Page:EB1911 - Volume 12.djvu/736

Rh which is also common in Bengali. Many words which have i in H. have a in G. and R., thus, H. likhē, G. lakhē, he writes; H. din, G. and R. dan, a day. Similarly we have a for u, as in H. tum, G., R. tamē, you. In colloquial G. ā often becomes ȧ, and ī becomes ē; thus, pȧṇī for pāṇī, water; mārēs for mārīs, I shall strike. As in most Indo-Aryan vernaculars an a after an accented syllable is very lightly pronounced, and is here represented by a small ᵃ above the line.

The Vedic cerebral l and the cerebral ṇ are very common as medial letters in both G. and R. (both being unknown to literary H.). The rule is, as elsewhere in western and southern intermediate and outer languages, that when n and l represent a double ṇṇ (or nn) or a double ll in Pr. they are dental, but when they represent single medial letters they are cerebralized. Thus Ap. soṇṇaũ, G. sōnũ, gold; Ap. ghaṇaũ, G. ghaṇũ, dense; Ap. callai, G. cālē, he goes; Ap. calai, G. caḷē, he moves. In northern G. and in some caste dialects dental and cerebral letters are absolutely interchangeable, as in ḍāhᵃdō or dahāḍō, a day; tũ or ṭũ, thou; dīdhō or dīḍhō, given. In G. and R. medial ḍ is pronounced as a rough cerebral ṛ, and is then so transcribed. We have seen that in the Marwari alphabet there are actually distinct letters for these two sounds. In colloquial G. c and ch are pronounced s, especially in the north, as in pā̃s for pā̃c, five; pusyō for puchyō, he asked. Similarly, in the north, j and jh become z, as in zāḍ for jhāḍ, a tree. In some localities (as in Marathi) we have ts and dz for these sounds, as in Tsarōtar (name of a tract of country) for Carōtar. On the other hand, k, kh and g, especially when preceded or followed by i, e or y, become in the north c, ch and j respectively; thus, dicᵃrō for dikᵃrō, a son; chētar for khētar, a field; lājyō for lāgyō, begun. A similar change is found in dialectic Marathi, and is, of course, one of the commonplaces of the philology of the Romance languages. The sibilants s and ś are colloquially pronounced h (as in several outer languages), especially in the north. Thus dēh for dēś, a country; hũ for śũ, what; hamᵃjāvyō for samᵃjāvyō, he explained. An original aspirate is, however, often dropped, as in ’ũ for hũ, I; ’ātē for hāthē, on the hand. Standard G. is at the same time fond of pronouncing an h where it is not written, as in amē, we, pronounced ahmē. In other respects both G. and R. closely agree in their phonetical systems with the Apabhraṁśa form of Śaurasēnī Prakrit from which the Midland language is derived.

Declension.—Gujarati agrees with Marathi (an outer language) as against Hindostani in retaining the neuter gender of Sanskrit and Prakrit. Moreover, the neuter gender is often employed to indicate living beings of which the sex is uncertain, as in the case of dikᵃrũ, a child, compared with dikᵃrō, a son, and dikᵃrī, a daughter. In R. there are only sporadic instances of the neuter, which grow more and more rare as we approach the Midland. Nouns in both G. and R. may be weak or strong as is fully explained in the article . We have there seen that the strong form of masculine nouns in Western Hindi generally ends in au, the ā of words like the Hindostani ghōṛā, a horse, being an accident due to the fact that the Hindostani dialect of Western Hindi borrows this termination from Panjabi. G. and R. follow Western Hindi, for their masculine strong forms end in ō. Feminine strong forms end in ī as elsewhere. Neuter strong forms in G. end in ũ, derived as follows: Skr, svarṇakam, Ap. soṇṇaũ, G. sōnũ, gold. As an example of the three genders of the same word we may take G. chōkᵃrō (masc.), a boy; chōkᵃrī (fem.), a girl; chōkᵃrũ (neut.), a child. Long forms corresponding to the Eastern Hindi ghoṛᵃwā, a horse, are not much used, but we not infrequently meet another long form made by suffixing the pleonastic termination ḍō or ṛō (fem. ḍī or ṛī; G. neut. ḍũ or ṛũ) which is directly descended from the Ap. pleonastic termination ḍaü, ḍaī, ḍaũ. We come across this most often in R., where it is used contemptuously, as in Turuk-ṛō, a Turk.

In the article it is shown that all the oblique cases of each number in Sanskrit and Prakrit became melted down in the modern languages into one general oblique case, which, in the Midland, is derived in the singular from the Ap. termination -hi or -hĩ, and that even this has survived only in the case of strong masculine nouns; thus, ghōṛā, obl. ghōṛē. In G. and R. this same termination has also survived, but for all nouns as the case sign of the agent and locative cases. The general oblique case is the same as the nominative, except in the case of strong masculine and neuter nouns in ō and ũ respectively, where it ends in ā, not ē. This ā-termination is characteristic of the outer band of languages, and is one of the survivals already referred to. It is derived from the Apabhraṁśa genitive form in -aha, corresponding to the Māgadhī Pr. (an outer Prakrit) termination -āha. Thus, G. chōkᵃrō, a son; chōkᵃrũ, a child; obl. sing. chōkᵃrā.

In G. the nominative and oblique plural for all nouns are formed by adding ō to the oblique form singular, but in the neuter strong forms the oblique singular is nasalized. The real plural is the same in form as the oblique singular in the case of masculines, and as a nasalized oblique singular in the case of neuter strong forms, as in other modern Indo-Aryan vernaculars, and the added ō is a further plural termination (making a double plural, exactly as it does in the Ardhamāgadhī Prakrit puttā-ō, sons) which is often dropped. The nasalization of the strong neuter plurals is inherited from Ap., in which the neuter nom. plural of such nouns ended in -aāĩ In R. the nominative plural of masculine nouns is the same in form as the oblique case singular, and the oblique plural ends in ā̃. The feminine has ā̃ both in the nominative and in the oblique plural. These are all explained in the article. We thus get the following paradigms of the declension of nouns.

The general oblique case can be employed for any case except the nominative, but, in order to define the meaning, it is customary to add postpositions as in Hindostani. These are:

The suffix nō of the genitive is believed to be a contraction of taṇō, which is found in old Gujarati poetry, and which, under the form tanas in Sanskrit and taṇaü in Apabhraṁśa, mean “belonging to.” It is an adjective, and agrees in gender, number and case with the thing possessed. Thus, rājā-nō dikᵃrō, the king’s son; rājā-nī dikᵃrī, the king’s daughter; rājā-nũ ghar, the king’s house; rājā-nā dikᵃrā-nē, to the king’s son (nā is in the oblique case masculine to agree with dikᵃrā); rājā-nē gharē, in the king’s house. The rō and kō of R. are similarly treated, but, of course, have no neuter. The dative postpositions are simply locatives of the genitive ones, as in all modern Indo-Aryan languages (see ). Thī, the postposition of the G. ablative, is connected with thawũ, to be, one of the verbs substantive in that language. The ablative suffix is made in this way in many modern Indo-Aryan languages (e.g., q.v.). It means literally “having been” and is to be ultimately referred to the Sanskrit root, sthā, stand. The derivation of the other postpositions is discussed in the article.

Strong adjectives agree with the nouns they qualify in gender, number and case, as in the examples of the genitive above. Weak adjectives are immutable.

Pronouns closely agree with those found in Hindostani. In the table on following page we give the first two personal pronouns, and the demonstrative pronoun “this.”

Similarly are formed the remaining pronouns, viz. G. ā, R. ũ, he, that; G. tē, R. sō (obl. sing. tī̃), that; G. jē, R. jō, who; G. kȧṇ (obl. kȧṇ, kō, or kē), R. kuṇ (obl. kuṇ), who?; G. śũ, R. kā̃ī̃, what?; G., R. kōī, anyone, someone, kāī̃ anything, something. G. has two other demonstratives, pēlō and ōlyō, both meaning “that.” The derivation of these and of śũ has been discussed without any decisive result. The rest are explained in the article. The