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. Classical Philology Volume XIV "January IQIQ Number i WORDS FOR 'BATTLE,' 'WAR,' 'ARMY,' AND 'SOLDIER' By Carl D. Buck In the midst of a world at war, even the plodding scholar finds it difficult to fix his attention upon his own proper field of research, unless it is upon a part -of it which bears some relation to war. Such points of contact obtrude themselves upon the historian, economist, sociologist, psychologist, philosopher, mathematician, chemist, physi- cist, and in every branch of medical science. So the philologist, unable to vie with his colleagues in contributions of practical value, may ponder innocuously on the history of the words which are on everyone's lips. Every change in the instruments and practices of warfare leaves its mark in language, and many of the new terms pass from one language to another and become international. The additions to vocabulary resulting from the present war bid fair to exceed those of any preceding period, not excepting that following the invention of gunpowder. However, it is not my intention to attempt to gather this crop before it is ripe, but rather to survey the history of the words (in the Indo-European languages) for a few of the most general concepts connected with war, such as are common to all periods, namely 'battle,' 'war,' 'army,' and 'soldier.' Such a survey must not be expected to yield important contribu- tions to philosophic thought on the subject of war, at least in my hands. No doubt an untrammeled imagination might readily find [Classical Philology XIV, January, 1919] 1 2 Cabl D. Buck in the history of these words an appropriate text for many a generali- zation on the causes and conditions of war, on militarism versus pacificism, etc. One might demonstrate anew the universality of war. Or, noting the frequent identity of 'army' with 'people,' one might observe that the conception of an army as the 'people in arms' is alike the most primitive and the most recent, contrasted with that of the purely professional army which prevailed in the long interim. Or again, one might note that 'war' sometimes is derived from 'judicial strife' and hopefully foretell the time when it will revert to that conception and when 'soldier' will mean nothing more than 'policeman,' like the French gendarme. But this survey does illustrate, quite superfluously it is true, the international character of war. For the chief interest in the study of this particular group of words lies in the extensive wanderings which many of them exhibit. Thus Latin has furnished one word for 'army' which in its various forms is now current from the Straits of Gibraltar to the banks of the Ganges (42), and another which is familiar in all Europe (44). A late Latin word for 'soldier' has spread over nearly all Europe and America, and the special Italian form of it over most of continental Europe and to remote parts of Asia (59). Germanic has given the word for 'war' to most of the Romance languages and English, though this word is not so used in the Germanic languages themselves (33). Conversely a Modern German word for 'battle' is of Latin origin, though the corresponding form is not so used in any of the Romance languages (15). 'battle' 1-12. From 'fight, strike,' etc. A battle is simple an organized fight, and the majority of the usual words for 'battle' are connected with verbs meaning 'fight, strike,' or the like. That this meaning itself is often secondary, coming from other more specific notions not always determinable, need not disturb our immediate grouping. 1. Skt-. yudh- 'fight, battle,' from yudh- 'to fight.' The root in this meaning is mainly Indo-Iranian (cf. Avest. yuidyeiti 'fights'), and is most prolific of derivatives in Indie (e.g., yodha 'battle' in modern Bengali and Mahratti, ydd 'battle' in Kashmiri). But it appears also in Homeric ianivq 'battle,' and probably in various Words for 'Battle,' 'War,' 'Army,' and 'Soldier' 3 Celtic proper names, as Old Welsh Iud-nerth, Iud-mael, etc., and in Old Irish idnae 'weapons' (Stokes, Urkelt. Sprachschatz, p. 224; Pedersen, Gram. d. kelt. Sprachen, I, 65). The cognates in other languages, as Lat. iubeo, Lith. judinu 'move, shake,' Lett, jauda 'might/ etc., show other meanings, which may be combined with 'fight' on the basis of 'be or set in violent motion.' 2. Avest. pzsana-, the word for 'battle,' once also psrat-, from parat- 'to fight'; OPers. partara- 'foe' or possibly 'battle'; 1 Skt. prt- 'battle' and prtana- 'battle,' and also used as a collective ('fighting body') of a specific type of army organization. The relationship of Lat. proelium (cf. Walde, Lat. et. Wth. 2 , s.v.) is very doubtful. 3. Gk. ixaxo, the regular word for 'battle' from Homer to the present day, from fidxonai 'fight,' of which the further connection is uncertain. For various possible combinations, see Boisacq, Did. klym. de la langue grecque, p. 616. 4. Lat. pugna is, as first pointed out by Breal {Mem. Soc. Ling., IV, 82), a back-formation from pugno, pugnare, which is itself a denominative from pugnus 'fist.' The semantic extension of fist fighting to fighting in general was effected in the verb. In late vulgar Latin pugna was displaced by the following. 5. The general Romance word for 'battle,' Ital. battaglia, Roum. b&talie, 2 Span, batalla, Port, batalha, Fr. bataille, whence also Eng. battle, is from Lat. batt{u)alia, a sporting term applied to boxing and fencing exercises, 3 and derived from battuo 'strike, beat,' occurring mainly as a boxing term. The verb is perhaps of Celtic origin, at any rate related to a group of Celtic words (cf. Walde, Lat. et. Wtb? with references, and Thurneysen in Thesaurus), and also to OE. 1 New reading partaram (prim) established by Weissbach for former hamaram in Naksh-i Rustam a 47. Weissbach, Keilinschriftm- der Achaemeniden, translates partaram patiyajata "hat den kaempfenden (Feind) gesohlagen" but notes that the corresponding phrase in the Elamitic and Babylonian versions means "hat Schlachten geliefert." It is possible that the Persian also means "waged battle," though the regular word for 'battle' is hamarana- (13). 2 But Roum. b&talie is a modern loan word from the French or Italian, replacing in this sense a Slavic word (23). The inherited Roumanian form bdtaie means 'blow, beating,' etc. 8 Gram. Lat., VII, 178: battualia, quae vulgo battalia dicunlur .... exercitionet aulem militum vel gladiatorum significat. 4 Cabl D. Buck beadu 'battle, war' (in poetry only), ON. bgd 'battle,' OHG. Batu- in proper names. 6. OE. gefeoht 'fight, battle,' the usual prose word, 1 from feohtan 'fight' (OHG. fehtan, NHG. fechten, etc.). Probably related to Lat. pecto, Lith. peszti 'pull hair'; cf. the use of Lith. pesztuves and pesztynes for 'scuffle, fight.' Connection with Lat. pugnus is less probable (cf. Walde, Lat. et. Wib.*, p. 622). 7. NHG. Schlacht 'battle,' OHG. slahta 'slaughter, fight,' from slahan 'smite'; cf. ON. slag 'blow, slaughter, fight,' OE. gesleht, gesliht 'fight, battle,' etc. As the specific military term, Schlacht is modern, Luther still using it mostly in the sense of slaughter. It is doubtless due to the influence of the German usage that in Danish and Swedish (and not in Icelandic) slag is now the regular word for 'battle.' 8. Goth, waihjo, which translates naxv (II Cor. 7:5), 2 and wigans(l) in du wigana=els iroXetwv (Luke 14:31), ON. vlg, OHG. vnc, OE. wig (one of the commonest words for 'battle or 'war,' with numerous compounds), all from the root seen in Goth, weihan 'fight,' OE. wigan 'fight,' Olr. fichim 'fight,' Lat. vinco, etc. 9. Lith. muszis 'battle,' from muszti 'strike.' 10. Lett, kauja 'battle,' from kaut 'smite, kill'; cf. Lith. kauti ' smite, fight,' NHG. hauen, Eng. hew, Lat. cudo, etc. 11. The Slavic words for 'battle,' as ChSl. bojl, Russ. boj, Serbo- Croat. b6j, Boh. boj, etc., or Boh. bitva, Pol. bitwa, etc. (cf. Berneker, Slav. et. Wtb., p. 68), are derived from biti 'strike.' 12. Russ. sraienie 'battle,' from srazat', srazit' 'strike down,' ChSl. (su-) raziti 'strike.' 1 In illustrating the wealth of synonymns in Old English poetry, Jespersen, Growth and Structure of the English Language, p. 51, remarks: "For 'battle' or 'fight' we have in Beowulf at least twelve synonymns." The twelve mentioned are: beadu (see 5); gud, cf. ON. gudr, gunnr related to Selva, Skt. han- 'smite,' etc.; hild; cf. OHG. hilta 'battle,' Mir. cellach 'war,' probably related to Lith. kalti 'smite,' Lat. clddes, etc.; nld (see 20); gealyht (see 7); orleg (see 33); vng (see 8); gewinn (see 16); rma 'rush, attack'; lindplega 'swordplay.' Where, however, it has not the military sense. While there is every probability mat Goth, waihjo and weihan were used for n&xv and paxo/tat. in the military sense, and also that harjis was the word for 'army' (see 41), the passages which would determine this are missing. The only Gothic word group quotable in distinctively military application is that comprising gadrauhts 'soldier,' drauhtinon 'wage war,' etc. (see 55). Wobds fob 'Battle,' 'War,' 'Army,' and 'Soldier' 5 13-21. From miscellaneous sources, such as 'meeting,' 'array,' 'field,' 'effort,' 'struggle,' 'action,' 'joy,' 'hate.' 13. OPers. hamarana- 'battle,' constantly used in the Behistun inscription, is identical with Skt. samarana-, samara- 'meeting,' usually 'hostile meeting, fight,' from sam-r- 'come together.' Com- pare the military use of NHG. Treffen, Eng. encounter, etc. 14. Arm. razm 'battle' is borrowed from Pers., Pahl. razm 'battle,' belonging with Avest. rasman- 'battle array,' from raz- 'direct, ar- range,' cognate with Lat. rego, etc. Compare Lat. acies 'battle array,' also 'battle.' 15. NHG. Kampf 'fight, battle,' in the older periods 'duel,' OE. camp 'contest, battle, war,' represent an early Germanic borrowing from Lat. campus. It is true that this derivation was mildly ques- tioned by Grimm (Deutsches Wtb.) and was positively rejected by Kluge in the earlier editions of his Etym. Wtb. der deutschen Sprache (but later restored to favor); and Paul (Deutsches Wtb. 2 ) still says: "Ableitung aus lat. campus ist zweifelhaft." But there is no suffi- cient ground for skepticism, and certainly most scholars today accept the derivation without question. • The connecting link, which Grimm felt was missing, is best seen in the special connotation which came to campus from the Roman Campus Martins. Among the Franks the name Campus Martius, obviously in imi- tation of the Roman use (but also associated with the name of the month, and after a change of date to May, also called Campus Madius) was given to the annual assembly which was "at once an army, a council, and a legal tribunal" {Cambridge Mediaeval History, II, 135). Furthermore, campus alone was used of lesser assemblies of military or judicial character and most frequently for a single combat in the arena, especially the trial at arms to determine the guilty party, like the duel described by Gregory of Tours (X, 10), which was ordered by the king (rex campum deiudicat) to settle a dispute as to guilt. Numerous other examples of campus 'duel,' sometimes expressly glossed by 'duellum' or 'certamen singulare,' are quoted by Du Cange. This is the prevailing meaning of OHG. and MHG. Kampf. It is also reflected in the derivatives, as campio (campiones 'gladia- tores, pugnatores,' glossae Isid.), OHG. chemphio, Fr., Eng. champion, 6 Cabl D. Buck and the verb OHG. chemfan, whence NHG. kaempfen. The broader use of NHG. Kampf is then clearly a modern extension. A similar extension, only much earlier, may be assumed for OE. camp in the sense of 'battle,' and likewise for OFr. champ (cans, chanz). For it seems more probable that here too this meaning has come through the medium of the late campus 'duel' than that it came more directly from campus in the sense of 'battlefield.' 16. OHG. and MHG. strlt, the usual word for 'battle' (cf. NHG. Streitwagen, etc., otherwise Streit now used mainly of verbal strife), also frequently 'war,' cognate with OE. strict 'strife, struggle,' ON. strld 'woe, grief, struggle,' also, by the end of the thirteenth century, 'war,' as in Modern Icelandic. For the semantic development compare also Welsh trin 'labor, trouble, action,' also 'battle' (ultimate root connection with OHG. strlt is probable), NHG. Krieg (see 31), OE. gewinn (see 32), OE- sacu 'strife,' especially 'verbal strife' (Goth, sakjo, etc.), also 'battle' or 'war.' 17. Words meaning 'action' are often applied specifically to mili- tary action, as Eng. action, Fr. action, Gk. epyov (e.g. Xen. Anab. vii. 8. 17). Such a use might easily become the dominant one and yield the regular expression for 'battle.' But I have no example of this. 18. ON. orrosta, the regular word for 'battle' in the Edda, as is Mod. Icel. orusta, is probably to be analyzed as *or-z-osta and con- nected with OE. eornost 'earnestness,' also sometimes 'battle,' from the root seen in Gk. opvvfii 'stir up, incite,' etc. Cf. Per Persson, Bezz. Beitr., XIX, 271, Beitr. zur idg. Wortforschung, pp. 636 ff. 19. Skt. rana- 'joy' is also a common word for 'battle,' the con- necting link being 'joy of battle.' Compare Gk. airj-17 'cry,' specialized to 'battle-cry,' and sub- stantially equivalent to 'battle' in the frequent Homeric phrase avrq re 7ttoX«ju6s re, and in Pindar's dvdvvos o£das avras. 20. Ir. cath, etc. The general Celtic word for 'battle,' lacking only in Breton, is seen in Ir. cath, Welsh cad, Corn, caz, Gall. Catu- rlges, cognate with OE. Heapo-, OHG. Hadu- in proper names. Further connection with Skt. gatru- 'enemy,' Gk. k6tos 'wrath' is probable, in which case the semantic development may well have Words fob 'Battle,' 'War,' 'Abmy,' and 'Soldibb' 7 been 'hatred, enmity, hostility, fight.' Cf. OE. ni£t 'hatred, spite, enmity' (NHG. Neid), also in poetry 'battle' or 'war.' 21. Mod. Pers jang, the usual word for 'battle' and 'war' (likewise in Hindustani) is obviously the same as Mod. Turk, jenk 'battle, war.' There is no trace of this word in earlier Iranian, and the borrowing in this case is probably not from Persian but from Turkish, just as some other Turkish military terms have passed into Persian (and Hindustani), as urdu 'camp.' 1 'wab' War is only a prolonged condition of organized fighting, and the distinction which we make between 'war' and 'battle' is secondary and by no means universal. Several of the words already mentioned are used for 'war' as well as for 'battle' (cf. 6, 8, 16, 21). Of the following words for 'war,' some are known to have been used for 'battle' (cf. 22, 23); some are derivatives of words for 'battle' (cf . 25, 26) ; and most of the others are from semantic sources similar to those which have been noted in the case of the words for 'battle.' 22. Gk. w6\efM>s 'war' was once simply 'fighting, battle,' as pre- vailingly in Homer, e.g. II. iii. 134: "now they rest quietly leaning on their shields, and the battle has ceased." It is connected with TreXejufw 'shake,' ir&KKu 'swing, brandish,' etc. 23. Roum. r&zboiu 'war' meant 'battle' in the earlier language, before the introduction of bdtalie (see 5). It is a Slavic word, ChSl. razboji 'murder, robbery,' a compound of razii- 'dis-' and the root of biti 'strike,' seen in boji 'fight' (11). 24. Albanian Vuftt 'war' is like Roum. luptd, Ital. lolta, Fr. lutte, from Lat. lucta 'wrestling match.' No doubt the Albanian word was once, and perhaps is sometimes still, used for 'battle,' which is, however, commonly expressed by niza, of Turkish origin. 25. Arm. paterazm 'war' (also 'battle') is from Pahl. patrazm, a compound (prefix pati- with vrddhi, as in Avest. paiti-vaka-) of razm 'battle' (14). 1 Horn, Grd. d. iran. Phil., II, 7, does not mention jang in his list of Turkish loan words in Persian. My belief in Turkish origin is based on the lack of Iranian connec- tion. Positive evidence, such as the existence of the word in the remoter languages of the Turkish family, may perhaps be supplied by those to whom more material in this field is accessible. 8 Carl D. Buck 26. Ir. cogadh 'war,' Ir. cocad 'battle,' is a compound (prefix co- 'con-') of cath 'battle' (20). Manx caggey is given as meaning both 'war' and 'battle.' 27. Breton brezel 'war' (cf. Corn, bresel 'strife') is from the root seen in bresa 'bruise,' Olr. brissim 'break, smash,' OE. berstan 'break, burst.' 28. The general Slavic word for 'war,' ChSl., Russ., Boh., etc. vojna, Pol. wojna, is from the root seen in Lith. veju, vyti, Skt. vayati, Avest. vayeiti, all meaning 'pursue, chase, drive.' 29. Skt. samprahara- 'war' is from sam-pra-har- 'hurl, cast, attack." 30. Skt. vigraha- 'separation, discord,' from vi-grabh-, is another frequent word for 'war.' 31. NHG. Krieg 'war' had the more general meaning 'strife' until about the end of the Middle High German period. It belongs with kriegen 'get,' which formerly meant 'strive' (cf. Paul, Deutsches Wtb.). Danish and Swedish krig 'war' are borrowed from German; and Dutch krijg owes its present meaning 'war' to the influence of the German usage. 32. OE. gewinn 'toil, struggle, strife' (cf. winnan 'strive,' Eng. win) is also the commonest prose word for 'war,' as Troiana gewin 'the Trojan war,' Punica gewin 'the Punic wars' in King Alfred's Orosius, where it is regularly used to translate bellum. 33. Ital., Span., Port, guerra, Fr. guerre, whence ME. werre, Eng. war, are borrowed from a Germanic word seen in OHG. werra 'confusion, strife' (cf. NHG. wirren, verwirren). 34. Dutch oorlog, etc. OHG. urliugi, MHG. urliuge, the usual word for 'war' until displaced by Krieg, also (with different grade of the root syllable) MHG. urlouge, MLG. orloge, OFris. orloge, and Dutch oorlog, which in spite of the encroachment of krijg, persists as the usual word for 'war,' have been connected with Goth, liuga 'marriage,' Ir. luige 'oath,' on the basis of a compound meaning' out of compact, breach of compact.' Cf. ON. ufridr 'un-peace, war.' So Torp and Falk, Fick, III 4 , 374, Norw.-Dan. et. Wtb., s.v. orlog. Dan. orlog is borrowed from MLG. orloge and specialized to 'naval war.' 1 1 Span, armada 'naval force,' contrasted with army, etc., is only apparently an example of similar specialization (see 44). Words fob 'Battle,' 'War,' 'Army,' and 'Soldier' 9 OE. orleg 'strife, war,' OS. orlag, etc., though not formally identical with the preceding, is a similar compound, connected with OE. lagu 'law.' 35. Lat. bellum, early duellum, is of uncertain origin. Cf. Walde, Lat. et. Wtb. 2 with references. The combination with Horn. Sat 'in battle,' S^toj 'hostile,' is perhaps the favorite. But the old derivation from duo, though the suffix formation is obscure, is seman- tically possible. For the objection that the earliest recorded 'sense' is 'war,' not 'duel,' 1 is inconclusive. An extension from 'duel' to 'battle,' such as took place in the historical period in the case of MHG. Kampf (see 15), and from 'battle' to 'war,' as in xoXejuoj (22), might have been completed in prehistoric times. Or the development might have been 'separation, discord' (cf. NHG. Zwist from zwei), hence 'war,' as in Skt. vigraha- (30). 36. Welsh rhyfel 'war' is, to me at least, of unknown derivation. For Lith. karas, Lett, karsch 'war, ' see 41. 'army' 37-41. From 'people, multitude,' etc. Many of the words for 'army' mean also, or are cognate with words meaning, 'people, host, band, multitude,' etc. The 'army' was 'the people (in arms),' or 'the (armed) host,' a specialization inherent in the situation in time of war. That the military application is secondary is sometimes plain from the evidence, and in general is the natural and reasonable assumption, unless there is specific evidence to the contrary. This is not to deny, of course, that the movement may at times be in the other direction, as in the case of Eng. host, from OFr. ost, Low Lat. hostis 'army' (43), or just as any word for 'army' may be used figura- tively to denote a great number. 37. The words of the Germanic group represented by Eng. folk, NHG. Volk, while usually meaning 'people,' are also used of an 'armed band' or 'army.' So often ON. folk, OHG. folc, OE. folc, and regularly OE. gefylce. The borrowed ChSl. pltikti is used in both senses, 'people' and 'armed band,' and the military use prevails in Boh. pluk, Russ., Pol. polk 'regiment.' 1 The old form duellum, preserved through the medium of poets, was revived in mediaeval times, under association with duo, in the sense of 'duel.' 10 Carl D. Buck Paul (Deutsches Wtb. 2 ) asserts that Volk "war ursprunglich Bezeichnung einer Kriegerschaar." But the more generally pre- vailing meaning 'people' even in Germanic, the meaning of the cognate Gk. ir\i}6os, Lat. plebs, populus, and their obvious connec- tion with pel-, pie- 'fill,' ought to make it sufficiently clear that the meaning 'crowd, multitude, people' is the earlier, and that of 'armed band, army' secondary. If Lat. popularl is derived from populus, as seems to me most plausible in spite of other suggestions (see Walde, s.v.), it would indicate at the most that populus was also once employed in the specialized sense of 'folk army,' until later replaced by the 'trained army,' exercitus. 38. Mid. Ir. sluag 'band, army,' Mod. Ir. sluagh 'army, multitude, people,' etc., Welsh llu 'throng, host, army,' Com. lu (luu, llu) 'multitude, army.' That the more general meaning 'band, throng' is also the earlier is apparent from the compounds Mid. Ir. teg-lach 'household,' Welsh, Corn, tei-lu 'family,' and the probably cognate ChSl. sluga 'servant' (originally a feminine collective 'familia'). Cf. Pedersen, Gram, der kelt. Sprachen, I, 84. Compare also Welsh byddin 'band, troup' and, besides llu, a common word for 'army,' cognate with Mod. Ir. buiden 'band, troup,' Old Breton bodin which glosses manus in hie Dolopum manus. OE. werod 'throng, band' (from wer 'man') is also used for 'army,' and occasionally, instead of the usual here, glosses exer- citus. 39. Pahlavi gund 'army,' whence were borrowed Arm. gund 'army' and Arab, jund 'army,' is identical with Mod. Pers. ghund 'crowd,' and Kurdish gund, jund 'village.' Although the meaning 'army' is the one quotable from the earlier time, this may be acci- dental. There are no certain cognates outside of Iranian (cf. Horn, Neupers. Etym., p. 179; Hiibschmann, Pers. Stud., p. 83). 40. Avest. spaSa- 'army, host,' OPers. spdda- in Taxma-spdda, Mod. Pers. sipah 'army,' is probably derived from spa-, Skt. cva 'swell,' and meant originally 'crowd, throng.' 41. NHG. Heer, etc. — The most widespread group of cognate words for 'army' is that which comprises the general Germanic word — Goth, harjis, ON. herr, Swed. har, Dan. hwr, OE. here, OFris. here, hiri, OS. heri, MD. here, OHG. hari, heri, NHG. Heer— also Words for 'Battle,' 'War,' 'Army,' and 'Soldier' 11 OPers. kara-, the regular word for 'army' in the Old Persian inscrip- tions, further Lith. karis, karias 'army' (now obsolete), OPrus. kragis, (read kargis, i.e. karjis; cf. carya-woytis 'Heerschau'), with which belong Lith. karas 'war' (kare in Russian Lithuanian), and Lett. karsch 'war.' The Germanic words in the older period are used not only for 'army' but also in the sense of 'host, multitude, people,' e.g. ON. Danskr herr 'Danish people,' etc. In fact Goth, harjis is quotable only in the broader sense, translating once arpand. obp&vios 'the heavenly host' (Luke 2:13), and once X«7€coj' (Luke 8:30), though this may well be accidental, the passages where it would be looked for in the sense of 'army,' e.g. Rev. 9: 16; 19: 19, being lacking. OPers. kara- also in several passages means simply 'people,' e.g., Bh. 1. 10 "the people had no knowledge that Smerdis was slain." In spite of the dominance in this group of the meaning 'army,' the probability is that this is derived from 'multitude, people,' as in the case of the words previously discussed. But the prevailing view is the opposite. 1 42-51. From various sources, as 'trained,' 'armed,' or 'attack- ing' (body), 'enemy,' 'expedition,' 'camp.' 42. Lat. exercitus was originally an abstract meaning 'training,' as in 'pro exercitu gymnastico et palaestrico" (Plaut. Rud. 21. 7). 1 In discussions of the Germanic group (Paul, Deutsches Wtb. 1 , Kluge, Etym. Wtb?, Falk-Torp in Fick, III 4 , and Norw.-Dan. etym. Wtb.) it is assumed (1) that the mean- ing 'army' is more original than 'host, people,' (2) that the Germanic word is a deriva- tive of one meaning 'war,' which (3) is attested by Lith. karas 'war,' this again (4) being cognate with ChSl. kara 'strife.' But, to take up these points in reverse order, ChSl. kara belongs to an extensive group of words used of verbal strife only, especially ' abuse, insult, mockery,' etc., and probably unrelated to the preceding group. Cf. Berneker, Slav. et. Wtb., pp. 578 ff., Walde, Lat. et. Wtb. 1 , s.v. carino. With this eliminated, there is no support for the priority of the meaning 'war.' Further, the notion of an o-stem meaning ' war ' and a derivative jo-stem meaning ' army ' works out for the Lith. karas and the Germanic forms, but, conversely, OPers. kara- means 'army,' not 'war,' and Lett, karsch 'war' is a io-stem =Lith. karias. Plainly the difference in meaning has nothing to do with the variation in stem, and it is simply a question of semantic development from 'war' to 'army,' or from 'army' to 'war.' Analogies may be found for a shift in either direction, but here the probability is certainly that the meaning 'war,' which is found only in Lithuanian and Lettic, is the secondary. Cf. the occasional use of ON. folk in the sense of 'battle' (Cleasby-Vigfusson) , and of OE. here as 'war, devastation.' If now the alleged derivation of the group in question from a word meaning 'war' is rejected no specific reason is left for assuming the priority of the meaning 'army' over the broader use 'multitude, people.' 12 Cabl D. Buck Hence in a concrete and specialized sense 'body of men trained to arms, army.' Borrowed from the Romans in the Orient, probably in Parthian times, exercitus is the source of Arabic c askar 'army' (with transposi- tion of ks to sk, as in al 1 skandar from 'AtefavSpos), 1 and, through the Arabic, of Persian and Hindustani lashkar. In colloquial Latin exercitus became obsolete, being displaced by hostis (43). But it was taken up again from the learned language (see under 43) and became the usual word for 'army' in Spanish and Portuguese (ejercito) and in Italian (esercito). 2 43. Low Lat. hostis, OFr. ost, Roum. oaste, etc. In early medi- aeval documents hostis occurs very frequently in the sense of 'army,' as hostem maximum collegit, cum hoste plurimo abiit, hoste commoto Francorum. See references in Du Cange and the indexes of Monu- menta Historiae Germaniae, especially, Script. Rer. Merov., II, Leges, I, II. Examples are most common in works like the Fredegar Chronicle and the Liber historiae Francorum, from the period between Gregory of Tours and Charles the Great, when a knowledge of literary Latin was virtually extinct. After the first revival of learning in the ninth century and the restoration of book Latin, the proper exercitus reappears, and in the great mass of mediaeval chronicles is the only word used. But hostis must have been the general vulgar Latin word for 'army.' For it is reflected not only in OFr. ost (whence Eng. host 'army' and also 'multitude,' now obsolete in the military sense), Oltal. oste, Span, hueste, Port, hoste, all now obsolete, but also in Roum. oaste, which is still the regular word for 'army,' and in Albanian ushtrl 'army.' By what stages did hostis 'enemy' come to mean 'army'? A transition from 'enemy in arms,' 'hostile army,' to any 'army,' would 1 Cf. Brockelmann, Grd. d. vergl. Gram. d. semit. Sprachen, pp. 269 S. Ignored by Haupt, AOS, XXXVI, 417. I am indebted to my colleague, Dr. Sprengling, for calling my attention to the Latin source of the oriental words. 2 The Italian form (with es, not sc as in scempio from exemplum) bears on its face evidence of learned origin. Not so the Spanish-Portuguese form, which Diez seems to have regarded as a genuine survival ("exercitus erhielt sich nur im Sildwesten"). But here too the older language has only hueste (Port, hoste), as my colleague Professor Pietsch assures me, and ejercito is certainly of learned origin with j for x after the analogy of inherited forms, just as in projimo beside proximo (Hanssen, Span. Gram. p. 60). Words for 'Battle,' 'War,' 'Army,' and 'Soldier' 13 be one's first thought and is taken for granted by some Romance scholars. But what seems more probably the true explanation was given by Diez. Etym. Wfb., I, 229: "Schon in altesten Mlatein bedentet hostis heer oder kriegsdienst; der begriff konnte sich aus der ublichen redensart ire in hostem, gegen den feind, d.i. zum heere gehen, entfaltet haben." The recruiting regulations of the Franks and the Visigoths abound in phrases like convenire in hostem, exire in hostem, proficisci in hostem, in hostem bannitus, etc. Cf. for example, from the capitularia of Charles the Great (Mon. Germ. Hist., Leges II, 166): "Quicumque liber homo in hostem bannitus fuerit et venire contempserit, plenum heribannum, id est solidos sexaginta, persol- vat," while another (p. 153) reads "Deperibanno Qui hostem facere potuit et non fecit, ipsum bannum componate," or from the Leges Visigothum (Mon. Germ. Hist., Leges I, 367): "conpulsores exercitus, quando Gotos in hostem exire conpellunt." The phrases containing in hostem would bear the translation "go forth against the enemy," "summoned against the enemy," etc., but were actually felt as "go forth on military service," "join the army," "summoned to military service," etc., whence also hostem facere, hostis bannitus (OFr. ost bannie), prepositus hostis, in hoste "on military service," "in the army," etc. 44. Fr. armee, Eng. army, etc. Starting from armata classis, current in early mediaeval chronicles, armata alone came to be used for 'naval force,' and also, by extension, for 'land force.' Oltal. armata is used most frequently of a naval force, e.g. I'esercito e V armata 'army and navy,' but also as an equivalent of esercito. Span.-Port. armada has only the former sense. Fr. armee is not quotable before the fourteenth century, and, while the form is French, its use is probably due to Italian influence (though such influence in military terminology is most conspicuous in the sixteenth century, see 59); otherwise it is strange that in earlier French we find, for example, armez 'armed men, soldiers, ' but no trace of the feminine form used substantively for 'armed force.' Fr. arm6e was used for a time of either a naval or a land force and passed into English in both senses. Eng. army (armee, armie) was used of an armed expedition by sea (so in the earliest quotable occurrence, Chaucer, Prol., 59) or by land ("made a great army into 14 Carl D. Buck Scotland," 1502), and of a naval or land force, the application to a naval force not becoming obsolete for some centuries (see NED). Later, Fr. armie, in its present established use, beside being the source of Breton arme, passed into many of the European languages, without, however, displacing the native words, e.g., NHG. Armee, Russ. armija (whence also armija in Russian Lithuanian, p. 15, note). 45. Dutch leger, cognate with NHG. Lager, Eng. lair, etc., and meaning 'resting place,' especially 'camp/ came to be used for 'army' in the sixteenth century, and is now the usual word {heir from MD. here being now obsolete, except in biblical language). Luther sometimes used Lager for 'encamped army,' e.g., Num. 10:5 where our King James version also has "the camps that lie on the east parts shall go forward." In one of the Rhaeto-Roman dialects the regular word for 'army' is (or was) camp, from Lat. campus. So in the Frankfurt New Testa- ment, based on the version of 1648 by Lucius Gabriel of Ilanz, e.g., Rev. 19:19, lur comps, sin comp. OSpan. fonsado, OPort. fossado 'army,' also tyxsavrov 'camp' and 'army' in Byzantine writers, are from Lat. fossatum 'ditch,' whence 'camp protected by a moat.' 46. Gk. (TTparoireSov ' camp ' was also used frequently of an army, e.g., Hdt. i. 76, Xen. Anab. iv. 4. 9, and even of a naval force, e.g., Thuc. i. 117. Cf. (TTpaT6ire8a vavrtxa. Kai we$tKa, Xen. Hell. vi. 3. 8. Gk. arparos the most usual word for 'army' from Homer to the present day, is from the root seen in (xropivvvfu Lat. sternd etc., and meant first the 'encamped army.' 47. OE. fierd, cognate with OSax. fard, OHG. fart, NHG, Fahrt, never means simply 'journey' but rather 'military expedition, campaign,' hence also 'army.' A sentence like Claudius, se casere, fyrde geladde on Breotone, "Claudius, the emperor, led an expedition [or army] into Britain" (Bede's Eccl. Hist, I, 3) illustrates the easy transition. In the Chronicles fierd is the word regularly employed to designate the English army in contrast to the Danish here. But elsewhere fierd is freely used without such distinction of native and hostile army. In early Middle English ferd appears to be more com- mon than h§re, and to have become the usual word, prior to the introduction of armie. Woeds for 'Battle,' 'War,' 'Army,' and 'Soldier' 15 Compare early NHG. Reise often = 'Kriegszug. ' The opposite transition is seen in Gk. Ta£eidioi> (from ra£is in sense of 'body of soldiers'), frequent in Byzantine writers in the sense of 'military expedition,' but now TaftSi, simply 'journey,' with verb ra&Seva 'travel.' 48. Skt. send-, the regular word for 'army' and Avest. haend-, OPers. haind-, which are applied only to a hostile army, are identical with Skt. send-, Avest. haend- 'missile' (cf.also Skt. sdyaka- 'missile'). The 'army' was the 'missile' or 'attack.' Compare the following. 49. The general Slavic word for 'army,' ChSl. vqjsko or vojska, Russ. vojska, Pol. wojsko, Boh. vojsko, Serbo-Croat, and Bulg. vojska, is from the same root meaning 'pursue,' etc., that is seen in vojna 'war' (see 28). Lith. vaiskas 'army' is a Polish loan word. 1 50. Arm. zor, OArm. zaur 'army' is an Iranian loan word. Cf. Avest. zdvar9, Mod. Pers. zor 'power, might.' Compare Eng. force and forces in military application. 51. Lett, kara-spehks 'army' is 'war might.' 'soldier' 52. ' Soldier ' = ' belonging to the army,' derivative (or compound) of a word for 'army.' Skt. sainika, from send (48). Mod. Pers. sipdhl, which passed into Hindustani and thence into the other Indian vernaculars (cf. Anglo-Indian sepoy), from sipdh (40). Mod. Pers. lashkarl from lashkar (42), OArm. zaurakan from zaur (50). Gk. crrpaTtwr^s from arparbs, arpana. (45). Roum. ostai (now obso- lete, replaced by soldat) from oaste (43). Alb. ushtdr, ushetdr from ushtrl (43). OE. here-mann, here-maecg, ON. her-madtr (Mod. Icel. her-maotur), her-mggr, from here (41). 53. ' Soldier ' = 'armed,' derivative of word for 'arms.' Arm. zinvor from zen 'arms, weapons,' which is an Iranian loan word (Avest. zaend-, Pahl. zen); cf. OFr. armez, etc. 54. 'Soldier' = 'warrior' or 'fighter,' derivative from word for 'war' or 'battle' or from the same root. The true Slavic words for 'The Polish loan words vaiskas and zalni'erius (58) are the usual words for 'army' and 'soldier' in texts from Prussian Lithuania. But our Lithuanian news- papers, which reflect the vocabulary of the Russian Lithuanian dialects (and show hosts of words which are bracketed or omitted in Kurschat's Dictionary) use the Russian armija (44) for 'army' and the native hareivis (54) for 'soldier.' 16 Cael D. Buck 'soldier' are derived from vojna 'war,' as vojnik in Russian, etc.; or from the same root, as ChSl. voji (plur.), vojinti, (sing.), Serbo- Croat., Russ. vojin, Boh., Serbo-Croat., Russ., vojak, Pol. wojak. (In Russian and Polish these words are now archaic or poetical, like Eng. warrior, having been replaced in ordinary use by soldat in Russian, iolnierz in Polish, see 58, 59.) OE. wiga ' fighter, warrior ' from wigan ' fight.' Welsh cadwar and OCorn. cadwur, which glosses Lat. miles, belong with Mir. cath-fer 'man of battle, warrior.' Cf. ON. orrosta-madr. Lett, kara-wihrs 'war man,' which is used for 'soldier' in the Lettic New Testament, and Lett, kareiwis, Lith. kareivis, 'one who goes to war' (cpd. of karas 'war' and eiti 'go'), the regular word for 'soldier' in Russian Lithuania. 55. Goth, gadrauhts, which renders orpartcbrT/j (cf. also drauhti- non 'to war,' drauhtinassus, drauhtiwitop 'warfare') is related to OE. drygt 'troop, company, army,' gedreag 'crowd,' OHG truht, 'troop, band,' Lith. draugas 'companion,' drauge 'with,' ChSl. drugti, 'friend,' etc., all having in common the notion of ' company, band ' or ' one of a company.' Specialization in a military sense is partial in the West Germanic forms, complete in the Gothic. 1 The application of gadrauhts to the individual soldier, in place of the collective use, probably started in the plural, just as the plural of OE. dryht means 'men.' 56. Lat. miles is most probably connected with Gk. o/uXos 'crowd,' 6jutXe'w 'consort with,' Skt. mil- 'meet, assemble,' mela- ' meeting, assembly.' One may suppose that the form was used first in the plural, mil-it-es "die haufenweise, scharweise marschieren- den" (so Walde, Lat. et. Wtb. 2 , s.v.). But another possibility is that miles is a secondary derivative, like eques from equus, from a stem *mllo (*meilo- like Skt. mela-) which had come to mean 'hostile meeting, battle.' Cf. OPers. hamarana-, Germ. Treffen (13), and for this particular group, the occasional use of dpikio) in a hostile sense, as dfj.i\ionev Aavaolaw Horn. II. xi. 523, etc. 1 So that they reacted on driugan, which occurs only in the meaning 'wage war.' This belongs with OE. dreogan 'perform, practice' and others of a group which covers a different semantic area than the above, though very likely ultimately related. Cf. Wood, Mod. Phil., V, 271 ff. Words fob 'Battle,' 'War,' 'Army,' and 'Soldier' 17 From Lat. miles are borrowed M.Ir. mil (also cathmil 'battle soldier'), and Welsh milwr (with -wr 'man' as in cadwr). 57-59. Eng. soldier, Fr. soldat, etc. From the time of Constantine the Great the gold coin which was the standard unit was known as a solidus. 1 From this came the use of solidum or more often solidata for stipendium militum, the regular pay of soldiers, and of the verb solidare 'to pay the soldier's stipend.' Hence the mercenary was called (miles) (1) solidarius or solidatarius, (2) solidenarius, or (3) simply solidatus. 2 57. Solidatarius is represented by OProv. soudadier and by OFr. soudoier (also spelled soldoier, sodoier, saudoier, soldeier), which is the source of Eng. soldier, quotable from 1300 on in the greatest possible variety of spellings (soudoier, soudier, sauder, souldier, etc., see NED). Ir. sdighdiur 'soldier' is a blend of the Middle English form (cf. especially the spelling sougeour) with a Mid.Ir., saigdeoir 'archer,' a derivative of saiget from Lat. sagitta. Bret, soudard is from the Old French. 58. Solidenarius is represented by Oltal. soldaniere, OFr. saudenier, and especially by MHG. soldenaere, soldener, soltner (NHG. Soldner), early Danish soldener. The German form is the source of Polish Solnierz, and this again of Lith. zalnierius. 59. Solidatus is represented by Ital. soldato, Span.-Port. soldado. The Italian is the source of Fr. soldat, NHG. Soldat, which replaced the earlier French and German forms (57, 58); and also, through the medium of French (and in part of German) of Dutch soldaat, Dan.-Swed. soldat, Lett, saldats, Russ. soldat, etc. (see 54) ; and again 1 So named, not because it was of ' solid metal,' as is still often stated, but because it was the basal unit, from solidus = integer. Cf. Daremberg et Saglio, Diet, des antiq., s.v. " cette piece nouvelle . . . designee officiellement pour Stre desonnais la base de tous les comptes en or, fut, pour ce motif, qualifiee aureus solidus, et bient&t on l'appela par abreviation, solidus." 2 Of the great number of forms occurring in mediaeval Latin texts (see Du Cange) many are doubtless due to varying artificial Latinization of the current vulgar forms. Of those cited above solidarius is frequent, solidatarius rare. Yet the latter is the proper antecedent of OFr. soudoier, OProv. soudadier, or, to put it more cautiously, it is, whether artificially constructed or not, of the type of formation represented by these. Hence we are justified in employing it as a caption at least; likewise solide- narius, though MHG. soldenaere appears to be formed after, and was certainly associated with, forms like bildenaere, etc. 18 Carl D. Buck through the medium of Russian the word has passed into remote languages of Asia, e.g. Yakut sollat. Ft. soldat belongs to that large group of military terms which were borrowed from Italian in the sixteenth century (Nyrop, Gram. SURVEY OP THE MOST DISTINCTIVE WORDS 'Battle' 'War* 'Army* 'Soldier' yudk (1), rana (19) p 9 & ana (2) hamarana(l3) jang (21) razm (14) niza (24) li&XQ (3) pugna (4) battaglia (5) bataille (5) batalla (5) b&talie (5) cath (20) cad (20) waihjd (8) orrosta (18) slag (7) strit (16) Scklacht (7), Kampf (15) slag (7) gefeoht (6) battle (5) muszis (9) kauja (10) boji (11) sraienie (12) boj (11) bitva (11) bitwa (11) vigraha (30), sam prahdra (29) send (48) spa$a (40), ha end (48) Jfeara (41), haind (48) *a*AA;ar (42) aor (50) wsAiri (43) arpariis (45) exercitus (42) eserc&o (42) arm^e (44) ejercito (42) oa#te (43) sZuaffA (38) Uu (38) karjis (41) Aerr (41) Atfr (41) Aer (41) Heer (41) Zeffer (45) Aere (41), /ierd (47) army (44) rats&as (49), armija (44) karaspekks (51) vojsko (49) vojsko (49) vojska (49) vojsko (49) wojsko (49) Modern Persian jang (21) paterazm (25) 1'ti/fe (24) x6X«^os (22) bellum (35) guerra (32) owerre (32) guerra (32) r&zboiu (23) copadA (26) rA W /eJ (36) wigans (8) girrrf (16) Arr'ff (31) urliuge (34) XWeff (31) oorlog (34) gewinn (32) war (33) karas, kari (41) karsch (41) vojna (28) vojna (28) vojna (28) vojna (28) wojna (28) atpfiAi (52) usAfdr (52) mt'Ies (56) soldato (59) •oldae (59) aoZdado (59) soJdat (59) sdighdiur (57) Welsh milwr (56) Gothic gadrauhts (55) hermadr (52) *oWaf (59) Middle High New High German . Dutch soldenaere (58) Soldo* (59) soldaat (59 J Old English New English Aeremann (52) sotater (57) ialnierius (58). feareivis (54) saldais (59) Old Church Slavonic uojl (54) soida( (59) Serbo-Croatian. . . . vojnik (54) ro/afc (54) Polish iolnierz (58) Jwsi. ete Zo Zangwe frangaise, I, 56 ff. Usts about 40 such), and which aroused the ire of Henry Estienne in his Dewx dialogues du nouveau langage frangaise italianize", published in 1578. As to NHG. Soldat, which dates from the sixteenth century, it was once suggested that the form found its way into German through the medium of the Italian and Spanish troups in the low countries. But it is more to Words fob 'Battle,' 'Wab,' 'Abmy,' and 'Soldieb' 19 the point to recall that the great sixteenth-century wars between Charles V and Francis I, in which the German, Spanish, French, and Italian troops were engaged, were waged chiefly on Italian soil. The original meaning 'mercenary soldier' was often lost sight of even in earlier periods, and such restriction, with the derogatory feeling attached to 'mercenary,' was wholly absent from soUai when it spread over Europe. The same is of course true of Eng. soldier and all the other modern forms mentioned in 57, 58, except that NHG. Sbldner was restricted to the older sense of 'mercenary,' after the introduction of soldat.