloquial type of nicknames like $ayas, rpeaas, xeaas, etc., which are not j'r-stems (xecas, x^o'Sj'ros in L. & S. is an error). Tr&XXas, waXKavros Veos,' attested only by Eustathius (84. 42, 1419. 50, 1742. 37), is if genuine a variant form of TrdXXaf (cf. Mod. Gr. iraWrjKo.pi. 'young fellow, warrior'). 4. Personal names. — The largest distinct group consists of the compounds in -oa/xas. Several of these are most familiar as names of legendary heroes, namely the Homeric 'A-, Apx/>i-, Evpv-, 'Iwwo-, Aao-, Uov\v-, XepovSap-as. But these, all but the last, occur also, with many others, in all over 35, as historical names. Only a few of these, as 'Iiriro-, Aao- (Aa-, Aeco-, Aeo-), IIoXu- (IIouXu-) Sa/xas, are at all common, but examples of the type are widely distributed. 1 On the other hand, the compound names in -as (-Qav-r-, from the first, which is inexplicable) KaKovcn, too &' iva obyniav, robs Si Tpeis rptavra. 9. 81. kclI Stiavra, oVep 'tari Sin xaX/toT, nal Tpit-apra, Sxep rpeis. The same coin is called rpias in the first passage, Tpi£as in the second. The latter is formed like St^as (cf. 5i£os, Tpi£6s = Surabs, rpurcrbs). The former occurs in Hesych. rpmpros irbpvrf Aayu/SApowa rpiavra & iarl \«rrd eucoo-i. rerpas Hesych. (coin and 'quadrant,' in latter sense also Vitruv.). i£as Hesych. and' Hdn. 1. 54. 17 (where the accentuation i£as is prescribed). 1 Miss Macurdy, JHS. 39. 64, comments on the overwhelmingly Trojan character of the Homeric names in -Sayuas, with the exception of 'ApufrtSanas, and quotes similar nameB of Dardanians, Thracians, and Thessalians, as part of the evidence of "coin- cidence between the typical Trojan names and those which are found most commonly in the ruling tribes of the northern part of the Balkan peninsula." The general thesis is of great interest and importance, and some of the evidence may not be illu- sory. But in this case (and likewise for the names beginning with Aao-) I can see no ground for asserting that the names were especially characteristic of the northern regions. For in the historical period names in -Sdp.as, only a few of which, as nouXuoayuas in this form, are clearly due to epic influence, are quotable from all parts of the Greek world, from Thessaly to Laconia, from Corcyra to Crete and Rhodes. Studies in Gbeek Noun-Formation 57 (fralvco, above, p. 52) are legendary only, at least Ilepi^as, "Ew^as, 'Tirepas, and probably also IIoXiK^aj on a Corinthian vase (SGDI. 3126). Likewise "ArXas, 'A/cd//as. Of the numerous uncompounded names like Aias, Bias, 06as, $eidas, etc., the great majority are legendary or foreign. Names which are derived from quotable verbs in -aiva, like Ilepi^as, etc., are: Aeijuas, 'Epi/3as (ftaivco; cf. 'Epi-rijuos, etc.), KuSas, Ke'pSas, KaXxas, O'iSas, 'Ovofias, Heipas. For Hpopalvci} like evpalvo3, and for some others actual verbs in -aivoi may have once existed. But, the type once started, such names were formed from other verbs, and from nouns, adjec- tives, or even adverbs, e.g., el5as (ei8o juat), Bias (jSta), 66as {dobs), Ei5as, epithet of Dionysus (evai). 1 A few names are merely back-formations from place-names with vt or v, e.g., "Apas, Tevdpas, reputed founders of 'Apavrla, Tevdpavia. Zds, Zd^ros, Hdn. 2. 635. 26, etc., is the late hyper-Doric form of Zijs, Zeis, with rr-inflection supplied, probably by the grammarians only, after the analogy of other names in -as, -avros. Apart from the names in -dafias, only about a dozen in -as, -avros are quotable as names of Greeks in the historical period. KvSavs (see above, p. 52) and 'Eviwas (cf. eviiri), evlwTui) are common names in Crete (SGDI. IV, pp. 1179, 1184), where also once BdXXas ('Ap X . 'E0. 1908, 230; nom. only, but Hdn. 2. 652. 34 gives gen. BdXXapros). There are a few scattered examples of Ooas, from Boeotia (IG. 7. 1092), Laconia (IG. 5. 1. 20 B 4), Aetolia (SGDI. 1730), Termessus (CIG. 4363, 4366t). Further, a Delphian 'Ept/3as (SGDI. 2757. 3; see above), a Spartan 'Ovbp.as (Arr. An. 3. 24. 4), a Macedonian KdXXas (Diod. 17. 7. 17, etc.), and Ai>p,as, Bpuas, 'Adafias in a list of "known Pythagoreans" (Iambi, vit. Pyth. 36). 'EXe^as occurs as a personal name (cf. AeW, Apancov, etc.) in IG. 4. 1484. 102, IG. 5. 1. 699, BCH. 29. 102, No. 3 (cf. also Polyb. 18. 24. 2). 'Apuearas is a well-attested Athenian name (cf. Kirchner, Att. Prosop. Nos. 2206-8; cf. also 'ApKias, 'P^/Jas, SdXYas, TevOpas, or mythical, as 'AXi>/3as, Tpuxas, Ilpas, B&ftpas, 'Ataxias. 3 1 Also accented rXiaos, but cf. Hdn. 1. 50. 18. For IIpAs, not Ilpas, cf. Hdn. 1. 399. 16. On the other hand, not TLb-pas (Ptol.), but nVrpas, belonging with Horn. jr«TpjJ««. 2 Tiflpas St. Byz., but Attic inscriptions have regularly 1ei.Spa.aioi. * 'Aptios Erastosthenes ap. Hdn. 2. 650. 10 is another place-name, of unknown locality. Studies in Greek Noun-Formation 59 'An&nas, 'Airiaas, "ArXas, Mijuas, 6pj8as are identical with the names of legendary heroes. KdXXas, a stream in Euboea, is presum- ably connected with /cdXXos. 'Aop/?as [p. 59. Kvp/Jas, Kvp/Jdvres, see Kopv/Jas, Kopv/3avT« Apicr/Jas ©iV/Jas Hdn. 2. 651. 1. 'AXv/3as [p. 58. 'Apv/Jas Kopv/3as Kopu/Javrts [p. 58. Axpdyas [p. 58. IYyavres [pp. 52, 58. yiyas Aesch., Hesych. [p. 52. a~vy- Byz. rpir Orph. rpaire^o- Byz. aivo- Nonn. dvSpo- Callim. BpiyavTes [p. 58. SdXyas Hdn. 2. 651. 23 [p. 58. (Zvyavm) [p. 58. CTSas) [p. 59. 3>a8as [p. 57. 'A<(!>eiSas [p. 58. 'A0ds) [p. 59. Aias [pp. 57, 59. Bias [p. 57. Aias ©«as niKiavrcs [p. 58. Aotds Iloias Ilapias dvSpids Pind.+[p. 53. IlpiixS Tpias Poll., Hesych. [p. 55. HtvKtTiavres [p. 58. A<£ids ireAexas Aristoph. [p. 56. IleAcKas (Mapi/cas) [p. 59. rdXas Hipponax [p. 51. aivo- Callim. [p. 51. AtiXas X^Xas Hdn. 2. 652. 19. ('Ap.v/cXas) [p. 59. BdXXas Hdn. 2. 652. 34, 'Ap X . "E- Delphin. in Milet. 138. 69. 'Eppo- Hevo IG. 11. 4. 587. 2. to£o- Aesch. Iiriro- Aa/gpo- Melanges gr6co-rom. 1. 441. 3. ctvSpo- Pind., Pliny. 'Avopo- ®t]pO- apyvpo- Plin. 'Io-o- IG. 12. 7. 83. AeovTo- Pind. 'Apurro- Airo- SGDI. 1682. 40. Ei- Ei0t>- IIoAv-, IIovAv- -oapas — Eupv- ©pao-v- Inscr. Pont. Eux. 1. 11. 3. 'Ao-rv- Aeco-, see Aao- 2c«- tt/cap-tts- Hom.+[p. 51. 'A/capas [p. 59. v A0apas [p. 57. Aifldpas Mnemos. 42. 332 [p. 59. Mapas Tapapavres, rapi'pavTes [p. 58. Ipxis Horn. + [p. 52. Aapas [p. 57. Mipas [p. 59. Bpipas Hdn. 2. 653. 22. TdApas Hdn. 2. 653. 19. 'OSdpxivTes [p. 58. 'Ovdpas [pp. 52, 57. Xappas ©avpas Avpas [p. 57. 'A/a