¹ The first phonetic trait that I shall examine below is quoted, e.g. by Bloomfield 469 f., relying on Linguistic survey of India (hereafter LSI) 4.278 ff. Of the two other suggested influences mentioned by Bloomfield, the first, that the confusion of Indo-European l and r in Sanskrit is perhaps due to a substratum language which possessed only one of these sounds, has nothing to recommend it, since we know of no language in India with this characteristic. The remaining one will be discussed below; it concerns the use of distinct singular and plural stems to which the same case endings are added.

² ‘India’ and ‘Indian’ will be used in what follows for the subcontinent, ignoring the political division into the Republic of India and Pakistan, and, when necessary, including Ceylon also.

³ For an attempt at reconstruction, see my article Linguistic prehistory of India, Proc. Amer. Philosophical Society 98.282–92 (1954).

⁴ This is the historical process to be invoked, rather than the too facile and unrealistic one of a general displacement of populations through expansions. Undoubtedly there were expansions involved, in the shape of marauding bands and of missionaries, but neither of these agencies had an interest in getting rid of earlier populations; it was to their advantage, political, economic, religious, to have subjects and proselytes. Absorption, not displacement, is the chief mechanism in radical language changes of the kind we are considering.

⁵ See also my article referred to in note 3 for a sketch of the history of this part of Indic scholarship and a bibliography.

⁶ Earlier in BSL 25.1–21 (1925) and BSOS 5.730–44 (1930).

⁷ Bloch (325 f.) attempts to use the presence of aspirated consonants in the northern Munda languages as somehow responsible for Indo-Aryan aspirated consonants. His argument is both confused and hypothetical ; moreover, we know too little of Munda to be sure of the nature and the history of these sounds there. — E. Prokosch, A comparative Germanic grammar 39 (1939), based part of his argument about the origin of Sanskrit voiced aspirated stops (and consequently the nature of the Indo-European phonemes involved) on his belief that voiced aspirated stops are found in Tibeto-Burmese, Dravidian, and Munda. This is true for Tibeto-Burmese, and very uncertain in its bearing for Munda (for the reason given above). As regards Dravidian, Prokosch was misled by E. H. Tuttle (Modern philology 18.52), who mistakenly assumed that Proto-Dravidian had aspirated stops because of the quite secondary occurrence of such sounds in modern Kannada and Telugu.

⁸ Trubetzkoy, Acta linguistica 1.84, set down as one of the six characteristics of an Indo-European language: ‘Das Wort muss nicht unbedingt mit der Wurzel beginnen. — Es gibt keine indogermanische Sprache ohne Präfixe. ... In den jüngeren indogermanischen Sprachen nimmt die Zahl solcher Präfixe stark zu.’ And further (85) : ‘Eine Sprache, die nicht alle genannten Strukturmerkmale besitzt, darf nicht als indogermanisch gelten.’ Did he know the Modern Indo-Aryan languages?

⁹ This form has neither morphologically nor in syntactic use anything to do with either Latin gerund or gerundive or participle; no really satisfactory term has yet been invented. Slavicists have been equally unhappy in calling Russian forms with a comparable syntactic use ‘past gerunds’. These forms in Russian and other Slavic languages are a modern innovation, not found in Old Church Slavic. There is no connection with the Indic phenomena.

¹⁰ Pace Bloch 327. See P. O. Bodding, Materials for a Bardali grammar, II. Mostly morphological 273 ff. (1929); J. Hoffmann, Mundari grammar 189 ff. (1903); G. V. Ramamurti, A manual of the So:ra: (or Savara) language 28 §59, 29 §§71–2, 44 §143, §145.

¹¹ Bloch, Structure grammaticale des langues dravidiennes 64–6; Emeneau, Lg. 24.321, 1st paragraph.

¹² P. O. Bodding 50–1 ; J. Hoffmann lv-lvi, 120–1,201–3; G. V. Ramamurti 49 §168, §§170–3.

¹³ Circular 3, 14 May 1928, 7–8; Circular 4, 25 June 1928, 2, 8–10, 13–14, 16.

¹⁴ The best accounts with examples are given by Suniti Kumar Chatterji, The origin and development of the Bengali language 777–81 (‘enclitic definitives or numeratives‘); Banikanta Kakati, Assamese, its formation and development 265–9 (1941); (Sir) George Abraham Grierson, Seven grammars of the dialects and sub-dialects of the Bihárí language. The matter was noticed by Bloch, L'indo-aryen du Veda aux temps modernes 189 (1934) ; he thinks of substratum influence and refers to Siamese.

¹⁵ The matter is hard to dig out of the Marathi grammars. A summary statement is found in H. M. Lambert, Marathi language course 243 (1943). An even more summary statement for Konkani is in S. M. Katre, The formation of Koṇkaṇī 117 §236 (1942). Bloch has no notice of this construction in La formation de la langue marathe ; in his reference to the Magadhan phenomena in L'indo-aryen 189, he says that the construction is limited to the eastern, i.e. the Magadhan, group of languages.

¹⁶ Oral communication for Hindi from my pupil Phillip Barker, on the basis of experience in several Hindi-speaking village communities at widely separated places. A few scattering examples turn up in LSI, but are not worth recording. It should however be noted that the LSI records jhan ‘people’ with jh- instead of j- for Chhattīsgaṛhī Hindi, and also for Halbi and the Bihari spoken in the Ranchi district (LSI 5.2.284). For Nepali, Turner's dictionary s.v. janā.

¹⁷ Wardha dialect in M. B. Emeneau, Kolami, a Dravidian language §4.63, §8.55 (University of California Publications in Linguistics, Vol. 12; 1955); other dialects reported so far as I know them in §§9.21–3. The Adilabad dialect is in P. Setumadhava Rao, A grammar of the Kolami language (1950).

¹⁸ T. Burrow and S. Bhattacharya, The Parji language (1953). References for jan are 71 sent. 15,104 sent. 1 and4,107 sent. 24,120 sent. 1 (2 exx.) ; for , 108 sent. 30, 133 sent. 89.

¹⁹ F. V. P. Schulze, A grammar of the Kuvi language 100 ff. (1911); A. G. Fitzgerald, Kuvin̄ga bassa: The Khond language as spoken by the Paṛjas ...of the Madras Presidency (1913) ; W. W. Winfield, A grammar of the Kui language (1928).

²⁰ A. Grignard, A grammar of the Oraon language 29 f. (1924).

²¹ Ernest Droese, Introduction to the Malto language 15–26 (1884).

²² In this etymology and that given in note 25, the sigilla for the languages are as in Lg. 29.339 fn. 1, or as in Kolami, a Dravidian language. Ta. maka, makavu child, young of animal; makaṉ son, man, husband; makaḷ daughter, woman, wife; makkaḷ children, human beings; mōṉai sonny (term of endearment in addressing a child); Ma. makan, (vulg.) mōn a son; makaḷ a daughter; makkaḷ children, the young of animals; Ko. mog (oblique mog-, mo·t-) child, wife; pe-mog woman (peṇ female); mo·ḷ daughter; To. mox (obl. mo·t-) child, son, male, daughter, woman; toz mox Toda woman or wife; Ka. maga son, male; magaḷ daughter; makkaḷ children; magavu, moga, mogu, moguvu infant; Koḍ. mo·vëⁿ son; mo·va daughter; makka children; pom makka women, wives (poṇṇï wife, female); Tu. mage, mōnu son; magaḷu, mōni daughter; makkaḷāṭigè childishness; Te. maga, coll. moga male; magãḍu husband, a man, a male, a hero; magadi male of any animal, beast, or bird; maganālu wife; Malt. maqe boy; magi girl; mago small, little, young; maqu a young one (animal); maqmaqo small ones (q = x); Kol. magvan husband (so correct Kolami, a Dravidian language).

²³ Information from Phillip Barker. Note the form with jh-, which is that found in Kurukh and in the neighboring Hindi dialects.

²⁴ Information from Bh. Krishnamurti.

²⁵ A. H. Arden, A progressive grammar of the Telugu language 88 §236 (4th ed., 1937), does not make the matter at all clear. Etymology of mandi : Ta. maṉṟu hall of assembly, court of justice, cow-stall, herd of cows, raised platform under a tree for village meetings, juncture of four roads; maṉṟam hall, assembly, court, meeting place under a tree, open space, cowshed, long street; mantai (not in old literature) flock, herd, common pasture, open space in middle of a village; Ma. maṉṉam, maṉṉu a place of judgment or discussion; Ko. manḏ Toda mund; burning place for dry funeral ; mandm meeting; To. moḏ locus of tribal activity, including village with dairy, dairy apart from village, funeral place; patrilineal clan; Ka. mande, mandi flock of sheep or goats, herd of cattle or buffaloes; open place (in the jungle or near a village) where a flock or herd stands, a pen, fold; persons, people; Koḍ. mandï village green; Tu. mandi, mandè people, persons; Te. manda flock, herd, drove, pack, (B also) place where flocks or herds are kept outside a village, hamlet inhabited by herdsmen; mandi a number, crowd, or collection of persons; retinue, infantry; Kol. (SR) mandï men; (Kin.) mandi man. If all these words are to be put together, the Te. forms require some phonological explanation. The correct phonetic correspondence is: PrDr *ṉṟ : Ta. ṉṟ, Ma. ṉṉ, Ko. ḏ (the forms in this etymology will not be discussed here), To. ḏ, Ka. Koḍ. nd, Tu. ṇḍ (and other developments), Te. ṇḍ, Kol. nd, Pa. nd, (NE dialect) nḍ, Oll. nḍ (see further Emeneau, Kolami, a Dravidian language §10.25). A simple example for Te. is the word for belleric myrobalan (Terminalia bellerica Roxb.): Ta. tāñṟi, Ma. tāññi, Ka. tāṟi (this is a form without nasal), Tu. dāṇḍi, Te. , Pa. dēindi. However, there a few noteworthy exceptions to the statement, including the word under discussion and the following two: ‘pig,‘ Ta. paṉṟi, Ma. paṉṉi, Ko. paj (*-ṉṟi > j), To. pody (probably a borrowing from Badaga, a Ka. dialect), Ka. Kod. pandi, Tu. pañji, Te. pandi, Go. paddī, Pa. pend, (NE) penḍ, Oll. panḍ, Kui paji; Ta. kaṉṟu young of various animals (calf, colt, etc.), sapling, Ma. kaṉṉu (oblique kaṟṟu) young of cattle (especially buffalo calf), young plantain trees round the mother plant, kaṟṟa boy, calf, To. koṟ female buffalo calf below one year, Ka. kaṟu, kaṟa, kaṟuvu calf, kanda young child, kandu calf, young plantain trees round the mother plant, Tu. kañji calf, Te. kandu, kanduvu infant, Pa. kar sapling, Kur. khadd child, young animal or plant, Malt. qade son. There seems to be no possibility of a contextual explanation for the Te. irregularity. Influence from Ka. might be invoked in all these cases, from Ta. also in the latter two.

²⁶ F. Kittel, A grammar of the Kannaḍa language 258 (1903) ; Harold Spencer, A Kanarese grammar 94 (1914).

²⁷ A. H. Arden, A progressive grammar of common Tamil 118 (4th ed. 1934) ; Julien Vinson, Manuel de la langue tamoule 102 (1903).

²⁸ This term ‘linguistic area’ may be defined as meaning an area which includes languages belonging to more than one family but showing traits in common which are found not to belong to the other members of (at least) one of the families. It is perhaps not quite satisfactory as a technical term, though it has the virtue of having been used previously in this sense by H. V. Veiten as a translation of Trubetzkoy's ‘Sprachbund’, Pacific Northwest quarterly 34.271–92 (1943). (My attention was called to this by Sebeok, UAL 10.214 [1944], and Voegelin, Word 1.58 [1945]). Among the disadvantages of the term is the lack of an adjective and the impossibility of using the reverse phrase ‘areal linguistics’, since this is preempted by the Italian neolinguistic school in another sense. Perhaps however it will do for the moment, until some more ingenious scholar invents better terminology.

²⁹ We must not, however, neglect Bloch's final remark and his reasons therefor (330): ‘Ainsi donc, si profondes qu'aient été les influences locales, elles n'ont pas conduit l'aryen de l'Inde ... à se différencier fortement des autres langues indo-européennes.‘