IIJ. ON A CASE OF SUPPLETIVE INDO-EUROPEAN SUFFIXES MAURICE BLOOMFIELD JOHNS HOPKINS UNIVERSITY The older Indo-European languages have preserved a considerable number of adjectives in oxytone ú of the type Ved. prthús, Avest. pǝrǝpuş, Gr. λarús 'broad'; or Ved. trşús, Goth. þaúrsus; OHG. durri 'dry.' For the most part they express very primary adjectival ideas, to a large extent qualities inherent in concrete objects, showing that they were formed at a very early time. They are generally made from the reduced grade of the root, due to their oxytone accent, but morfological and lexical (semantic) influences have tended occasionally, though not very often, to efface the inter-play between oxytonesis and reduced root-form (Vedic açús Gr. Kús, etc.). Adjective function and the position of the accent preserve their distinctive class character to an uncommon degree. = In the accented texts of the Veda we find the majority (though not all) of these Indo-European formations: prthús 'wide'; trşús 'dry'; mrdús 'soft'; raghús, laghús 'light'; tanús 'thin'; bahús 'much'; anhús 'narrow'; ripús1 'tricky'; açús 'swift'; dhṛṣṇús 'bold' (for *dhrşús =Opaσus; MHG. türre, the n coming from the present stem dhṛṣṇu-)2; katus, kátuka 'sharp' (Lith. kartus 'bitter'); dhrigus 'poor' in ádhrigu 'rich' Avest. driyuş 'poor' (superlative, drāejișta)3; yahús 'swift' (cf. Avest. yazivi; Ved. fem. yahvi); vāyús 'blowing,' 'wind' (Avest. vayuş 'wind'); tákus 'swift' (with changed accent, cf. Taxis); rjús 'straight' (Avest. ǝrǝzuş); grhús 'begging' (cf. Lith. grabus 'light-fingered'); vidhús 'solitary' (cf. Lith. vidus 'inner part'); anu-şthús 'following,' susthús 'well-conditioned' (cf. Lith. at-stùs 'being far'). = Less definitely connected, or farther removed from the primary period of formation are the stems rbhú- 'active'; rşú- 'glowing'; kŕdhú 'Providing this is really comparable to Lith. lipus ‘sticky.' But scarcely any Lith. adjective in -us need be from older times; see the sequel. 'Cf. the congeneric grdhnús, "eager," below p. 93. 'The author, AJPh. 17. 429ff.; Neisser, Zum Wörterbuch des Rigveda 29. 'short'; nṛtú- 'dancing'; madgú- 'diving <bird>'; krīdú- 'playing'; vankú- 'mobile'; çañkú- 'tottering'; valgú- 'pleasing'; mandú- 'joyous'; vanú 'eager'; phalgú- 'glittering'; bhindú- 'cleaving'; vindu'finding'; tápu- 'hot' (with changed accent); bhfgu- 'shining' (with changed accent, cf. Lat. fulgor); vīḍú- ‘firm'; sādhú- 'straight'; kārú'singer'; çayú- ‘lying'; jāyú- 'victorious'; cāyú- 'honoring'; tanyú'thundering'; stāyú-, tāyú- ‘thief'; pāyú- 'protecting'; dārú- 'breaking'; piyú- 'hating'; māyú- 'lowing'; dhāyú- 'thirsting'; perú- ‘saving'; jigyú- 'victorious'; cikitsú- 'intelligent'; çikşú- 'helping'; ditsú'liberal.' The type has become indefinitely productive with the secondary conjugations, see Whitney, Skt. Gramm. 1178 g ff. There is in this long list, to which other instances might be added from the later, unaccented texts (such as patu- 'sharp'), not a single case of a word containing radical u; the u-roots obviously avoid this formation, and one's sense of the formative possibilities of the language instinctively forbids the expectation of such stems as *budhú- 'perceptive,' or *çucú- 'bright,' or *ugú- 'strong.' To this there are three apparent exceptions, namely, urú- 'wide,' gurú- 'heavy,' and purú'much.' No one has, as far as I know, noted that these three words are congeneric rime words, expressing the idea of size or extent. We find them again in Avestan in a rime state, voru-, go"ru-, and po"ru-, forms which are none too easy to bring into fonetic correspondence with their Vedic sisters. They occur with various changes in the older IndoEuropean languages and are founded upon the types *urrú-, *g*rrú-, and *pllú-, with linguals, and not u, in the root. It is not important to enter here upon the intricacies of the special changes that have determined the final shape of Greek evpú-, Baρú-, Toλú-, or Lat. gravis, or Goth. filu, because Vedic váriyas, gáriyas, and párīņas, páriman, etc., determine definitely their root vocalism. Moreover these three rime words are not alone in their semantic class. I.E. *plt(h)u- 'broad,' corresponds, more particularly with urrú- 'broad'; I.E. *bhnghú- (Skt. bahú- 'much,' Gr. Tax) with *pllú- 'much', as do, a little more remotely, ON. þykkr (accus. þykkuan), OHG. dicchi, Erse tiug 'thick'; Gr. daous 'thick,' and rappus 'thick.' Compare also Gr. ßpaxús 'short': Lat. brevis and OBulg. kratňků 'short': Ved. krdhú- 'short.' As opposites we have Skt. anhú-, Goth. aggwus, OHG. engi 'enge,' to confront with *p]t(h)ú 'Well worth noting is the Vedic triad párīņas, ápnas, and rékņas, all of them with the rare suffix -nas, and all of them in the sense of 'wealth' or 'abundance.' The suffix -nas is the abstract -as, into which is infused a touch of passivity by the n of suffix -nd, a past participle suffix, usually with passive sense. 'broad'; I.E. tņnú- 'thin,' Skt. tanú-, Gr. Tavú- in ravú-yλwooos, Lat. tenuis, OHG. dunni to confront with bhnghú- (Gr. Taxi- 'thick,' and ON. pykkr. I.E. Ing"hú- 'light,' and g❝rrú- 'heavy' are opposites which coquet with one another down to modern times in dialectic Italian greve (for grave), more congenial to leve. None of these show u in the root. The situation in Greek is essentially the same. Most of the ú-adjectives come from the same prehistoric time as the older Skt. words: πλατύς, βραδύς, θρασύς, βραχύς, παχύς, βαρύς, ἐλαχύς, κρατύς (Goth. hardus, OHG. herti), τανυ- (in τανύγλωσσος), μαλδυ- (in ἀμαλδύνω), πολύς, ὠκύς, and adus. We may add from a later period: Myús ‘shrill,' rappus 'thick,' δασύς ‘thick, ἀμβλύς 'dull, βαθύς ‘deep, τραχύς ‘rough.’ Ευρύς (with hysterogenous evpos 'breadth') is controlled by Skt. urú- Avest. vouru-. The only form with apparent u, yλUKús, in the light of Lat. dulcis (both with two sonorous sounds, u and 1, in the root) points to in the root and not to *doleuk-, as Hirt (Der Indogermanische Ablaut, 8527) assumes by a process of addition. TeûKOS 'sweetness,' is seconddary to γλυκύς.. Lat. levis, gravis, tenuis, brevis, mollis, pinguis, suavis, acu-pedius, dulcis (above) reflect the I.E. stock without u in the root. The same is true of its Teutonic continuants: Goth. þaúrsus, kaúrus, aggwus, hardus, quaírrus (MHG. kürre) 'tame,' tulgus 'firm' (ul), OHG. dunni, ON. pykkr. Goth. suts discreetly avoids the suffix -ú of svādús. The Balto-Slavic treatment of the u-stems will be touched upon later. The number of I.E. roots with radical u is, of course, not to be compared with the mass that contain other vowels. Nevertheless, if we regard, as we well may, the Vedic list as a faithful reflex of I.E. conditions, we find a very respectable list of u-roots. Some of them belong to the oldest and most familiar stock: ruc 'shine,' yuj ‘join,' bhuj 'enjoy,' bhuj 'bend,' pruth 'snort,' plu 'flow,' çru 'hear,' sru 'flow,' dhú 'shake,' bhu 'be,' pū 'purify,' tud 'thrust,' nud 'push,' mud 'rejoice,' ud 'flow,' rud 'howl,' budh 'perceive,' kup 'be angry,' rup (lup) 'break,' lubh 'desire,' uş 'shine,' juş 'enjoy,' tuş 'enjoy,' çuş 'be dry,' pruş 'sprinkle.' There is no occasion here to be exhaustive, nor is it necessary to list the 50 or more u-roots of the Sanskrit language easily gathered from Whitney, Roots, Verb-forms, etc. 243ff., which are not I.E., or, at least, not obviously so: e.g. kuc 'curl,' tuc 'generate,' muc 'release,' mruc (mluc) 'set,' çuc 'shine,' tuj 'thrust,' ruj 'break,' Sanctioned by the English and German mind in the alliterative expressions, 'thru thick and thin,' 'durch dick und dünn.' çcut 'fall,' dyut 'shine,' khud 'futuere,' cud 'impel,' etc. Not a single one of these roots makes an adjective in -ú: just as the u-adjectives are never made from u-roots, so u-roots never make adjectives in u-. This is evidently a case for suppletion: the u-roots must resort to other afformatives. It is not to be expected that this should take place on a concerted scheme on the part of the u-roots. Thus the very primary idea "dry" is mobilized in the Aryan languages by the suffix -ko-, very rare as a primary suffix (Vedic çuşka-, Iranian huşka-6 from an I.E. root sus,) in other languages by the suffix -o- (Gr. avos, AgS. séar, Lith. saūsas, OBulg. suchů). Or, the I.E. root lubh 'love,' makes a primary adjective in -o- Goth. liuba, OHG. liob 'lieb,' OBulg. ljubŭ. Nevertheless, the far more common suppletion of the suffix -ú in roots that contain u lies with the old primary suffix, I.E. -ró-. From the Indic point of view some of the adjectives in -rá- are I.E.; some are Aryan; and quite a number appear only in Vedic and Sanskrit. I.E. age are rudhirá 'red' (*rudhró-) which may have given rise to the later congeneric rucirá- 'bright"; udrá- 'water animal': Gr. vôpos, iSpa 'water-snake,' OHG. ottar, Lith. údra, OBulg. vydra 'otter'; cubhrá- 'bright': Arm. surb 'pure'; usrá- 'shining,' 'matutinal': Goth. austro-, in Austro-goti; Avest. zāeni-budra- 'eagerly watchful': Lith. budrus, OBulg. budru 'watchful.' Vedic and Avestan are Ved. ugrá-, Avest. uyra- 'strong'; Ved. kṣudrá- 'small,' Avest. xṣudra'semen's; Ved. çukrá- (çuklá-) 'bright,' Avest. suxra-, Old Pers. puxra'red.' Purely Indic are Ved. Rudrá- 'Howler,' túmra- 'swelling' (with irregular accent: Lat. tumeo), mudrá- "joyous" (AV. 18.3. 19), dudhrá- 'stiff.' From a later time: mucira- 'liberal,' muhira- 'foolish,' cf. also çūdrá- 'member of fourth caste' (of obscure belongings). Well worthy of note is the frequent use of the suffix -ró- with roots ending in u, long or short: Ved. sthūrá- (sthūlá-) 'thick,' OHG. stiuri, ON. stür 'strong'; Ved. krūrá-, Avest. xrūra- 'bloody'; Ved. dūrá-, Old Pers. dūra- 'far'; Ved. çúra- 'strong' (accent to match çávira'mighty'), Avest. sūra-, Gr. ǎ-Kupos 'weak'; Ved. rurá- 'hot'; Ved. mūrá- 'foolish'; and perhaps another murá- 'hasty' (to Lat. movēre); Ved. kşurá- 'scratching,' 'razor,' Gr. §upóv, 'razor'; Ved. turá-, 'strong.' From the roots of these words, or from any other root in final й there Cf. their congeneric words for 'dry': Lat. siccus for *sit-qo-s; Cymric hysp from *sis-kyo-s. 'Both words occur as designations of 'saffron.' 'H. Pedersen, IF 5. 61, identifies, not implausibly, Gr. Vudpós, 'false' with this pair. exists not a single instance of an adjective in -ú-; such a formation is, indeed, typically unimaginable. More or less antique cases of -ró- adjectives with u in the root are, of course, to be found in all the languages: OHG. slupfer 'slippery' (Goth. sliupan 'slip'), Lat. lubricus; Goth. hlutrs, OHG. hlutar 'lauter' (cf. Gr. kλúšw 'rinse'); Goth. snutrs, OHG. snottar, O.N. snotr, Middle Engl. snoter 'wise' OHG. sur 'sour,' Lith. súras 'salty,' OBulg. syru 'raw'; Lat. pūrus (cf. Skt. pūtás 'purified'); Lat. obscūrus (cf. OHG. skür, 'cover'); Lat. muger (Paul. Diac.), 'qui talis male ludit' (cf. mufrius, 'gambler,' Petronius); MHG. miuchel 'secret'; Umbr. vufru 'votivum' (Buck, Grammar of Oscan and Umbrian, §§152; 257.1); Sabinian cuprum 'bonum,' Cupra Bona Dea'; Umbr. Cubra (Buck, ibid. 310, collection, nr. 83), from the root of cupio 'desire'; Gr. vypós 'moist' (cf. ON. vokua 'moisture'), Gr. udpós 'false' (above); Gr. Avypós 'mournful' (Epic); Neulepos, Lat. liber (stem leufero-). That the Balto-Slavic adjectives in -u- continue to some extent the corresponding I.E. adjectives is guaranteed rather by isolated nouns in -u- than by the u-adjectives themselves. Such nouns are Lith. medus, Old Pruss. meddo, OBulg. medй 'honey'; Lith. pekùs, Old Pruss. pecku 'cattle'; Lith. alùs, Old Pruss. alu, OBulg. olu, ON. gl 'beer,' 'ale'; or with suffix -nu-, Lith. sunùs OBulg. synů 'son.' As regards the u-adjectives, the long lists in Leskien, Die Bildung der Nomina im Litauischen 244ff., bring, on the one hand, surprisingly few cases that can claim antiquity; on the other hand, the great spread of this type shows obvious secondary propagation. This appears most clearly in doublets like szaunùs and szaúnas 'excellent," darbus and darbas laboring,"10 rupùs and rùpas 'rough," romùs and romas 'gentle," etc.; see especially Leskien, ibid. 173ff., 355ff. The stock of old ú-adjectives should be headed by pilus 'full' (cf. pilnas), if Leskien's statement (p. 248) did not cast doubt upon it. However, saldus, OBulg. sladňků 'sweet' have relieved I.E. suādú- by substituting the stem sald- 'salt,'18 as the root of the word (cf. the OBulg. opposite bridŭké ‘bitter'). The triangle Lith. platùs, Gr. πλarús, Ved. prthú- intrigues (one would expect Lith. piltùs), but doubtless establishes the stem as I.E. fundamentally. 'Leskien, Nomina 356. 10 Leskien, ibid. 173. 11 Ibid. 258. 12 Ibid. 179. 13 Lett. salds has passed the word into the a-class. |