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. Stbatton, and prepared under the supervision of Cakl D. Buck 1 DENTAL TERMINATIONS I. 1 Neuters with Genitive in -otos (Exclusive of -/ia, -narcs) By Carl D. Buck The great mass of neuters with genitive in -aros consists of those in -fia, -juaros, which have already been discussed and listed, in con- nection with other /i-suffixes, by Stratton (Chicago) Studies in Classical Philology 2. 115 ff. They are clearly the Greek repre- sentatives of the IE. neuter men-stems, as in opo/xa = Lat. nomen, Skt. ndma, etc., however opinions may differ as to the source of the T-inflection, for which see references in Brugmann-Thumb, Griech. Gram. 233. The r-inflection also represents a neuter n-stem in the case of words in -ap or -ap, gen. -aros, and in some more isolated words such as ovs, ovaros, Kapr}, tcpaaros, Krearecrcn. From these words reflecting neuter n-stems, especially the over- whelmingly dominant class in -p.a, -fiaros (over 1,000 in number), the r-inflection spread at the expense of the minor classes of third- declension neuters, namely, to ybvv and dopv (already in Homer), to those in -as (post-Homeric), even to some in -os, and to a few others. -ap or -ap, gen. -aros This class reflects a well-defined composite type (nom.-acc. in -r, other cases from an n-stem), which must have become established in the parent speech. Cf. Brugmann, Grundriss 2 2. 1. 578 ff.; also Fraenkel, KZ. 42. 114 ff.; Schwyzer, KZ. 46. 165 ff. Specific corre- spondence is seen in: fjirap, tfiraros with Skt. ydkrt, gen. yaknds, Avest. ydkars, Lat. iecur, iecineris (based on Hecinis, like feminis to femur); ovdap, ovdaros with Skt. udhar, gen. udhnds (Lat. uber, 1 See Introductory Note, CP. 5. 323 ff. [Classical Philology XII, January, 1917] 21 22 Carl D. Buck uberis, OE. uder, with generalization of the r); itScop, iSaros with Umbr. utur. abl. une from *udne (cf. also Skt. uda-ka-m, gen. udnds, Goth, waio, watins, OE. wester, etc., with generalization of either the w- or the r-stem). The rare ap (also aKei^a, see below, p. 27), dXei^aroj, from a\eUf>o}, is a specifically Greek derivative, formed after the analogy of some word of related meaning such as irtap. 3 From the semantic point of view the IE. r/n type is too diverse for precise definition, but it comprises mainly words for familiar, concrete objects and divisions of time, the most noticeable feature being the number of words for parts of the body. The following words are from -fap, -Faros: ireipap from *irepFap (cf. Skt. parvan- 'knot, period'); eldap, Hesych. Uap, from *e8Fap to eSa; Att. piap, pia.Tos, Horn, (frpeiara, from *pr)Fap (cf. Arm. albiur 'spring,' Goth, brunna); dekeap from *5e\eFap (cf. SeKevpov, htktTpov), ep. SeTXap from *8e\Fap, Aeol. 0Ar?pfrom *@\r)Fap; a, never -var. It is impossible to determine whether the situation in Sanskrit is secondary, owing to a generalization of the n-stem at the expense of a once existing r-form, or whether it reflects the original habit of the IE. ^en-stems. In the latter case the Avestan r will be due to the analogy of the certainly inherited forms like yakan. Similarly in Greek the -p of ireipap, as contrasted with Skt. parva, may be secondary, after the analogy of 7$7rap, etc., and conversely in one or another word the -fap may be inherited and the -far- analogical. ovtiap (ovqap) is obviously a Greek derivative from ova-, ovr}- (cf. &VTIO-IS, ovaais). But whether it was once *6va-fap, formed after the analogy of others in -Fap before their loss of f, or was at the out- set a hiatus form, as Brugmann, Ber. sachs. Ges. 1913. 201, IF. 35. 96, prefers, cannot be determined. For words which have every appearance of being specifically Greek analogical formations exist both among the -ap forms, e.g., aXe«£ap, and among the -fap forms, e.g., eldap. 1 In general, the type in -p, -aroj, not to distinguish further between -op, -wp, and -fap, enjoyed some degree of productivity in proethnic Greek. Callim. (cf. Sicilian 5es in medical writers. On these last cf. Solmsen, KZ. 34. 7ff., who prefers the derivation from *pias for 4>piap (Choerob. 1. 360. 18) is probably fictitious; yovvas, 8ovpas (Gram. Gr. 2. 407; dovpas also Anth. Pal. 6. 97) are obviously artificial creations to yovvaros, dovparos. But ouas as an analogical substitute for ovs has more claim to genuine- ness. Apart from its occurrence in our text of Simon. 37. 14 Bergk, also Kaibel, Epigr. Gr. 1046. 81, it is given by the grammarians, includ- ing Herodian, e.g., 1. 392. 4, in a list containing no suspicious forms (a few lines before, 1. 391. 27 ff., it is expressly stated that Seas and o-weas, in spite of deaTos, aweaTeaai, are not in use). 1 In Modern Greek, vdap, Ijirap, o$0ap, and aros from Kve4>as. From depas the grammarians give depcos and Mparos, the latter, e.g., Hdn. 1. 401. 11; so perhaps to be retained in Diod. 4. 56. For epvaiirekas the r-inflection is to be inferred from epvo-iweXariidijs, etc., in medical literature. From Kpeas the v 'ArriKU>v avvaipovaa to ao eis co, rod Kpecos, rod yrjpws. 1 Cf. also Mod. Gr. to Kpeara, etc. The same sources give yrjpas, yrjparos, though the latter seems not to be notable otherwise before the Byzantine period. That the encroachment of r-forms become general in later Greek is apparent from the statements of the grammarians, like that just quoted, and also from the situation in Modern Greek, where the few remaining representatives in the spoken language of the old neuters in -as are regularly inflected as r- or ro-stems. Cf . Thumb, Handbook, § 105. 1 The statements of Moeris 336 Koch, ripa koX repdv ' Attikws, rtpara. koX repdrwv 'EXXijwkOs, and of Thom. Mag. 348 Ritschl, ripa ical repCbv 'AttikoI, as Kpia, Kpt&v rtpara Si Kal repdrtov dirXfls"EXXi)i'es, rest upon a similar observation of the difference between Attic and Hellenistic inflection, but are unfortunate in the example chosen, since ripat has T-inflection even in Attic. 26 Cakl D. Buck The case of Kepas is different from the preceding on account of the a in Keparos, 1 etc., which are the forms regularly employed in Attic, except in the military phrase eirl Kipus. According to the view now current tciparos is from *Kepaaaros, which is thought to represent an inherited n-stem extension (see below on Kpaaros, Skt. girsnds), or in any case (as giving it with intervocalic aos the r-forms, coros, etc. (beside aovs, etc.), were obviously formed to the nominative subsequently to its con- traction. There is no trace of *aaros to aos. ra^ari—rfKJiet, seems to occur in pap. Brit. Mus. 1. 102. 556: dpdt w vfias Kara rod iv rco raif/ari. But apart from the xf/ which, if correct, is due to the influence of Baipai, the isolated example of transfer to the r-type at this late date is difficult to credit. 1 The stem Ktpar- is also seen in Keparlvqs and other derivatives, likewise in the rare compound adjective forms, as i^iKipara Pind. fr. 321 Bergk (Ar. Nub. 597), irvpyoK^para Bacch. fr. 51, XevKOictparcs Hesych., xP v < roK ^P aT ' Eur. Hel. 382, dt-vxtpari Phot. s.v. eifojrpvVv Tatipy. Cf. especially Danielsson, Gram, und Etym. Stud. 26 ff. 2 The late epic xepaarot is an artificial imitation of Kpaaros. 3 *XPV 0S is not necessarily from *xPVf°s< s ti u ' ess XP$ aTa from *xp^fara. Cf. Brugmann, Ber. sacks. Ges. 1913. 200. Studies in Greek Notjn-Formation 27 •v, -aros The neuters ybvv and S6pv were originally inflected *yovf-os, *8opf-os, etc., whence Horn, yovvbs, 8ovp6s and Att. (poet.) 5op6s. But this was an isolated type, and after the analogy of the numerous neuters in -aros arose as collateral forms *yovf-aros, *8opf-aros, whence yovvaTos, bohparos in Homer, though much less frequent than yovvbs, SovpSs, and Att. yovaros, Sbparos, the regular prose forms. 1 -a(-7j), -aros and -a, -aros Regarding the complicated and much-discussed 2 inflection of /capo, Horn. K&pr), no doubt exists that the r-forms reflect an n-stem, as seen in Skt. gtr§an.- 'head,' gen. cir§nds (but nom.-acc. sg. giras like Kepas), no matter whether Horn, updaros, whence Kpards, be taken from *Kpaa (Aesch.-)-) 3 beside &\eiap (Hes., Theoc), dAet^aros (Hom.-f-) has been accepted as an isolated instance of a nom.-acc. in -a from n, parallel to the frequent -p,a from -mn (Skt. -ma, Lat. -men). Cf. J. Schmidt, Pluralbildung 108; Brugmann, Grundristf 2. 1. 311. But on the evidence of other languages an IE. nom.-acc. sg. in -n, exclusive of -mn or -y,n, must have been exceedingly rare if indeed it existed. 4 It is doubly suspicious that the sole alleged representative 1 Any historical connection between *yovfaros, *Sopfaros and the Sanskrit «-stem forms to neuter w-stems as gen. jdnunas, darunas to janu, daru, as is assumed by Peder- sen, KZ. 32. 253, is wholly improbable. 2 Danielsson, Gram, und Etym. Stud. 1 ff . ; J. Schmidt, Pluralbild. d. idg. Neutra 363 ff . ; Brugmann, IF. 18. 428 ff . ; Ehrlich KZ. 38. 86 ff. 3 To the references in Liddell and Scott add the important SGDI. 5495. 34. In Hes. Th. 553 the preferred MSS reading is SXeupap. * Certainly most neuter n-stems, other than those with the suffixes -men or -uen, belonged to the composite type, with nom.-acc. sg. supplied by a different stem. In Sanskrit only those in -man and in -van have nom. -ace. sg. in -a, never those in an (cf . Lanman, Noun-Inflection in the Veda 530) . There are no forms pointing to -n in Germanic or Balto-Slavic (OPruss. wundan 'water' is an o-stem form). In Latin, outside of those in -men, and inguen, unguen, early sanguen, there are only pollen, a word of doubtful history and in this form quotable only from the grammarians (cf. Stolz, Lat. GrA 181 with references to discussions) , gluten, which might easily owe its form to the analogy of unguen, and the late and obviously secondary turben and circen. 28 Carl D. Buck of such a form in Greek, apart from the hypothetical *Kapaaa, should be a word which is so plainly a specifically Greek derivative as is akeufra from dXet^co. I prefer to believe that aheupa, beside aKeuf>ap for which see above, p. 22, either owes its form to the analogy of some form in -jua, 1 perhaps aXetjujua, though this is not quotable until later, or is in origin an o-stem plurale tantum. The names of the letters like a\p, o-Karos Sophron (emend.), gram. [p. 22. .Uas, -aros N.T.+ [p. 25. rump, -aros Hom.+ [p. 25. yoVu, -aTos Hom.-|-[p. 27. rprap, -aTOS Hom.+[p. 21. irpoo-wwov, -olto. Horn. [p. 28. ycpas, -dTa Kaibel Epig. Gr. 1046. 29 [p. 26. Stpas, -aTos Diod.(?), Hdn. [p. 25. Kepas, -aTos Aesch.+ [p. 26. -Ktpar- in adj. cpds [p. 26. 1 itf/i- Pind. rrvpyo- Bacch. kcvKo- Hesych. Xpvo-v- Eur. 6£v- Phot, n-epas, see irapap. Tcpas, -aros Hdt. + [p. 25. y?pas, -otos Hdn.+[p. 25. ovapov, -aTa Hom.+[p. 28. irapap, n-etpas, 7r«pas, -aTos Hom.+ [pp. 22, 24. Sdpv, -aTos Horn. + [p. 27. Se\.Ta, -aTos Schol. Dion. Thrax [p. 28. Orjra, -aTos ibid. [p. 28. oSs (2»s, ovas), -otos Horn. + [p. 24. kv{