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. A SEMANTIC NOTE By Carl D. Buck A Greek boy of the present day, who had lost a pet animal through neglect, might say: Qixa-va- va to Ta.i\, k' e^o^cre, 'I forgot to feed it and it died.' The sentence would not be understood by one conversant only with ancient Greek, yet it contains no foreign ele- ment, but only words sprung from ancient Greek material. The source of e£exar](re, was briefly indicated long since by the first critical student of the modem Greek vocabulary, Koraes. It is the purpose of this note to review their semantic history, noticing some parallels, and in connection therewith some of the many different sources of the notions 'forget' and 'die' or 'kill.' &xavw or &xv&, aor. i^exacra, 'forget.' The ancient Greek expressions for 'forget/ \avdkvonai, Aijflojucti, euX^0ojuai, etc., were first displaced by an allied secondary derivative Xijer/xoj'w (cf . \r)riv ixopyyovvT ! or CIG. 4697. 15, crvvTa^eis cnruias re /cat apyvptKO-s. The history of rayr) may now be carried farther back, through the evidence of the papyri. It is used of a fixed quantity of provisions, e.g., Ox. Pap. viii. 1139. 3 (fourth century a.d.), d&re 'kpnabU? iKcrKeirTopi Xax&vcav rayijv niav; Berl. griech. Urkunden 1186. 16 (23/22 B.C.) icpdffov yXwcews rayas X'. A Semantic Note 41 In another passage rayq occurs in the sense of 'allotment, assign- ment,' not referring to food. At least this interpretation seems to me clear, though the word is not so understood in the only translation of the passage which I know of. In a letter regarding a dispute with workmen in a quarry, Flinders Petrie Pap. ii. 4. 2 (third century B.C.), we read: eypa\f/a o4>3> 'die.' The regular Modern Greek word for 'die' is airedava in the aorist, with a new analogical present airodaivo) or iredaivw in place of the ancient aToOvfono). But ipocfxa, in ancient Greek meaning 'make an inarticulate noise,' came to be used colloquially in the sense of 'die,' especially of animals or of men dying miserably, as from starvation. This use is attested for the twelfth century at least (Prodromus, I, 317, \f/bovi> etc tt\v vetvav 'I was dying of hunger'), and is doubtless much earlier. Koraes, Atakta, I, 264 ff., connected this use with the noise made by the body falling in death, comparing the Homeric use of 5ovtt£o) to denote the dull thud of the corpse, e.g., SeSovroros OidiroSao 'when Oedipus had fallen.' But it is rather the inarticulate gasp of death that furnishes the transition, for which the closest parallel is the slang croak in the sense of 'die.' Compare also the use in the sense of 'die,' and with the same application and tone as \j/o&, of Fr. crever and Ital. crepare, whence NHG. krepieren. While this use may of course be derived from the usual meaning of the French and Italian verbs, namely 'burst,' it more probably represents an old colloquial expression which, like croak, grew out of the notion of noise that was dominant in the Latin verb (crepare 'crack, creak, rattle,' etc.). Another association underlying words pertaining to death is illustrated by Mod. Gk. (tkot&vo) 'kill,' from ancient 6 tx0<>s M*"<« ^ naKurri. 7« tyov (coXttffflot, Plut. Mot. 667 F), and the colloquial form inpi-ptov displaced ixMs- Hence Mod. Gk. ^Apt 'fish.' A Semantic Note 43 blind' ('kill' already in Byzantine writers), from us 'death,' which probably comes from the root *dheu- through the medium of the notion of 'mist' or the like (cf. Skt. dhuma- 'smoke, vapor,' Latin fumus, etc.). In NHG. sterben, OE. steorfan 'die' (now starve) the underlying meaning was the 'stiffness' of death. Cf. NHG. starr, Gk. (rrtpifrvios (Hesych.), arepeds, etc. A more offensive allusion to the appearance of the corpse is the vulgar 'turned up his toes.' OE. cwelan 'die,' whence also cwellan 'kill' (now quell), belongs with OE. cwealm 'pain' and 'death' (now only the mild qualm), OHG. quelan 'suffer pain,' quala 'pain' and sometimes 'death,' ON. kvelja 'cause to suffer,' Lith. gela 'pain,' gelti 'ache,' Olr. at-bail 'dies.' In this group, in which the meaning fluctuates between 'pain' and death,' it is probable that the former meaning is the earlier, from which arose 'death' as the climax of 'pain.' 1 To 'die' may be, euphemistically expressed, to 'depart, pass away,' etc., as in Eng. pass away of reverent, or pass out of irrev- erent, speech. So Lat. pereo, Eng. perish. Ir. bds 'death' and related verbal forms are perhaps cognate with Skt. ga- 'go,' Gk. ejfop, etc. (so Pedersen, op. cit., II, 458, 461; otherwise Stokes, Fick II, 159). Avest. irid- 'die' (partic. irista- 'dead,' the blessed dead, the departed), used, in contrast to miryeite (below), only of good beings, is cognate with Goth, -leipan, OE. lipan 'go.' Cf. also Lat. inter eo 'go into the midst of (get between, be cut off from view), whence 'perish, die'; likewise intercido 'fall between, perish.' And of still other euphemistic expressions which may be current, like Gk. cbraXXao-, NHG. quellen, Skt. galaii. If there is any such relationship, a more plausible semantic connection is suggested by the intransitive use and its evolution in Skt. gal- ' drip, drop, vanish, pass away. See Wood, "IE. a* 99," Mod. Lang. Notes, XXXIV, 205. 44 Carl D. Buck already in the parent speech and is beyond the reach of any certain further semantic analysis. 1 Thus Skt. mriyate, Avest. miryeite, OPers. amariyata, Mod. Pers. murdan, Arm. meranim, Lith. mirti, Lett, mirt, ChSl. mreti, and so in all the Slavic languages (Russ. y-meret, etc.), Lat. morior and so in all the Romance languages (Fr. mourir, etc.). In the Celtic, Greek, and Germanic branches other expressions for 'die' displaced this root, which survived only in derivatives, as Olr. marb 'dead,' Welsh marw 'dead' (whence also marw 'die'), Gk. j8por6s 'mortal,' Goth, maurpr, OHG. mord, OE. morp, morpor, Eng. murder, etc. The Germanic forms, originally equivalent to Lat. mors, developed a pejorative sense and were used of unjustifiable, especially secret homicide, murder. To 'kill,' as the causative pendant to 'die,' is often expressed by a derivative of the word for 'die,' 'death,' or 'dead.' Thus NHG. toten, Dutch dooden, Swed. doda, Goth, af-daupjan; OE. cwellan 'kill,' .causative of cwelan 'die'; Skt. marayati 'kills,' causative of mriyate 'dies'; Lith. zudyti 'kill,' causative of zuti 'perish' (though this is not the regular word for 'die'); Lat. interficio 'slay, kill,' which is virtually a causative of intereo 'perish'; Olr. marbaim 'kill,' from marb 'dead'; OPruss. gallintwey 'kill,' from gallan 'death'; likewise Mod. Gk. o-kotoww, Eng. murder, etc., previously mentioned. But many of the words for 'kill' are unrelated to those for 'die,' and go back to the more general meaning 'strike, smite' or the like, as in Eng. slay in its present restricted meaning compared with that of OE. slean or NHG. schlagen. Eng. kill goes back to a form which in its earliest occurrences means 'strike, hit,' as ofte me hine culde 'often one struck him.' Skt. han- means 'strike,' but is also the commonest word for 'kill' (likewise Avest. and OPers. jan-), while in the Greek cognates the meaning 'strike' prevails in deivu, but 'kill' in hreve, ir&j>a.T6vos, etc.; cf. also, from the same root, Olr. gonim 'wound, kill.' Gk. urdvui, biroKreivu, the latter the regular prose word for 'kill,' are cognate with Skt. ksan- 'hurt, wound, destroy.' 1 Though ultimate identity with a root meaning 'pound, crush,' etc., is commonly assumed and is probable. A Semantic Note 45 Compare also Mod. Pers. kushtan 'kill,' cognate with Skt. kus- 'tear, gnaw'; Lett, no-kaut 'kill,' from kaut 'strike'; Dan. draebe 'kill,' cf. ON. drepa, OE. drepan 'strike' (sometimes also 'kill'); Welsh lladd 'kill' (also 'cut, strike'), Bret, laza 'kill,' cognate with Olr. slaidim 'smite'; ChSl. iz-biti, u-biti 'kill,' from biti 'strike,' and similarly Russ. ubyt, Serbo-Croat, ubiti, Boh.z ahiti, Pol. zabic. Similarly Lat. caedo 'cut, strike,' also 'kill,' and occldo regularly 'kill,' whence Ital. uccidere, Roum. ucide, OFr. ocire. This general Romance word for 'kill' has been replaced by malar in Spanish and Portuguese, and by tuer in French. The derivation of the former from. Lat. mactare 'slaughter, kill' (Diez) is at variance with the Spanish development of Lat. ct, and the word is now believed to be from the Persian (originally Arabic) mat ' dead,' which became famil- iar in Europe, with the introduction of chess, in the phrase shah mat 'the king is dead' (whence Fr. Schec et mat, Eng. checkmate). Fr. tuer was derived by Diez, starting from the phrase tuer le feu, from Lat. tutare, through such stages as 'protect,' 'cover' (the fire) and so 'quench' it, then generally 'destroy/ with which one might compare the occasional figurative use of Eng. snuff out. But there is no reason to believe that the use of tuer in tuer le feu is anything but a secondary one; cf. Lat. aquae fiammas necant, Pliny, Eng. kill the engine, deaden the sound, Ital. smorzare 'quench, allay,' etc. The use of tuer in the sense of 'kill' (a person) is quotable from the twelfth century at least, as my colleague Professor Jenkins informs me, referring to Vie St. Georges, 427, 1281. Connection with Lat. tundo 'strike' (so Littre, deriving tuer from Hudere or tuditare) is obviously the most plausible on semantic grounds, and I am convinced that tuer is actually derived from tuditare despite the phonetic difficulty involved in Fr. w=Lat. H. University op Chicaqo