Pramāṇa · Search · About · API

A Comparative Grammar of the Modern Aryan Languages of India v1.c6: Chapter III

A Comparative Grammar of the Modern Aryan Languages of India v1.c6

Language
English

Chapter III

v1.c6.1

CONTENTS.—§ 48. GENERAL REMARKS: POSITIONAL AND ORGANIC CHANGES.—§ 49. POSITIONAL CHANGES—INITIAL CONSONANTS.—§ 50. MEDIAL CONSONANTS.—§ 51. (1) RETENTION.—§ 52. (2) SOFTENING OF TENUES TO MEDIAE.—§ 53. (3) ELISION.—§ 54. LAWS OF THE THREE PROCESSES.—§ 55. FINAL CONSONANTS.—§ 56. ORGANIC CHANGES—GUTTURALS.—§ 57. RELATION OF PALATALS TO LINGUALS.—§ 58. MARATHI SUBSTITUTION OF SIBILANTS FOR PALATALS.—§ 59. CONNEXION BETWEEN CEREBRALS AND DENTALS.—§ 60. FURTHER TRANSITION OF CEREBRALS INTO *l* AND *ḷ*.—§ 61. MODIFICATIONS AND CHANGES OF SEMIVOWELS, (1) *ल*.—§ 62. SEMIVOWELS, (2) *य*.—§ 63. SEMIVOWELS, (3) *र*.—§ 64. SEMIVOWELS, (4) *व*.—§ 65. THE NASALS.—§ 66. THE SIBILANTS.—§ 67. *ह*.—§§ 68, 69. THE ASPIRATES.—§ 70. VISARGA AND ANUSVÂRA.—§ 71. INVERSION OF WORDS.

v1.c6.2

§ 48. By single consonants are implied those consonants which stand alone in a word, and which being in a measure unsupported are subjected to many and various changes. These changes may conveniently be considered under two classes. First, those whose operation varies with the position of the letter in a word, which are therefore called positional changes; and, secondly, those whose operation is not influenced by position, and which, being chiefly changes from one organ of speech to another, will be called organic changes.

v1.c6.3

Both classes are sometimes seen in activity in the same word, and their influences are very conflicting and confusing, but there is as a rule much uniformity in the way in which the whole seven languages undergo positional changes; while in the case of organic changes, on the other hand, the individual peculiarities of the several languages stand out in strong relief. Positional changes are more universal in their application, organic ones more restricted to special instances; the former are based upon general principles, the latter on the various and sometimes irregular developments of provincial pronunciation.

v1.c6.4

§ 49. Positional changes follow the general principle which rules all the languages of the Indo-European family, and which has been thus concisely stated by Grimm: "Initial consonants retain the grade of each organ in the purest and truest way, medial consonants have a tendency to soften, finals to harden."¹ By the expression "grade" must be understood the two classes of tenues and mediæ; and it is thus laid down as a rule that tenues *k, p, t*, etc., in the middle of a word have a tendency to soften into their corresponding mediæ *g, b, d*; while medial *g, b, d*, at the end of a word have a propensity to harden into *k, p, t*. In the Indian languages the rule holds good in the main, and although a great variety of modifications is observable, they can all be traced by careful examination to the operation of this law, either in the exact method stated above, or in the still more advanced developments of it.

v1.c6.5

¹ *Geschichte d. Deutschen Sprache*, vol. i., p. 251, "Anlaut hält die stufen jedes organs am reinsten und treusten, inlaut ist geneigt es zu erweichen, auslaut zu erhärten." We have unfortunately no equivalents for anlaut, inlaut, or auslaut. These words, meaning on-sound, in-sound, and out-sound, respectively, are also applied to vowels, but in the passage quoted only consonants are referred to.

v1.c6.6

Initial consonants remain undisturbed, whether in Tatsamas or Tadbhavas.

v1.c6.7

*a.* In Tatsamas, as कमल, कन्या, गृहस्थ, गिरि, चकोर, चन्दन, जीवन, तूरी, मान, मेघ, सिन्दूर, रुचि, हार. Here are instances of tenues, mediæ, semivowels, nasals, sibilants, and ह. In fact, from the very nature of Tatsamas, which are words in their pure Sanskrit form, it follows that no change can have taken place, otherwise they would cease to be Tatsamas.

v1.c6.8

β. In Tadbhavas, as कडाही, कौडी, कोइल्, जोंक, तिरछा, नेव, नंगा, मूठ, रोना, लोंग, सूई, हाथ, from कटाह, कपई, कोकिल, जलूक, तिर्यक्, नेमि, नम्र, मुष्टि, रोदनं, लवङ्ग, सूचि, हस्त, where, in spite of violent changes in the interior of the word, the initial letter remains unaltered. A very large number of these words may be found in every page of the dictionaries of these languages. When initial changes do occur, they are generally found to be due to some organic disturbance in the body of the word, which has exercised a retrospective influence on the initial letter. Such changes are, however, altogether exceptional, and comparatively few in number. The most prominent and common of them will be noticed below. To be omitted from the present section, however, are all those words which in Sanskrit begin with स and स. The former generally appear with an initial फ, the latter with थ or ठ. The aspiration is due to the influence of the sibilant, and, as we are here considering single consonants only, this process does not fall within our present subject.

v1.c6.9

There are, however, instances where the sibilant forms the first member of a nexus in the middle of a word, and in going out has affected, not only the letter to which it was joined, but also the initial. Thus, Skr. पुष्प “flower,” becomes in Pr. पुप्फ, but in Old H. पुझप, and finally फुत or फूप. Skr. कस्तूरी “civet,” S. ख्युरी; Skr. वाष्प “vapour,” H. भाफ, and the same in P., B., and O., where both letters are aspirated. The form बाफ also occurs in B. and H., and in S., G., and M. it is the only form in use. The Skr. कच्छू “a scab,” is perhaps the origin of H. खाज “itch”; also खुजली “a diminutive,” where there has been epenthesis of u, as described in § 34; S. खाजी, M. खाजणें “to scratch,” B. खसु. Skr. कूप “a well,” is S. खूह and खुझ; Skr. कुसुम “saffron,” = S. खुझंबो; Skr. कांस “cough,” H., P., and S. खांसी, खांसना, खांसणु.

v1.c6.10

In some cases an aspirate in the middle of a word is thrown back to the initial and amalgamates with it, as Skr. गृह “house,” original form गह, still preserved in Kashmiri गरह. This becomes in all the modern languages घर.

v1.c6.11

Skr. दुहिता “daughter,” becomes in Pali धितर and धीतर, and even in Skr. a form धीदा is in use. The word is oxytone, and the coalition of the two short toneless syllables into one is therefore natural. Pr. धीआ, धूदा, H. धी, धीया, P., G. id., S. धिउ, धिय, B. झी, O. झिओ (jhio); the two last forms have arisen from the close connexion between jh and dh, which, especially when followed by a palatal vowel or semivowel, is very frequent. To the parallel connexion between ज and द may be ascribed the isolated H. form जामाद, Skr. जामाता “son-in-law,” where all the other languages have ज, as M. जामात, B. जामाई, O. जुआँइ and ज्वाइँ, G. जमाई, S. ज्ञाटो.¹

v1.c6.12

¹ The Hindi word is borrowed immediately from the Persian داماد, which accounts for the change; Skr. ज often becomes ज in Persian through an intermediate j.

v1.c6.13

The Skr. पनस “the Jack-tree,” becomes in Pr. फणसो, H. फनस, but in this case it may be assumed that in Skr. also the original form was फणस, derived from फण “the expanded hood of the cobra,” to which the leaves of this tree bear a close resemblance. Sindhi has झंगु and झंगली, from Skr. जंगल, in which irregularity it stands alone, unless the name of the Panjab town Jhang be derived from the same, which is highly probable, as it stands in the centre of a vast desert.

v1.c6.14

Skr. महिष “buffalo,” becomes H. भैंसा, fem. भैंस, G. भेंस, B. भंइष, O. भयेष. In many parts of Hindustan म्हेंस is commonly used. M. has only म्हैस, while P. preserves the initial म in मञ्झ; so also does S. in मंझ and मंहि. The form in भ has of course arisen from a corruption of म्ह. The reverse of this process is seen in M. म्हणे, from Skr. भण् “to speak”; similar to it is Pr. भरणे “remembrance,” for म्हरण, from स्मृ०.

v1.c6.15

In Prakrit writings changes of initial consonants are more common. The total rejection of an initial consonant occurs with frequency. Especially is this the case in the effeminate dialect of the Saptaśataka, where, however, it must be noted that the examples given by Weber, *muhaanda*=*mukhachandra*, *kuḍumvaïṭṭhi*=*kuṭumbakṛiṣṭi*, *nahaala*=*nabhastala*, are not cases in point, inasmuch as the words *chandra*, *kṛiṣṭi*, *stala*, being the last members of a compound, their initial letter is no longer an initial, but becomes subject to the laws which rule medial letters. The supposition that the Prakrit root *achh* is from *gachh* (*gam*), by rejection of the initial, is directly opposed to Vararuchi, who explains it by *as*. In all the passages where this word occurs the context requires that it should be translated by "stay," "remain," "stop,"—precisely the reverse of the idea of *gachh*. The modern languages exhibit two forms of the substantive verb, or perhaps two separate verbs: अस् and अच्छे; the latter of which seems to point to अछ. But reserving the discussion of this question to a more suitable place, the Magadhi of the Jains, as well as the language of the Saptaśataka, presents instances of rejection of initial consonants; thus, *a*=*cha*, *uno*=*punar*, *aḍham*=*gāḍham*, *auṇā*=*chatur*, and the like. This process is absolutely unknown in the modern languages, and it may be doubted whether it was ever really in vogue even in Prakrit beyond the limits of literary composition.

v1.c6.16

The softening of initial tenues into mediæ, or, in the case of the labials, into the semivowels, is found to occur in instances where the modern languages retain the letter in its original Sanskrit grade. Thus *vaï*=*pati*, *vaä*=*pada*, *vaṇḍu*=*pāṇḍu*, *vaṭṭa*=*patra*, *vaḍāā*=*paṭākā*. The process appears to have been almost confined to *p*, a letter which, as we shall frequently see in this chapter, is peculiarly weak and liable to change.

v1.c6.17

On the whole, however, Grimm's law is observed faithfully, and the instances where it is not followed may be regarded as exceptions, in each case of which some special reason exists for the change; and it may be further observed that the modern languages in no single well-authenticated case follow the Prakrit custom of entirely rejecting the initial or softening it into its corresponding media.

v1.c6.18

§ 50. It is in medial consonants that the greatest variety of change takes place. They are either retained unaltered, or weakened into softer sounds, tenues to mediæ, mediæ to semi-vowels and vowels, or they disappear altogether. In the latter instance the consonant sometimes carries away with it the following, and sometimes even the preceding, vowel, so that a whole syllable disappears.

v1.c6.19

In Prakrit it is stated that there are instances of hardening of medial consonants. This is so opposed to all analogy, and so unsupported by the example of the modern languages, that it may be as well to examine the instances adduced before proceeding to inquire into the three regular processes mentioned above.

v1.c6.20

Weber expresses himself as doubtful whether the instances of hardening found in the Bhagavati be really genuine, or mere errors of the writer of the manuscript. The instances are *jati=yadi* and a whole series of words in which it is dimly and indistinctly conjectured that the corruption of a certain Sanskrit word would have produced a media, whereas it is found in the text with a tenuis. These are not decisive proofs; though ingenious, they are based upon conjecture. *Jati=yadi* is the only *bonâ fide* instance, and that appears to be an error of the scribe, who has a partiality for *t*, and sticks it in where it has no right to be. In Saptaśataka, too, there are only conjectures; as *dhakkei* for *sthagayati*, where the modern धांकना and झांकना "to cover," seem rather to point to a separate root. It does not follow, because a Prakrit word is interpreted in Sanskrit by a somewhat similar word, that it should be connected therewith: *lukka* is supposed to be either for *lagna* or *yukta*, but there is a root *luk* which will answer the purpose better than either. From Vararuchi are quoted *vachchai* for *vrajyati*, *ruchchai* for *rudyate*, *chakkh* for *yaksh* (*jaksh*). These are also vague, and the affiliation of the Prakrit words is not certain. Upon the whole, then, it appears that there are only a few instances of this process, and the majority of those few are doubtful, so that no rule can be founded on them. The analogy of the languages is against it; and where the manuscripts are so carelessly written, and the forms of some of the letters differ so much from those in use in printed works as to mislead the reader, it is safer to suppose a mistake in writing than a systematic violation of analogy.¹

v1.c6.21

¹ Even if it be conceded that the instances quoted are real cases of hardening, it must at the same time be admitted that they are too few to establish a rule, and the discussion of them has therefore been shortened in this work, which, as before mentioned, deals only with Prakrit as a secondary subject in so far as it throws some, though often a confused and misleading, light on the modern languages.

v1.c6.22

§ 51. (1.) Proceeding then to the first class in which a single letter is retained unaltered, we find that च is the most tenacious of all the tenues. We find पचनं, रचनं, वचनं, मोचनं, सोचनं, याचनं, सेचनं, रुचि, in the modern languages, though in Prakrit they all lose the च. Thus, *raāṇam* appears for *rachanam*, though it also represents *radanam*, or even *ratna*. The confusion introduced into Prakrit by this omission of consonants renders it very difficult to believe that these dialects were ever really spoken languages.

v1.c6.23

Vararuchi's rule (ii. 2) is to the effect that the lenes of the four organs, excluding the cerebrals, are elided: but Bhâmaha's comment infers from the use of the word *prâyas*, or "generally," that where euphony is satisfied there need be no elision, and he adduces the following words in proof of his inference: *sukusamam*, *piagamaṇam*, *sachâvam*, *avajalam*, *atulam*, *âdaro*, *apâro*, *ajaso*, *savahumânam*, for *sakusuma*, *priyagamana*, *sachâpa*, *apajala*, *atula*, *âdara*, *apâra*, *ayaśas*, *sabahumâna*. But all these words are compounds, and it is observable that there is much irregularity in their case. Sometimes the first consonant of the latter part of a compound is treated as though still an initial, and escapes elision; while in other cases it is treated as no longer an initial, but a medial, and is elided accordingly. The decision seems to depend on the degree of amalgamation attained by the two elements of the compound; in a word which is well known and frequently used the consonant is elided, showing that though a compound it had got to be treated as a single word; but in those words whose two elements have not so coalesced, as, for instance, in occasional compounds, a sentiment of the separate existence of the two words has operated to preserve the initial of the second from elision. Thus, in a common compound like *supurusha*, a word of every-day occurrence, the *p* is elided, and we have *suüriso*; whereas in the less commonly used compounds quoted above the initial is preserved. I do not think euphony has anything to do with it, because *supurusha* is quite as easy to pronounce as *sukusuma*.

v1.c6.24

Apart from compounds, however, we find instances of retention. Thus, (*α*) tenues: *kotuhallâiṃ*, *kapolam*, *iti*, *vatuna*, *sutâ*, *âpelo*, *satam*, *samiti*, and a few others, for *kutûhalâni*, *kapola*, *iti*, *patana*, *sutâ*, *âpîḍah*, *śatam*, *samiti*; (*β*) mediæ: *juguchha*, *gagaṇa*, for *jugupsa*, *gagaṇa*, and the like. But these are exceptions. The language of the Bhagavatis, like Pali, retains single consonants much more frequently than scenic Prakrit, or works written in imitation of it.

v1.c6.25

In the modern languages, even in Tadbhavas, retention is to be found; as in—

v1.c6.26

Skr. क्षिति “earth,” H., P. क्षिति. Skr. क्षेप “moving,” खेप, in all. Skr. जट “matted hair,” जटा, in all. Skr. जपनं “muttering,” H. जपना, जंपना, M. जाप, जापणीं.

v1.c6.27

Skr. ज्योतिः “light,” H. जोत, जोति, S. *id.*, G. जोत, P. जोतना. Skr. तिलक “sectarial mark,” H. टीका, M., B., O. टिका, S. टिको. Skr. स्थगन “covering,” H. ठगना, and so in all. Skr. यजमान “a client,”¹ H. जिजमान, in the rest जजमान. Skr. स्फाटिक “alum,” H. फिटकरी, M. फटकी, S. फिटको, the rest फटकरी or ॰डी.

v1.c6.28

¹ Literally a man who gets Brahmans to hold a sacrifice for him and pays for it, but in modern times applied to any one who has a right to certain services, as those of the barber, shoemaker, etc.

v1.c6.29

Although these words are classed as Tadbhavas, yet it is observable that they differ very little from the Sanskrit form, and are in many cases modern, though not universally so.

v1.c6.30

I am disposed to think that a single consonant is more often preserved when followed by a long or accented vowel and preceded by a short or unaccented one, than when the reverse is the case. This supposition will be reverted to when all three forms of treatment have been reviewed, and is introduced here in order that the reader may bear it in mind through the next few pages.

v1.c6.31

It is curious that the letters which we should suppose to be the weakest and most liable to rejection are precisely those which keep their ground most persistently. The nasals, semivowels, sibilants, and ह are almost always retained. Thus, in Prakrit *kâla*, *komala*, *kusuma*, *kula*, *pavaṇa*, *paaví*, *paṭhama*, *palâsa*, *mahilâ*, *gaha*, exhibit these letters in their original state, while if any of the letters of the four vargas had been in those situations they would have been softened or rejected, or in some way or other changed, as will be seen in the next two sections. The modern languages follow the Prakrit lead in this respect with but few deviations. The nasal is the most ineradicable of all; so much so that the insertion of an anuswâra even will generally suffice to preserve a single consonant which would otherwise have disappeared. Compare such words as ताल, तारा, तान, with ताप, तात, काक, and it will be seen that while the former preserve their semivowels and nasal unchanged throughout all sorts of compounds and derivatives, the latter change their medial consonant in many ways. Thus, from ताल “hand-clapping,” come H. ताल, ताली, and so in all; but from ताप “heat,” we get either ताव or ताउ. It may be surmised that in the effeminate speech of those days, or rather in the fantastically refined utterances of that particular school of writers by whom the plays and Prakrit songs were written, the strong consonants seemed too harsh and grating, and were therefore omitted, while the soft liquid semivowels were retained as not being too stiff or hard to break the delicate warbling cadences in which they so much delighted; and even in the speech of the masses somewhat of this feeling must have prevailed, as we find it to a certain extent true of prose works which are written in a severer style and without any great seeking after euphony.

v1.c6.32

The changes which the nasals, semivowels, and sibilants undergo are of an organic, not of a positional nature, and will be discussed in their proper place.

v1.c6.33

§ 52. (2.) The softening of tenues to mediæ is a very frequent characteristic of genuine Tadbhavas, but it occurs more regularly in some letters than in others; *k*, *ṭ*, *p*, go regularly into *g*, *ḍ*, *b*, but changes from *c* into *j* are rare, and those from *t* into *d* nearly equally so. Examples are—

v1.c6.34

*k* into *g*.

v1.c6.35

Skr. काक “a crow,” Pr. काओ, H. काग, P., M., G., B. *id.*, S. कांगु, G. dimin. कागडो.

v1.c6.36

Skr. शाक “potherb,” Pr. साओ, H. साग, P. *id.*, S. सागु.

v1.c6.37

Skr. शकट “cart,” Pr. सअढो, H. सगड, but B., O. छकडा, and vulgo हकडी “hackery.”

v1.c6.38

Skr. कंकाल “skeleton,” H. कंगाल “starving,” and so in all.

v1.c6.39

Skr. कंकण "bracelet," Pr. कंकणो, H. कंगन, and so in all. Skr. पर्यंक "bedstead," Pr. पज्जंको, H. पलंग, and so in all.

v1.c6.40

च into ज.

v1.c6.41

Skr. कुंचिक "key," H. कुंजी, P., S., M. *id.*, B. कूजी, O., G. कुंची, O. also कुंजी. Skr. काच "glass," M. काज (vulg.).

v1.c6.42

ट into ड.

v1.c6.43

Skr. कटि "worm," Pr. कीडो, H. कीडा, B., O., P. *id.*, M. कीड, किडा, G. कीडो, S. किडो. Skr. कर्पट "cloth" [Pr. कप्पडो], H. कपडा, P. *id.*, S. कापडो, G., M. कापड, B., O. *id.* Skr. कटाह "pan" [Pr. कडाहो], H. कडाहो, P., S. *id.*, G. कढा (कडहा) कढई, O. कराई, कह्राइ, कढेइ, B. कड, कडाई. Skr. वट "banyan-tree," Pr. वडो, H. बड, and so in all, but M. and G. वड. Skr. भाटक "hire," H. भाडा, S. भांडा, in all as H. Skr. घट "jar," Pr. घडो, H. घडा, M. घडी *id.*, B. *id.*, and so in all. Skr. घोटक "horse," Pr. घोडओ, in all घोडा.

v1.c6.44

त into द.

v1.c6.45

जामाता "son-in-law," Pr. जामाअ, H. दामाद (but through Persian داماد).

v1.c6.46

प seldom stops short at ब. It would appear that ब itself differed very little in sound from व in most parts of India. प therefore modulates into व, and still further into उ, and sometimes combines with a preceding *a* or *ā* into ओ, or औ. Thus:

v1.c6.47

Skr. स्वपन¹ "sleeping," Pr. सिविणो, सिविण्णो, H. सोना, P. सौणा, S. सुम्हणु, G. सुवुं, B. सोइते, O. शोइबा.

v1.c6.48

¹ I use this form, and not स्वप्न, because the latter contains a mixed nexus, and the nasal would consequently disappear; the words in the text could not come from such a form.

v1.c6.49

Skr. वपनं “sowing,” H. बोना, B. बूनन, O. बोइबा. Skr. शपथ “oath,” Pr. सवहो, H. सौंह, P. सोहं, S. सुंहं. Skr. छेपनं “moving,” Pr. छेप्पणं, H. खेवना “to row,” P. id., G. खेवा, M. खेवा “rowing.” Skr. कपाट “door,” Pr. कपाटो, H. कवाड, P., M. id., B., O. कबाट. Skr. कपई “cowrie,” H. कौडी, M. कवडी, and so in all. Skr. ताप “heat,” H. ताव, ताउ, and so in all. Skr. सपत्नी “co-wife,” Pr. सवत्ती, H. सौतिन, सौत, M. सवत. Skr. सपादिक “with a quarter,” H. सवा, and so in all.

v1.c6.50

Vararuchi does not make this rule general in Prakrit, but confines its operation to *t*, *p*, and *t*, giving as examples *udú*, *raadam*, *âado*, *nivudí*, and others, for *ritu*, *rajatam*, *âgata*, *nivriti*, etc. He makes transition from *p* to *v* general, and gives instances: *sâvo*, *savaho*, *ulavo*, *uvasaggo*, for *śâpa*, *śapatha*, *ulapa*, *upasarga*. Upa is universally changed into *uva*, and even *ua* (see § 53). The change of *t* to *d* is illustrated by *nado*, *vidavo*, for *naṭa*, *riṭapa*, but there are hundreds of instances to be found in Prakrit works.

v1.c6.51

§ 53. (3.) Elision is in Prakrit the rule; retention and weakening, to a certain extent, the exceptions. Vararuchi’s rule (ii. 2) is very sweeping, and includes all the unaspirated letters of the four organs, except the cerebrals, as stated before (§ 51). *य* and *व* are added probably because they are so closely connected with *ज* and *ब* respectively. The instances given are *maülo*, *ṇaülo*, *sâaro*, *ṇaaram*, *vaaṇam*, *súi*, *gao*, *raadam*, *kaam*, *viâṇam*, *gaâ*, *mao*, *kaí*, *viulam*, *vâüṇâ*, *ṇaạṇam*, *jiam*, for *makula*, *nakula*, *sâgara*, *nagara*, *vachanam*, *sûchi*, *gaja*, *rajatam*, *kṛita*, *vitâna*, *gaḍâ*, *mada*, *kapi*, *vipula*, *vâyuṇâ*, *nayana*, *jîva*.

v1.c6.52

The confusion arising in Prakrit from this constant elision is extraordinary; thus, *vaaṇa* stands for *vachana*, *vadana*, *vapana*; *vaa* for *pada*, *vayas*, and *Vraja*; *râi* for *râji*, *râtri*; *raa* for *raya*, rajas, rata; and the accumulation of vowels with no intervening consonant is in striking contrast to the Sanskrit, which tolerates no hiatus. Juaïjana जुआइअण = yuvatijana, उआइआरअ uaâraa = upakâraka, uaa = udaka, aïrahaa = abhirataka, aïujjua = atyrijuka; so that we seem to be listening to some Maori or other Polynesian dialect, rather than to anything Aryan; and I cannot bring myself to believe that the people of India at any stage of their history ever spoke such a form of speech as this.

v1.c6.53

In the modern languages instances of elision are tolerably frequent, but they do not result in hiatus to such an extent as in Prakrit. Either one of the vowels goes out with the consonant or the two vowels which are left behind coalesce into one, or hiatus is avoided, as it is also in some kinds of prose Prakrit, by the insertion of य, व, or even ह. For the treatment of vowels in hiatus see §

v1.c6.54

क.

v1.c6.55

Skr. कोकिल “koil,” Pr. कोइल्लो, H. कोइल, P., S. id., G. कोयळ, O. कोयिल.

v1.c6.56

Skr. सुवर्णकार “goldsmith,” Pr. सोण्णारो, H. सोनार, सुनार, P. सुनिआर, S. सोंनारो, M. सोनार. Similarly are all names of trades ending in कार treated, as कुम्भकार “potter,” and others.

v1.c6.57

Skr. मुकुल “bud,” Pr. नउलो, H., B. मौल.

v1.c6.58

Skr. नकुल “weasel,” Pr. णउलो, H. नेवल, P., B., O. नेउल, G. नोळियुं, S. नोळू.

v1.c6.59

Skr. नारिकेल “cocoa-nut,” Pr. णारिएलो, H. नारियल, M. नारळ, G. नारियळ, P. नरेलु, नलेरु, S. नारेलु and नाइरु.

v1.c6.60

ग.

v1.c6.61

Skr. द्विगुण “twofold,” Pr. दूउणो, H. दूना, P. id., S. ड़ूणा, M. दूण.

v1.c6.62

Skr. नगर “city,” Pr. णअरो, नयरी, H. नैर,¹ G. नयरी.

v1.c6.63

¹ In Old H. नयर. The word is seen in the terminations of a few names of towns, as Bikaner, Bhatner.

v1.c6.64

Skr. सुगंध “fragrant,” Pr. सुअंध, H. सींधा, P. *id.* Skr. भगिनी “sister,” H. बहिन (for भइन्), P. भैण, बैन्ह, S. भेणु, G. बेहेन, M. बहीण, B. बह्न, O. भौणी, भउणी.

v1.c6.65

च.

v1.c6.66

Skr. सूची “needle,” Pr. सूई, H. सूई, P., S., M. *id.*, G. सोय, O. and B. retain च. Skr. काच “glass,” S. काचो.

v1.c6.67

ज.

v1.c6.68

Skr. राजा “king,” Pr. राआ, H. राउ, राव in all, so also in राजकुल = राउल, राजदूत = राउत, राजपुत्र = रावत. Skr. बीज “seed,” H. बिहन, M. बी, बीं, P. बीह, S. बीहणु, O. बिहन.

v1.c6.69

त.

v1.c6.70

Skr. पिता “father,” Pr. पिआ, H. पिउ. Skr. माता “mother,” Pr. माआ, H. मा, माइ, माउ. Skr. भ्राता “brother,” Pr. भाआ, H. भाई. Skr. घात “wound,” Pr. घाओ, H. घाव, घाउ, so in all. Skr. शत “hundred,” Pr. सअ, H., P. सौ (सउ) सै. } See Chap. II. § 43.

v1.c6.71

द.

v1.c6.72

Skr. खादनं “eating,” Pr. खाअणं, H. खाना, P. खाणा, S. खाइणु, M. खाणें, G. खावुं, O. खाइबा, B. खाइते. Skr. रोदनं “weeping,” Pr. रुअणं, H. रोना, P. रोणा, S. रुअणु, G. रोवुं, B. रोइते. Skr. हृदय “heart,” Pr. हिअअं, H. हिया, P. हियां, हिआउं, S. हिँआँउ, M. हिव्या. Skr. कदली “plantain,” H. केला (कयला), P., S., B. *id.*, S. केविडो (diminutive), G. केळ, केर, M. केळ.

v1.c6.73

प.

v1.c6.74

Skr. प्रापनं “getting,” H. पाना, though also पाउना, P. पाउणा, S. पाइणु, G. पामवुं, M. पावणे, B. पाइते, O. पाइबा.

v1.c6.75

Skr. कूप "a well," H. कूआं, P. id., S. खुह, G., M. कुवो, B., O. कूआ. Skr. पिपासु "thirsty," H. पियासा, so in all. Skr. दीप "lamp," H. दिया, P. दीआ, M. दिवा, G. दिवो,

v1.c6.76

ट and ड are never elided; in non-Aryan words they would naturally hold their own, and in Aryan words they would generally spring from त्र, द्र, or र्त, र्द, and so being already, as it were, on the second step of development, they would not ordinarily be any further corrupted, except in the case of ड, which being now in the majority of cases pronounced as a harsh *r*, is not unfrequently confounded with र; its further change into ळ comes under the head of organic changes.

v1.c6.77

With regard to ब, it must be observed that in Prakrit there is much difficulty in distinguishing between it and व. It is not correct to say that they are quite identical, however, as some words are always written with व, others always with ब. Cowell, in his edition of Vararuchi, makes no distinction, putting both letters under व, and he is to a certain extent justified in this course by his author, who is very hazy on the subject. Bengali and Oṛiyā among the moderns are the only two languages which make no distinction between these two letters, but they make them both into ब. The only notion the Bengali or Oṛiya peasant has of *v* is that it is the same as *w*, which again is to him only *u* pronounced quickly between the two vowels (see what has been said on this subject in Chapter I. § 23). If we wish really to know which words ought to be spelt with *v* and which with *b*, we must go to the Marathi and Gujarati, which keep the two sounds distinct.

v1.c6.78

Thus, Sanskrit has बदर, and not वदर; though Cowell gives बोरं as the Prakrit, it clearly ought to be वोरं, and M. has accordingly बोर, and not वोर.

v1.c6.79

It is probable, however, that though the distinction undoubtedly exists even in Prakrit, it was not very carefully observed, and if the harder प was softened into a vowel, it is not likely that व would escape. Examples are not to be looked for so much in Bengali and Oriya as in the western languages.

v1.c6.80

व when elided leaves its mark behind in the labial vowel o, and य similarly in the palatal vowel e, in cases where they are both preceded and followed by a, as in ava, aya.

v1.c6.81

अव, Prakrit oâra=avatâra, oâsa=avakâśa, oinna=avatirṇa, ohi=avadhi; but ava, arising from softening of apa, does not undergo further change, as avasaüṇa=apaśakuna, avarâha=aparâdha=, avaraṇha=aparâhṇa.

v1.c6.82

अय in Prakrit occurs most frequently in the causal verb; thus: rochemi rochemo, rochesi rochetho, rochedi rochenti, for Skr. rochayâmi, rochayasi, rochayati, etc.

v1.c6.83

In other positions, however, aya not unfrequently becomes aä by simple elision of the y; examples are jaä=jaya, jâä=jâyâ, aäso=ayaśas, raä=rayas. In these cases it was probably pronounced as j, just as it is in the present day in many parts of India. In the moderns no such process as this is to be found.

v1.c6.84

§ 54. It is now necessary to inquire why these three processes—retention, weakening, and elision—exist side by side, and what is the law which decides in every case which process shall be followed. It is easy to talk, as some authors do, of the "lawless licence" of Indian etymology; but this is only a confession of ignorance; it amounts to saying that because we cannot find the reasons for any particular change, therefore there are no reasons at all; the blind mole says there is no sun because he cannot see the daylight. Reasons there must be, and it is our business to try and find them out; or at any rate in this early stage of inquiry into the elements of the modern Indian languages, we may perhaps be satisfied if we can point out some slight indications which, if followed up hereafter, may lead

v1.c6.85

later inquirers to a discovery of the full and perfect system. The following hints are given in this hope, and do not pretend to be anything more than hints.

v1.c6.86

The cases of retention from their form are nearly all seen to be Tatsamas, or such very late Tadbhavas that they have not yet had time to make any great divergence from the Tatsama form. The principal difficulty lies between weakening and elision. The Prakrits may be cleared away at once by saying that they always elide, and we may further get rid of the cerebrals, which are never elided. The inherent weakness of य, which leads it almost always to be softened into the semi-vowel, places this letter also on a different footing from the rest. When it has become व, and by a step further औ, ओ, and उ, its total elision becomes rather a question of vowels than of consonants. Thus, प्रापनं having become पाउना, the further change to पाना is a matter of vowels, and more especially in those languages where the verbal base ends in a vowel, while the termination begins with one, as in the case of Bengali ite, ilâm, or Sindhi iṇu, iṇdo; where, to avoid too great a clash of vowels, the u naturally disappears. The same remarks apply in a still greater degree to व, as has been pointed out in the preceding section. The semivowels, nasals, sibilants, and ह, do not come within the scope of this inquiry, as they are seldom if ever elided, except य and व, for which we should probably understand ज and ब, and they cannot be weakened, as they have no corresponding weak letters, being in fact mediæ themselves; consequently their changes are not positional but organic.

v1.c6.87

Having cleared the way somewhat by getting rid of the above-mentioned classes, there remain क, ग; च, ज; त, and द; and in the case of these six letters the rule appears to be that they are generally elided when preceded by a long or accented vowel, generally retained if mediæ, or weakened into mediæ if tenues, when preceded by a short or unaccented vowel.

v1.c6.88

Contrast *kókila*, *khádanam*, *ródanam*,—where the preceding vowel is long, and in which therefore the consonant is elided, sometimes even together with the following vowel, as *koïl*, *khânâ*, *ronâ*,—with *kapóta*, *kadâ*, *chakóra*, which are followed by a long vowel, and in which, therefore, the consonant is retained, as *kapot*, *kad*, *kadhi*, *chakor*. By the operation of this rule *hṛidaya*, *kádali*, *prápanam*, *kûpa*, *râjâ*, *sûchi*, *vîja*, lose the single consonant altogether, as shown in the last section. It would further appear that when two long or accented syllables come together, the intervening consonant goes out. Thus, in the class of words expressive of trades, *kumbhákâra*, *sûtrádhâra*, become *kumhâr* (through *kumhaâr*), *chhutâr*; and even where a short vowel intervenes, as *suvárṇakâra* = *sonâr*, *lóhakâra* = *lohâr*. Again, there are cases where the word having been in existence in Prakrit has elided its consonant in accordance with Prakrit rules, such as *mákula* = *maul*, *maül*, *sugandhá* = *saundhá*. Even here there is sometimes a tendency to revert to the rule above, as in *nakulá*, “a weasel,” which should by rule retain its consonant; having, however, lost it in Prakrit, the preceding vowel is lengthened, and we have *neval*, *neul*, etc. The words *pitâ*, *mâtâ*, are oxytone, but having lost their *t* in Prakrit they remain without it in modern times. There might seem to be an exception in *dviguṇa*, but *dvi* has first become *dú*; thus, *dúguṇa* is like *kokila*, and the consonant goes out. Further instances are *súkara* = *súar*, *káka* = *kawwâ* (i.e. काउ + आ, the form *kág* given in § 52 is rare and well-nigh obsolete), *sitakâla* = *siáro*, S., (i.e. *siaáro*), where the *t* is elided through the preceding long vowel, and the *k* through Prakrit influence, *jágaran* = *jâgnâ*.

v1.c6.89

On the other hand, the consonant is retained in a number of words derived from Sanskrit causals because the accent is on the first vowel of the causal characteristic: *bhedáyati*, *ropáyati*, *mápáyati*, form *bhejnâ*, *ropnâ*, *mâpnâ*. In the causal verbs which retain still a causal signification, the causal characteristic appears as long *â*, as in *chalâna*; in the above verbs it does not appear, though it leaves its mark behind in the retention of the consonant, which would otherwise be rejected, because preceded by a long vowel. Compare *ronâ*, from *rodanam*, with *ropnâ*, from *ropáy-*; also *chhânâ*, from *chhâdanam*, with *bâjnâ*, from *vâdáy-*.

v1.c6.90

The difficulty in this theory is that it was formerly stated (Chap. I., § 7) that early Tadbhavas—i.e. words which had come down through the medium of Prakrit—were distinguished by the fidelity with which they retained the accent, and it is part of this hypothesis that Prakrit also retained the accent. Yet here we find words retaining the accent, that is, showing traces of having felt its influence, and, therefore, having been in existence at a time when the Sanskrit accent was still known and heard, and which should, therefore, agree in form with Prakrit words, which yet do not agree in form with Prakrit. The words in § 53, where elision is practised, do, it is true, agree, but not those in § 52, where the consonant is merely weakened. Such forms, for instance, as H. *sagar*, Pr. *saadho*; H. *sâg*, Pr. *sâo*, seem to militate against the above theory. To this it may be replied, that the fact of the divergence of Prakrit in this respect from the modern languages is an additional argument in favour of the theory of the unreal and merely literary character of the constant elisions in that language, and that it is chiefly in Prakrit poetry that these elisions are found; in Prakrit prose they are much less frequent, and in Pali and the earlier forms of Prakrit they are almost unknown. Still, I must confess that this theory of the effect of the accent and the relation between long syllables and the preservation or elision of consonants, though it will probably eventually turn out to be correct, is at present in rather a crude state, and will require to be worked out at greater length when fuller materials are available.

v1.c6.91

§ 55. The rejection of the final inherent *a* in the very large

v1.c6.92

class of Sanskrit nouns which end with that vowel, has had the effect of producing a great number of words with consonantal endings in the modern languages; and, following Grimm's law, we should expect that Sanskrit words ending in ga, ja, da, ba, would have the consonants hardened to k, ch, t, p, respectively. The fact is, however, the reverse. Occasionally we find instances where this does take place, as in Marathi दाप for दाव, but these are not numerous. The cause of this appears to be that the final a was retained till very recent times. In poetry it is even now required to be pronounced, and in Bengali and Oriya, though not heard in ordinary rapid talking, directly a man speaks slowly and distinctly, the short final a, there changed to ŏ, becomes audible. Thus, it has happened that these letters have always been regarded as medials, and treated as such, with a tendency to weakening rather than strengthening. On the other hand, in those Sanskrit nouns which end in a consonant, it is generally only the nominative case to which the description applies; the other cases having vocalic case-endings lose the consonantal type, and in Prakrit (Var. iv. 6) we have the absolute rule that a final consonant is always elided. Thus, Skr. सरित “a river,” becomes in Old Hindi सरिता.

v1.c6.93

The majority of instances of consonantal endings wherein a soft or sonant letter has been hardened is to be found in Panjabi, where Persian and Arabic words have been so long in use. These words having in those languages a true consonant ending have in some instances been hardened. It is true that in Arabic the nouns have technically their vowel case-endings in the shape of tanwîn; but, as is well known, tanwîn has for many centuries been a mere grammarian's fiction. No Arab ever says rajulun, rajulin, rajulan, in conversation, whatever he may do when reading the Kur'an; so that practically these words met the ear of the Panjabi as true consonantly terminated words, and he has hardened the final consonant in some cases. Thus, he says किताब for کتاب, मसीत for مسجد, सहित for شہد, and the like. The same practice is also occasionally found in Aryan words, as पत “dignity,” from पद; सेप “service,” from सेव (i.e. सेब). The same process exists also in Sindhi, in spite of the fact that in that language all nouns without exception end in a vowel. This vowel is often so short and light as hardly to be audible.

v1.c6.94

But on the whole the hardening of final consonants is rare and not sufficiently regular to constitute a rule, though we are justified by analogy in supposing that if consonantal endings were more frequent the process of hardening would be more often evident.

v1.c6.95

§ 56. Organic changes for the most part operate without reference to position, being found nearly as often in initial as in medial consonants. They are also in many cases confined to particular languages or dialects.

v1.c6.96

In the case of the gutturals there appear to be no organic changes in the Prakrits or modern languages, with the exception of the compound च (क + ष). In Sanskrit a connexion appears to exist between the letters of this organ and the palatals. When a verb beginning with a guttural is reduplicated, the corresponding palatal is used; as कुट् “to sever,” = चुकुट्; गम् “to go,” = जगम्. It is probably this practice, taken in connexion with the similar custom in nouns of changing च when terminating a base into क before certain case-endings, as वाक्, acc. वाचं, that has led to the commonly received idea that the genitive postposition in Marathi, चा, is derived from or connected with the corresponding Hindi का. If this be so, we should expect to find that क was regularly replaced by च in Marathi. After considerable search, however, I am unable to find any such instance. If, then, the above supposition be correct, it must be an isolated case.

v1.c6.97

It must be remembered that the modern languages have almost entirely abandoned the Sanskrit verbal system, and form their verbs from one or two tenses only of the Sanskrit verb, chiefly from the present tense and certain participles and verbal nouns, so that the reduplication of the ancient verb would not be reproduced in modern speech. Similarly the noun takes for its base one form, and that generally the nominative case of the Sanskrit; so that here also the euphonic changes required by the structural peculiarities of that language would not be reproduced.

v1.c6.98

The solitary instances of H. काका and चाचा “uncle,” and O., B. चास = ऋषि, are all that can be brought forward, and काका, though an Aryan word, comes into the modern languages through the Persian.

v1.c6.99

§ 57. In connexion with the palatals, however, there is another and, at first sight, less explainable tendency. In a considerable class of words they modulate into cerebrals or dentals, that is to say, into one or other of the departments of the lingual range of sounds. The instances of जामाता, H. दामाद, and घी = झी, have been mentioned above (§ 49). A more widespread example is afforded by a class of words meaning “to press,” “stamp,” and the like. The earliest type of this group is perhaps the Skr. root कप्, or कंप, which is said to mean “to go”; but after making all due allowances for the copiousness of Sanskrit, every third root in that language can hardly mean “to go,” though the lexicographers calmly assert that it is so. Perhaps this root is only a dialectic form of चप्, the causal of चि “to strike,” which would well enough agree with the modern meanings, “to stamp,” “press,” “tap,” etc., in this way that stamping would naturally be defined as the act of causing a seal or stamp to strike the paper or other article. This idea will not appear unreasonable or far-fetched to those who remember how constantly ideas which in European languages are expressed by simple verbs

v1.c6.100

are in the Indian languages rendered by causals, or rather by words which still retain a causal form; thus, "to call" or "summon," बुलाना, causal of बोलना, *i.e.* "to cause to speak," as उसको बुलाओ "call him," in full, "cause him to (come and) speak (to me)"; डुबाना "to drown anything," causal of डूबना "to sink," *i.e.* "to cause anything to sink"; बचाना "to rescue," causal of बचना "to escape," "be saved," *i.e.* "to cause one to escape"; समझाना "to explain," from समझना "to understand," *i.e.* "to cause to understand"; and very many others.

v1.c6.101

With the palatal initial, then, we have: Hindi छाप "a stamp" or "seal"; छापा "a stamp," "an edition of a work," also the sectarial marks stamped by Hindus on various parts of their bodies; छापना "to print"; छापती "a seal" or mark made of cowdung and put on a heap of grain to prevent its removal; छप "a splash," or sound made by an object striking the water, and derivative छपाका "a splash," "squash"; छपक the same; छपकाना "to dash or splash water"; छपरी "a puddle"; and other words.

v1.c6.102

As one of the senses of छिप is "to disappear," the causal would mean "to cause to disappear," *i.e.* "to hide," hence with the *i*-sound comes छिपाना "to hide," छिपना "to lie hid," छिपाव or छिपाउ "concealment." These words are also written with *a*; from the cognate idea of "covering" comes छप्पर "a thatch" or "thatched roof," छप्परखट "a bedstead with curtains," छिपकी "a lizard," from its hiding in crevices of walls, etc.

v1.c6.103

With rejection of the aspirate, by no means an unusual process in the vulgar speech, we get a long array of words, which may, however, be referred also to Sanskrit चापय, causal of चि "to heap," "to collect." This चि, however, is probably even in Sanskrit connected with चि, and some of the meanings of the following words agree better with the latter than with the former. It is easy to see the connexion between all these meanings. To strike, press down, press a covering down over anything, cover, conceal, hide; from pressing down by another turn of thought come the ideas of pressing down into a vessel, heaping up and pressing tight together, collecting in a heap, squeezing, and the like. From one primary idea the Aryan mind runs off down many radiating lines of thought, so that derivatives widely apart in meaning now-a-days may often be clearly traced to one central root.

v1.c6.104

¹ From this word comes the now vulgar expression, "first-chop," meaning the best kind of any article, that which bears the highest stamp.

v1.c6.105

चाप "a bow" (also in Skr.). चापडा "cake of cowdung," made by stamping and flattening the dung between the hands. चांपना "to stuff," "press," "squeeze." चांप "the lock of a gun,"—that part, to wit, which is pressed down on the nipple. It also means the stocks, or other instrument of punishment. चपाती "a chapaty," or thin cake of unleavened bread, made by patting and flattening dough with the hands. चपटा or चिपटा "flat." चपटाना "to flatten." Then a string of words with the meanings of being pressed close to, adhering, clinging. चिपटा "clammy," "viscous." चिपटना "to stick to," चिपकना the same; also "to be compressed." चिमटा "tongs." चिमटना "to cling to" (you say to a child, मत चिमटो "Don't tease!"). चुप कन "a tightly-fitting coat or cassock" (the French soutane). चपरास "a buckle," subsequently "a badge." From the idea of repression comes चपना "to be abashed" or "shamefast," "to be silent"; चुप "silence!" चुप रहना "to be silent." I omit a vast host of derivatives which would occupy several pages.

v1.c6.106

Marathi has छाप, छापणें, छापा, छपणें, छपणें, छप्पर, छप्पा; and with च, चांप, चापट "squat," "dumpy"; चापटणें "to flatten by beating"; and the usual quantity of derivatives. With i: चिपणें, चिपडा, चिपड, चिपडणें, चिपार्ट, चिप्प, चिमटना; also चिवडणें "to crush," "squeeze"; चिवडा "a mass of pulp"; चिवळणें the same as in M. छ regularly reappears as श; we may here perhaps affiliate शिवण "to thatch" (H. छप्पर), which should be kept distinct from शिवण "to sew," which is from Skr. सीवन. शिंपण "to sprinkle," is probably from Skr. क्षिप् "to throw"; but this root also may be no more than an ancient causal of क्षि "to strike." To follow up this thread would, however, lead us too far away from the present subject.

v1.c6.107

In Panjabi most of the words quoted under Hindi occur. It is useless to repeat them.

v1.c6.108

Sindhi gives छपरु "eyelid"; also the cognate sense of "thatch," which is probably the primary one, that of "eyelid" being secondary and metaphorical; छपि "the crouching of a beast of prey," as in छपि मारे विहणु "to lie in ambush"; छाप, छापणु, छापो; and with च: चापणु "to press or shampoo the limbs," चापहि "a knot" or "lump," चापटो "a treadle," चापोडो "a flat clod of earth or plaster," चपर "a wedge," चपिडासु "a chaprass," चपिडो "a lever," चिपणु "to press," "mash," चिपतिरो "flat," चिमटो "tongs."

v1.c6.109

Gujarati has the principal words given under Hindi, and perhaps the whole of them, if the dictionary-maker had only put them in his book. In Bengali are found चप्पर, छापिने, छाप, छापा, छापाइते, "to conceal"; छापी "suppressed," "concealed," छिपि "a cork" or "stopper," छिपिते "to hide," छिबरा "pulp"; and with च: चपरास, चपट "the open palm of the hand," चपेट "a blow with the palm," चपकन "a chapkan" or "cassock," चापिते "to weigh down" or "press in a vice," चाप "a clod," "block," "lump," "burden," छापाइते "to print," चापुया "curdled," "coagulated," चाबडा "clod"; with i: चिप, चिपइ "to squeeze," "express juice," "wring out water," चिमडिते "to cling to," and derivatives in crowds.

v1.c6.110

Lastly, Oriya has the same words as the other languages; also छापमुदि "a signet ring," चापडा "a slap," चापडाइबा "to slap," चपरा "muddy," "viscous," "slimy," चपषि "the running or blotting of ink on paper."

v1.c6.111

The above instances prove the existence of a large group of words with a palatal initial, which are probably connected with the Sanskrit roots क्षि "to throw" (vulgar English "to shy") with its two ancient causals क्षप° and क्षिप. This extensive group has for its fundamental meaning "to press," and includes all the varied forms of pressing, as stamping, sealing, crushing, flattening, clinging, beating, and the secondary ideas of re-pressing, suppressing, compressing, and impressing.

v1.c6.112

We also find an equally large and varied group beginning with a lingual, either त or ट, and running parallel to the palatal group in all its meanings.¹ This group contains the following leading words: Hindi टप्पा "post-office," *i.e.* place where letters are stamped, टप्पाल "letter-post"² (टप्पा + आलय = "stamping-house"), टपक "throbbing," "dripping," टपका "a drop of rain," टपकना "to drop," "drip," टिपडी "the stocks," टोपना and तोपना "to bury," "cover with earth," टाप "a tap," "sound of beating," टापा "a coop," टापना "to tap," "flatten," "beat down," तपक्क "a sledge-hammer," टीप "pressing," also "a note of hand" or "bill," टीपना "to press," ठेपी "a cork" (comp. B. क्षिप), ठप्पा "a die" or "stamp," ठपना "to beat."

v1.c6.113

¹ See also a series of words of the types डप, डब, and दब, with the central idea of "compressing," at No. (2) of § 59.

v1.c6.114

² This word, though apparently Aryan, is not much used in Northern India, though it is the common word in Madras and the South for our ordinary word dāk, —the name of the place having been extended to the whole system.

v1.c6.115

Marathi टपकना "to drip," टपटप् and other derivatives; टप्पा, टप्पाल "post-office," टाप, टापटीप, टापटेप, टापरणे "to nail or peg down," टापसी "butting," टिपण "a note," टिपणे "to note down," also "to dab," "daub," "smear," टिपरी "stocks," also ठपका, ठिपका, ठिपकणे, etc.

v1.c6.116

Sindhi टपड "tapping," टपडणु टीप, ठपणु "to beat out metals," "to seal," "to print," ठपणी "stamping," "printing," ठपो "a seal."

v1.c6.117

Bengali टपटपानि "dripping rain," टापर "a tap," टापा “a coop,” टिप, टीप, and टुप as verbal roots with the same meanings as above, टीप “a bond” or “note of hand,” टेप, टेपिते “to squeeze,” टोपाइते “to distil,” टोपान “fallen in drops,” “distilled.”

v1.c6.118

Oriya टिपिबा “to press,” “pinch,” टिप “a bond,” टिपणा “a finger mark” or “notch.”

v1.c6.119

The other languages, Panjabi and Gujarati, have in general the same class of words as Hindi.

v1.c6.120

It is evident that there is some similarity and, to all appearance, close connexion between these two groups of words. The latter group has the sense of pressing, stamping, tapping, beating, dropping, dripping, and the like, which are too akin to the senses of the group in च to be mere accidental resemblances.

v1.c6.121

A few other instances may here be added:

v1.c6.122

Skr. चंचु “beak,” H. चोंच and ठोंठ, O. घण्ट, थोंट, B. ठोंट.

v1.c6.123

H. टाट “canvass” (perhaps Skr. चाच), B. टाट and चट, O. चट, H. चटाई.

v1.c6.124

H. छोकडा “boy,” O. टोका (Skr. शावक).

v1.c6.125

Skr. कृषि “cultivation,” O. चास and तास.

v1.c6.126

We may also compare with this the change from H. चालीस “forty,” to इकतालीस “forty-one,” पैंतालीस “forty-five.” This change takes place throughout the forties in Panjabi, Sindhi, Gujarati, but not in Marathi or Oriya, and only in two words, तेतालिस, पैंतालिस, in Bengali.

v1.c6.127

The substantive verb in Oriya has two forms, अछि, and अटे or अटइ, and it might be thought in the light of the remarks in this section that there was some connexion between the two. This would, however, be an erroneous supposition. The former, like B. आछे, Tirhut छे, G. छे, and several others, is from अच्छति, Prakrit from a Skr. root अच्छ “to appear”; whereas the latter is from वर्तते, root वृत. The व is still preserved in B. बटे; this is shown by the fact that this tense has ट in all three persons, sing. अटे, अट, अटे, pl. अटटुं, अट, अटन्ति, which it could not have were it derived from चक्. It may not be premature here to state that I find four Sanskrit roots regularly used as auxiliaries or substantive verbs by the moderns, viz. अस्, अच्छ्, भू and स्था, and that the root वृत् is not used in this way by any of the languages: thus while O. achhi is in Oriya used to define a present or past, just like the H. hai, as in heüachhi “he is becoming,” dekhi achhi “he has seen,” ate is used alone, and never as an auxiliary; in fact, in Bengali अटे, the sole remaining trace of this verb, has nearly lost its verbal meaning, and is now merely an emphatic affirmative, “it is so,” “truly,” “indeed,” and the like.

First 127 of 461 passages. Read on

Cites (15)