We learn from our Greek grammars that the second syllable in δείκνῡμι is long, but in the plural, δείκνῠμεν, it is short. This cannot be by accident, and we may observe the same change in δάμνημι and δάμναμεν, and similar words. Nothing, however, but a study of Sanskrit would have enabled us to discover the reason of this change, which is really the accent in its most primitive working, such as we can watch it in the Vedic Sanskrit, where it produces exactly the same change, only with far greater regularity and perspicuity.
Why, again, do we say in Greek, οἶδα, I know, but ἴσ-μεν, we know? Why τέτληκα, but τέτλαμεν? Why μέμονα, but μέμαμεν? There is no recollection in the minds of the Greeks of the motive power that was once at work, and left its traces in these grammatical convulsions; but in Sanskrit we still see, as it were, a lower stratum of grammatical growth, and we can there watch the regular working of laws which required these changes, and which have left their impress not only on Greek, but on Sanskrit, and even on German. The same necessity which made Homer say οἶδα and ἴδμεν, and the Vedic poet véda and vidmás, still holds good, and makes us say in German, Ich weiss, I know, but wir wissen, we know.
All this becomes clear and intelligible by the light of Comparative Grammar; anomalies vanish, exceptions prove the rule, and we perceive more plainly every day how in language, as elsewhere, the conflict between the freedom claimed by each individual and the resistance offered by the community at large, establishes in the end a reign of law most wonderful, yet perfectly rational and intelligible.
These are but a few small specimens to show you what Comparative Philology can do for Greek and Latin; and how it has given a new life to the study of languages by discovering, so to say, and laying bare, the traces of that old life, that prehistoric growth, which made language what we find it in the oldest literary monuments, and which still supplies the vigor of the language of our own time. A knowledge of the mere facts of language is interesting enough; nay, if you ask yourself what grammars really are—those very Greek and Latin grammars which we hated so much in our schoolboy days—you will find that they are store-houses, richer than the richest museums of plants or minerals, more carefully classified and labeled than the productions of any of the great kingdoms of nature. Every form of declension and conjugation, every genitive and every so-called infinitive and gerund, is the result of a long succession of efforts, and of intelligent efforts. There is nothing accidental, nothing irregular, nothing without a purpose and meaning in any part of Greek or Latin grammar. No one who has once discovered this hidden life of language, no one who has once found out that what seemed to be merely anomalous and whimsical in language is but, as it were, a petrification of thought, of deep, curious, poetical, philosophical thought, will ever rest again till he has descended as far as he can descend into the ancient shafts of human speech, exploring level after level, and testing every successive foundation which supports the surface of each spoken language.
One of the great charms of this new science is that there is still so much to explore, so much to sift, so much to arrange. I shall not, therefore, be satisfied with merely lecturing on Comparative Philology, but I hope I shall be able to form a small philological society of more advanced students, who will come and work with me, and bring the results of their special studies as materials for the advancement of our science. If there are scholars here who have devoted their attention to the study of Homer, Comparative Philology will place in their hands a light with which to explore the dark crypt on which the temple of the Homeric language was erected. If there are scholars who know their Plautus or Lucretius, Comparative Philology will give them a key to grammatical forms in ancient Latin, which, even if supported by an Ambrosian palimpsest, might still seem hazardous and problematical. As there is no field and no garden that has not its geological antecedents, there is no language and no dialect which does not receive light from a study of Comparative Philology, and reflect light in return on more general problems. As in geology again, so in Comparative Philology, no progress is possible without a division of labor, and without the most general coöperation. The most experienced geologist may learn something from a miner or from a ploughboy; the most experienced comparative philologist may learn something from a schoolboy or from a child.
I have thus explained to you what, if you will but assist me, I should like to do as the first occupant of this new chair of Comparative Philology. In my public lectures I must be satisfied with teaching. In my private lectures, I hope I shall not only teach, but also learn, and receive back as much as I have to give.
NOTES.
NOTE A.text
On the Final Dental of the Pronominal
Stem tad.
One or two instances may here suffice to show how compassless even the best comparative philologists find themselves if, without a knowledge of Sanskrit, they venture into the deep waters of grammatical research. What can be clearer at first sight than that the demonstrative pronoun that has the same base in Sanskrit, Greek, Latin, and German? Bopp places together (§ 349) the following forms of the neuter:—
| Sanskrit | Zend | Greek | Latin | Gothic |
| tat | taḍ. | τό | is-tud | thata |
and he draws from them the following conclusions:—
In the Sanskrit ta-t we have the same pronominal element repeated twice, and this repeated pronominal element became afterwards the general sign of the neuter after other pronominal stems, such as ya-t, ka-t.
Such a conclusion seems extremely probable, particularly when we compare the masculine form sa-s, the old nom. sing., instead of the ordinary sa. But the first question that has to be answered is, whether this is phonetically possible, and how.
If tat in Sanskrit is ta + ta, then we expect in Gothic tha + tha, instead of which we find tha + ta. We expect in Latin istut, not istud, illut, not illud, it, not id, for Latin represents final t in Sanskrit by t, not by d. The old Latin ablative in d is not a case in point, as we shall see afterwards.
Both Gothic tha-ta, therefore, and Latin istud, postulate a Sanskrit tad, while Zend and Greek at all events do not conflict with an original final media. Everything therefore depends on what was the original form in Sanskrit; and here no Sanskrit scholar would hesitate for one moment between tat and tad. Whatever the origin of tat may have been, it is quite certain that Sanskrit knows only of tad, never of tat. There are various ways of testing the original surd or sonant nature of final consonants in Sanskrit. One of the safest seems to me to see how those consonants behave before taddhita or secondary suffixes, which require no change in the final consonant of the base. Thus before the suffix îya (called cha by Pâṇini) the final consonant is never changed, yet we find tad-îya, like mad-îya, tvad-îya, asmad-îya, yushmad-îya, etc. Again, before the possessive suffix vat final consonants of nominal bases suffer no change. This is distinctly stated by Pâṇini, I. 4, 19. Hence we have vidyut-vân, from vidyut, lightning, from the root dyut; we have udaśvit-vân, from uda-śvi-t. In both cases the original final tenuis remains unchanged. Hence, if we find tad-vân, kad-vân, our test shows us again that the final consonant in tad and kad is a media, and that the d of these words is not a modification of t.
Taking our stand therefore on the undoubted facts of Sanskrit grammar, we cannot recognize t as the termination of the neuter of pronominal stems, but only d;24 nor can we accept Bopp’s explanation of tad as a compound of ta + t, unless the transition of an original t into a Sanskrit and Latin d can be established by sufficient evidence. Even then that transition would have to be referred to a time before Sanskrit and Gothic became distinct languages, for the Gothic tha-ta is the counterpart of the Sanskrit tad, and not of tat.
Bopp endeavors to defend the transition of an original t into Latin d by the termination of the old ablatives, such as gnaivod, etc. But here again it is certain that the original termination was d, and not t. It is so in Latin, it may be so in Zend, where, as Justi points out, the d of the ablative is probably a media.25 In Sanskrit it is certainly a media in such forms as mad, tvad, asmad, which Bopp considers as old ablatives, and which in madîya, etc., show the original media. In other cases it is impossible in Sanskrit to test the nature of the final dental in the ablative, because d is always determined by its position in a sentence. But under no circumstances could we appeal to Latin gnaivod in order to prove a transition of an original t into d; while on the contrary all the evidence at present is in favor of a media, as the final letter both of the ablative and of the neuter bases of pronouns, such as tad and yad.
These may seem minutiæ, but the whole of Comparative Grammar is made up of minutiæ, which, nevertheless, if carefully joined together and cemented, lead to conclusions of unexpected magnitude.
NOTE B.text
Did Feminine Bases in â take s in the Nominative Singular?
I add one other instance to show how a more accurate knowledge of Sanskrit would have guarded comparative philologists against rash conclusions. With regard to the nominative singular of feminine bases ending in derivative â, the question arose, whether words like bona in Latin, ἀγαθά in Greek, sivâ in Sanskrit, had originally an s as the sign of the nom. sing., which was afterwards lost, or whether they never took that termination. Bopp (§ 136), Schleicher (§ 246), and others seem to believe in the loss of the s, chiefly, it would seem, because the s is added to feminine bases ending in î and û. Benfey26 takes the opposite view, viz. that feminines in â never took the s of the nom. sing. But he adds one exception, the Vedic gnâ-s. This remark has caused much mischief. Without verifying Benfey’s statements, Schleicher (l.c.) quotes the same exception, though cautiously referring to the Sanskrit dictionary of Boehtlingk and Roth as his authority. Later writers, for instance Merguet,27 leave out all restrictions, simply appealing to this Vedic form gnâ-s in support of the theory that feminine bases in â too took originally s as sign of the nom. sing. and afterwards dropped it. Even so careful a scholar as Büchler28 speaks of the s as lost.
There is, first of all, no reason whatever why the s should have been added29; secondly, there is none why it should have been lost. But, whatever opinion we may hold in this respect, the appeal to the Vedic gnâ-s cannot certainly be sustained, and the word should at all events be obelized till there is better evidence for it than we possess at present.30
The passage which is always quoted from the Rv. IV. 9, 4, as showing gnâ-s to be a nom. sing. in s, is extremely difficult, and as it stands at present, most likely corrupt:—
Utá gnấḥ agníḥ adhvaré utó gṛhá-patiḥ dáme, utá brahmấ ní sídati.
This could only be translated:—
“Agni sits down at the sacrifice as a woman, as lord in the house, and as priest.”
This, however, is impossible, for Agni, the god of fire, is never represented in the Veda as a woman. If we took gnâḥ as a genitive, we might translate, “Agni sits down in the sacrifice of the lady of the house,” but this again would be utterly incongruous in Vedic poetry.
I believe the verse is corrupt, and I should propose to read:—
Utá agnấv agníḥ adhvaré.
“Agni sits down at the sacrifice in the fire, as lord in the house, and as a priest.”
The ideas that Agni, the god of fire, sits down in the fire, or that Agni is lighted by Agni, or that Agni is both the sacrificial fire and the priest, are familiar to every reader of the Veda. Thus we read, I. 12, 6, agnínâ agníḥ sám idhyate, “Agni is lighted by Agni;” X. 88, 1, we find Agni invoked as ấ-hutam agnáu, etc.
But whether this emendation be right or wrong, it must be quite clear how unsafe it would be to support the theory that feminine bases in â ended originally in s by this solitary passage from the Veda.
NOTE C.text
Grammatical Forms in Sanskrit corresponding to
so-called Infinitives in Greek and Latin.
There is no trace of such a term as infinitive in Sanskrit, and yet exactly the same forms, or, at all events, forms strictly analogous to those which we call infinitives in Greek and Latin, exist in Sanskrit. Here, however, they are treated in the simplest way.
Sanskrit grammarians when giving the rules according to which nouns and adjectives are derived from verbal roots by means of primary suffixes (Kṛt), mention among the rest the suffixes tum (Pâṇ., III. 3, 10), se, ase, adhyai, tavai, tave, shyai, e, am, tos, as (IV. 4, 9–17), defining their meaning in general by that of tum (III. 3, 10). This tum is said to express immediate futurity in a verb, if governed by another word conveying an intention. An example will make this clearer. In order to say he goes to cook, where “he goes” expresses an intention, and “to cook” is the object of that intention which is to follow immediately, we place the suffix tum at the end of the verb pak, to cook, and say in Sanskrit, vrajati pak-tum. We might also say pâcako vrajati, he goes as one who means to cook, or vrajati pâkâya, he goes to the act of cooking, placing the abstract noun in the dative; and all these constructions are mentioned together by Sanskrit grammarians. The same takes place after verbs which express a wish (III. 3, 158); e.g., icchati paktum, he wishes to cook, and after such words as kâla, time, samaya, opportunity, velâ, right moment (III. 3, 167); e.g., kâlaḥ paktum, it is time to cook, etc. Other verbs which govern forms in tum are (III. 4, 65) śak, to be able; dhṛsh, to dare; jñâ, to know; glai, to be weary; ghaṭ, to endeavor; ârabh, to begin; labh, to get; prakram, to begin; utsah, to endure; arh, to deserve; and words like asti, there is; e.g., asti bhoktum, it is (possible) to eat; not, it is (necessary) to eat. The forms in tum are also enjoined (III. 4, 66) after words like alam, expressing fitness, e.g., paryâpto bhoktum, alam bhoktum, kuśalo bhoktum, fit or able to eat.
Here we have everything that is given by Sanskrit grammarians in place of what we should call the Chapter on the Infinitive in Greek and Latin. The only thing that has to be added is the provision, understood in Pâṇini’s grammar, that such suffixes as tum, etc., are indeclinable.
And why are they indeclinable? For the simple reason that they are themselves case terminations. Whether Pâṇini was aware of this, we cannot tell with certainty. From some of his remarks it would seem to be so. When treating of the cases, Pâṇini (I. 4, 32) explains what we should call the dative by Sampradâna. Sampradâna means giving (δοτική), but Pâṇini uses it here as a technical term, and assigns to it the definite meaning of “he whom one looks to by any act” (not only the act of giving, as the commentators imply). It is therefore what we should call “the remote object.” Ex. Brâhmaṇâya dhanam dadâti, he gives wealth to the Brâhman. This is afterwards extended by several rules explaining that the Sampradâna comes in after verbs expressive of pleasure caused to somebody (I. 4, 33); after ślâgh, to applaud, hnu, to dissemble, to conceal, sthâ,31 to reveal, śap, to curse (I. 4, 34); after dhâray, to owe (I. 4, 35); spṛh, to long for (I. 4, 36); after verbs expressive of anger, ill-will, envy, detraction (I. 4, 37); after râdh and îksh, if they mean to consider concerning a person (I. 4, 39); after pratiśru and âśru, in the sense of according (I. 4, 40); anugṛ and pratigṛ, in the sense of acting in accordance with (I. 4, 41); after parikrî, to buy, to hire (I. 4, 44). Other cases of Sampradâna are mentioned after such words as namaḥ, salutation to, svasti, hail, svâhâ, salutation to the gods, svadhâ, salutation to the manes, alam, sufficient for, vashaṭ, offered to, a sacrificial invocation, etc. (II. 3, 16); and in such expressions as na tvam triṇâya manye, I do not value thee a straw (II. 3, 17); grâmâya gacchati, he goes to the village (II. 2, 12): where, however, the accusative, too, is equally admissible. Some other cases of Sampradâna are mentioned in the Vârttikas; e.g., I. 4, 44, muktaye harim bhajati, for the sake of liberation he worships Hari; vâtâya kapilâ vidyut, a dark red lightning indicates wind. Very interesting, too, is the construction with the prohibitive mâ; e.g. mâ câpalâya, lit. not for unsteadiness, i.e., do not act unsteadily.32
In all these cases we easily recognize the identity of Sampradâna with the dative in Greek and Latin. If therefore we see that Pâṇini in some of his rules states that Sampradâna takes the place of tum, the so called infinitive, we can hardly doubt that he had perceived the similarity in the functions of what we call dative and infinitive. Thus he says that instead of phalâny âhartum yâti, he goes to take the fruits, we may use the dative and say phalebhyo yâti, he goes for the fruits; instead of yashṭum yâti, he goes to sacrifice, yâgâya yâti, he goes to the act of sacrificing (II. 3, 14–15).
But whether Pâṇini recognized this fact or not, certain it is that we have only to look at the forms which in the Veda take the place of tum, in order to convince ourselves that most of them are datives of verbal nouns. As far as Sanskrit grammar is concerned, we may safely cancel the name of infinitive altogether, and speak instead boldly of datives and other cases of verbal nouns. Whether these verbal nouns admit of the dative case only, and whether some of those datival terminations have become obsolete, are questions which do not concern the grammarian, and nothing would be more unphilosophical than to make such points the specific characteristic of a new grammatical category, the infinitive. The very idea that every noun must possess a complete set of cases, is contrary to all the lessons of the history of language; and though the fact that some of these forms belong to an antiquated phase of language has undoubtedly contributed towards their being used more readily for certain syntactical purposes, the fact remains that in their origin and their original intention they were datives and nothing else. Neither could the fact that these datives of verbal nouns may govern the same case which is governed by the verb, be used as a specific mark, because it is well known that, in Sanskrit more particularly, many nouns retain the power of governing the accusative. We shall now examine some of these so-called infinitives in Sanskrit.
Datives in e.
The simplest dative is that in e, after verbal bases ending in consonants or â, e.g., dṛśé, for the sake of seeing, to see; vid-é, to know, paribhveê,33 to overcome; śraddhé kám, to believe.
Datives in ai.
After some verbs ending in â, the dative is irregularly (Grammar, §§ 239, 240) formed in ai; Rv. VII. 19, 7, parâdái, to surrender. III. 60, 4, pratimái, to compare, and the important form vayodhái, of which more by and by.
Accusatives in am. Genitives and Ablatives in as. Locatives in i.
By the side of these datives we have analogous accusatives in am, genitives and ablatives in as, locatives in i.
Accusative: I. 73, 10, śakéma yámam, May we be able to get. I. 94, 3, śakéma tvâ samídhan, May we be able to light thee. This may be the Oscan and Umbrian infinitive in um, om (u, o), if we take yama as a base in a, and m as the sign of the accusative. In Sanskrit it is impossible to determine this question, for that bases in a also are used for similar purposes is clearly seen in datives like dábhâya; e.g., Rv. V. 44, 2, ná dábhâya, not to conquer; VIII. 96, 1, nṛ́bhyâḥ tárâya síndhavaḥ su-pârấḥ, the rivers easy to cross for men. Whether the Vedic imperatives in âya (śâyac) admit of a similar explanation is doubtful on account of the accent.
Genitive: vilikhaḥ, in îśvaro vilikhaḥ, cognizant of drawing; and possibly X. 108, 2, atiskádaḥ bhiyásâ, from fear of crossing.
Ablative: Rv. VIII. 1, 12, purâ âtṛ́daḥ, before striking.
Locative: Rv. V. 52, 12, dṛśí tvishé, to shine in glancing(?)
Datives in s-e.
The same termination of the dative is added to verbal bases which have taken the increment of the aorist, the s. Thus from ji, to conquer, we have ji-sh, and je-sh, and from both datival forms with infinitival function. I. 111, 4, té naḥ hinvantu sâtáye dhiyé jishé, May they bring us to wealth, wisdom, victory!
I. 100, 11, apấm tokásya tánayasya jeshé, May Indra help us for getting water, children, and descendants. Cf. VI. 44, 18.
Or, after bases ending in consonants, upaprakshé; V. 47, 6, upa-prakshé vṛ́shaṇaḥ - - - vadhvấḥ yanti áccha, the men go towards their wives to embrace.
These forms correspond to Greek infinitives like λῦσαι and τύψαι, possibly to Latin infinitives like ferre, for fer-se, velle for vel-se, and voluis-se; for se, following immediately on a consonant, can never represent the Sanskrit ase. With regard to infinitives like fac-se, dic-se, I do not venture to decide whether they are primitive forms, or contracted, though fac-se could hardly be called a contraction of fecisse. The 2d pers. sing. of the imperative of the 1st aorist middle, λῦσαι, is identical with the infinitive in form, and the transition of meaning from the infinitive to the imperative is well known in Greek and other languages. (Παῖδα δ’ ἐμοὶ λῦσαί τε φίλην τά τ’ ἄποινα δέχεσθαι, Deliver up my dear child and accept the ransom). Several of these aoristic forms are sometimes very perplexing in Sanskrit. If we find, for instance, stushé, we cannot always tell whether it is the infinitive (λῦσαι); or the 1st pers. sing. of the aor. Âtmanep. in the subjunctive (for stushai), Let me praise (λύσωμαι); or lastly, the 2d pers. sing. Âtmanep. in the indicative (λύῃ). If stushe has no accent, we know, of course, that it cannot be the infinitive, as in X. 93, 9; but when it has the accent on the last, it may, in certain constructions, be either infinitive, or 1st pers. sing. aor. Âtm. subj. Here we want far more careful grammatical studies on the language of the Veda, before we can venture to translate with certainty. In places, for instance, where as in I. 122, 7 we have a nominative with stushé, it is clear that it must be taken as an infinitive, stushé sâ vâm - - - râtíḥ, your gift, Varuṇa and Mitra, is to be praised; but in other places, such as VIII. 5, 4, the choice is difficult. In VIII. 65, 5, índra griṇîshé u stushé, I should propose to translate, Indra, thou longest for praising, thou desirest to be praised, cf. VIII. 71, 15; while in II. 20, 4, tám u stushe índram tám gṛṇîshe, I translate, Let me praise Indra, let me laud him, admitting here, the irregular retention of Vikaraṇa in the aorist, which can be defended by analogous forms such as gṛ́-ṇî-sh-áṇi, stṛ́-ṇî-sh-áṇi, of which more hereafter. However, all these translations, as every real scholar knows, are, and can be tentative only. Nothing but a complete Vedic grammar, such as we may soon expect from Professor Benfey, will give us safe ground to stand on.
Datives in âyai.
Feminine bases in â form their dative in âyai, and thus we find carâyai used in the Veda, VII. 77, 1, as what we should call an infinitive, in the sense of to go. No other cases of carâ have as yet been met with. A similar form is jârâyai, to praise, I. 38, 13.
Datives in aye.
We have next to consider bases in i, forming their dative in áye. Here, whenever we are acquainted with the word in other cases, we naturally take aye as a simple dative of a noun. Thus in I. 31, 8, we should translate sanáye dhánânâm, for the acquisition of treasures, because we are accustomed to other cases, such as I. 100, 13, sanáyas, acquisitions, V. 27, 3, saním, wealth. But if we find, V. 80, 5, dṛśáye naḥ asthât, she stood to be seen by us, lit., for our seeing, then we prefer, though wrongly, to look upon such datives as infinitives, simply because we have not met with other cases of dṛśi-s.
Datives in taye.
What applies to datives of nouns in i, applies with still greater force to datives of nouns in ti. There is no reason why in IX. 96, 4 we should call áhataye, to be without hurt, an infinitive, simply because no other case of áhati-s occurs in the Rig-Veda; while ájîtaye, not to fail, in the same line, is called a dative of ájîti-s, because it occurs again in the accusative ájîti-m.
Datives in tyai.
In ityái, to go, I. 113, 6; 124, 1, we have a dative of iti-s, the act of going, of which the instrumental ityâ occurs likewise, I. 167, 5. This tyâ, shortened to tya, became afterwards the regular termination of the gerund of compound verbs in tya (Grammar § 446), while ya (§ 445) points to an original ya or yai.
Datives in as-e.
Next follow datives from bases in as, partly with accent on the first syllable, like neuter nouns in as, partly with the accent on as; partly with Guṇa, partly without. With regard to them it becomes still clearer how impossible it would be to distinguish between datives of abstract nouns, and other grammatical forms, to be called infinitives. Thus Rv. I. 7, 3 we read dîrghâya cákshase, Indra made the sun rise for long glancing, i.e., that it might glance far and wide. It is quite true that no other cases of cákshas, seeing, occur, on which ground modern grammarians would probably class it as an infinitive; but the qualifying dative dîrghâya, clearly shows that the poet felt cákshase as the dative of a noun, and did not trouble himself, whether that noun was defective in other cases or not.
These datives of verbal nouns in as, correspond exactly to Latin infinitives in ĕre, like vivere (jîváse), and explain likewise infinitives in âre, êre, and îre, forms which cannot be separated. It has been thought that the nearest approach to an infinitive is to be found in such forms as jîváse, bhiyáse, to fear (V. 29, 4), because in such cases the ordinary nominal form would be bháyas-e. There is, however, the instrumental bhiyása, X. 108, 2.
Datives in mane.
Next follow datives from nouns in man, van, and an. The suffix man is very common in Sanskrit, for forming verbal nouns, such as kar-man, doing, deed, from kar. Van is almost restricted to forming nomina agentis, such as druh-van, hating; but we find also substantives like pat-van, still used in the sense of flying. An also is generally used like van, but we can see traces of its employment to form nomina actionis in Greek ἀγών, Lat. turbo, etc.
Datives of nouns in man, used with infinitival functions, are very common in the Veda; e.g. I. 164, 6, pṛccâmi vidmane, I ask to know; VIII. 93, 8, dâmane kṛtáḥ, made to give. We find also the instrumental case vidmánâ, e.g., VI. 14, 5, vidmánâ urushyáti, he protects by his knowledge. These correspond to Homeric infinitives, like ἴδμεναι, δόμεναι, etc., old datives and not locatives, as Schleicher and Curtius supposed; while forms like δόμεν are to be explained either as abbreviated, or as obsolete accusatives.
Datives in vane.
Of datives in váne I only know dāvâne, a most valuable grammatical relic, by which Professor Benfey was enabled to explain the Greek δοῦναι, i.e., δοϝέναι.34
Datives in ane.
Of datives in áne I pointed out (l.c.) dhûrv-ane and vibhv-áne, VI. 61, 13, taking the latter as synonymous with vibhvế, and translating, Sarasvatî, the great, made to conquer, like a chariot. Professor Roth, s.v. vibhván, takes the dative for an instrumental, and translates “made by an artificer.” It is, however, not the chariot that is spoken of, but Sarasvatî, and of her it could hardly be said that she was made either by or for an artificer.
Locatives in sani.
As we saw before that aoristic bases in s take the datival e, so that we had prák-sh-e by the side of pṛ́c-e, we shall have to consider here aoristic bases in s, taking the suffix an, not however with the termination of the dative, but with that of the locative i. Thus we read X. 126, 3, náyishṭhâḥ u naḥ nesháṇi párshishṭhâḥ u naḥ parsháṇi áti dvíshaḥ, they who are the best leaders to lead us, the best helpers to help us to overcome our enemies, lit. in leading us, in helping us. In VIII. 12, 19, gṛṇîsháni, i.e. gṛ-ṇî-sháṇ-i stands parallel with turv-án-e, thus showing how both cases can answer nearly the same purpose. If these forms existed in Greek, they would, after consonantal bases, be identical with the infinitives of the future.
Cases of Verbal Nouns in tu.
We next come to a large number of datives, ablatives, or genitives, and accusatives of verbal nouns in tu. This tu occurs in Sanskrit in abstract nouns such as gâtú, going, way, etc., in Latin in adven-tus, etc. As these forms have been often treated, and as some of them occur frequently in later Sanskrit also, it will suffice to give one example of each:—
Dative in tave: gántave, to go, I. 46, 7.
Old form in ai: gántavái, X. 95, 14.
Genitive in toḥ: dâtoḥ, governed by îśe, VII. 4, 6.
Ablative in toḥ: gántoḥ, I. 89, 9.
Accusative in tum: gántum. This is the supine in tum in Latin.
Cases of Verbal Nouns in tva.
Next follow cases of verbal nouns in tvá, the accent being on the suffix.
Datives in tvấya: hatvấya, X. 84, 2.
Instrumental in tvấ: hatvấ, I. 100, 18.
Older form in tvî́: hatvî́, II. 17, 6; gatvî́, IV. 41, 5.
Datives in dhai and dhyai.
I have left to the end datives in dhai and dhyai, which properly belong to the datives in ai, treated before, but differ from them as being datives of compound nouns. As from máyaḥ, delight, we have mayaskará, delight-making, mayobhú, delight-causing, and constructions like máyo dádhe, so from váyas, life, vigor, we have váyaskṛ́t, life-giving, and constructions like váyo dhât. From dhâ we can frame two substantival frame, dhâ and dhi-s, e.g. puro-dhâ, and puro-dhis, like vi-dhi-s. As an ordinary substantive, purodhâ takes the feminine termination â, and is declined like śivâ. But if the verbal base remains at the end of a compound without the feminine suffix, a compound like vayodhâ would form its dative vayodhe (Grammar, § 239); and as in analogous cases we found old datives in ai, instead of e, e.g. parâdai, nothing can be said against vayodhai, as a Vedic dative of vayodhâ. The dative of purodhi would be purodhaye, but here again, as, besides forms like dṛśaye, we met with datives, such as ityai, rohishyai, there is no difficulty in admitting an analogous dative of purodhi, viz., purodhyai.
The old dative dhai has been preserved to us in one form only, which for that reason is all the more valuable and important, offering the key to the mysterious Greek infinitives in θαι, I mean vayodhái, which occurs twice in the Rig-Veda, X. 55, 1, and X. 67, 11. The importance of this relic would have been perceived long ago, if there had not been some uncertainty as to whether such a form really existed in the Veda. By some accident or other, Professor Aufrecht had printed in both passages vayodhaiḥ, instead of vayodhai. But for this, no one, I believe, would have doubted that in this form vayodhai we have not only the most valuable prototype of the Greek infinitives in (σ)θαι, but at the same time their full explanation. Vayodhai stands for vayas-dhai, in which composition the first part vayas is a neuter base in as, the second a dative of the auxiliary verb dhâ, used as a substantive. If, therefore, we find corresponding to vayodhai a Greek infinitive βέεσθαι, we must divide it into βέεσ-θαι, as we divide ψεύδεσθαι into ψεύδεσ-θαι, and translate it literally by “to do lying.”
It has been common to identify Greek infinitives in σθαι with corresponding Sanskrit forms ending in dhyai. No doubt these forms in dhyai are much more frequent than forms in dhai, but as we can only take them as old datives of substantives in dhi, it would be difficult to identify the two. The Sanskrit dhy appears, no doubt, in Greek, as σσ, dh being represented by the surd θ, and then assibilated by y; but we could hardly attempt to explain σθ = θy, because σδ = ζ = δy. Therefore, unless we are prepared to see with Bopp in the σ before θ, in this and similar forms, a remnant of the reflexive pronoun, nothing remains but to accept the explanation offered by the Vedic vayodhai, and to separate ψεύδεσθαι into ψεύδεσ-θαι lying to do. That this grammatical compound, if once found successful, should have been repeated in other tenses, giving us not only γράφεσ-θαι, but γράψεσ-θαι, γράψασ-θαι, and even γραφθήσεσ-θαι, is no more than what we may see again and again in the grammatical development of ancient and modern languages. Some scholars have objected on the same ground to Bopp’s explanation of ama-mini, as the nom. plur. of a participle, because they think it impossible to look upon amemini, amabâmini, amaremini, amabimini as participial formations. But if a mould is once made in language, it is used again and again, and little account is taken of its original intention. If we object to γράψεσ-θαι, why not to κελευ-σέ-μεναι or τεθνά-μεναι or μιχθή-μεναι? In Sanskrit, too, we should hesitate to form a compound of a modified verbal base, such as pṛṇa, with dhi, doing; yet as the Sanskrit ear was accustomed to yajadhyai from yaja, gamadhyai from gama, it did not protest against pṛṇadhyai, vâvṛdhadhyai, etc.
Historical Importance of these Grammatical Forms.
And while these ancient grammatical forms which supply the foundation of what in Greek, Latin, and other languages we are accustomed to call infinitives are of the highest interest to the grammarian and the logician, their importance is hardly less in the eyes of the historian. Every honest student of antiquity, whether his special field be India, Persia, Assyria, or Egypt, knows how often he is filled with fear and trembling when he meets with thoughts and expressions which, as he is apt to say, cannot be ancient. I have frequently confessed to that feeling with regard to some of the hymns of the Rig-Veda, and I well remember the time when I felt inclined to throw up the whole work as modern and unworthy of the time and labor bestowed upon it. At that time I was always comforted by these so-called infinitives and other relics of ancient language. They could not have been fabricated in India. They are unknown in ordinary Sanskrit, they are unintelligible as far as their origin is concerned in Greek and Latin, and yet in the Vedic language we find these forms, not only identical with Greek and Latin forms, but furnishing the key to their formation in Greece and Italy. The Vedic vayas-dhái compared with Greek βεεσ-θαι, the Vedic stushe compared with λυσαι are to my mind evidence in support of the antiquity and genuineness of the Veda that cannot be shaken by any arguments.
The Infinitive in English.
I add a few words on the infinitive in English, though it has been well treated by Dr. March in his “Grammar of the Anglo-Saxon Language,” by Dr. Morris, and others. We find in Anglo-Saxon two forms, one generally called the infinitive, nim-an, to take, the other the gerund, to nim-anne, to take. Dr. March explains the first as identical with Greek νέμ-ειν and νέμ-εν-αι, i.e., as an oblique case, probably the dative, of a verbal noun in an. He himself quotes only the dative of nominal bases in a, e.g. namanâya, because he was probably unacquainted with the nearer forms in an-e supplied by the Veda. This infinitive exists in Gothic as nim-an, in Old Saxon as nim-an, in Old Norse as nem-a, in Old High German as nem-an. The so-called gerund, to nimanne, is rightly traced back by Dr. March to Old Saxon nim-annia, but he can hardly be right in identifying these old datival forms with the Sanskrit base nam-anîya. In the Second Period of English (1100–1250)35 the termination of the infinitive became en, and frequently dropped the final n, as smelle = smellen; while the termination of the gerund at the same time became enne, (ende), ene, en, or e, so that outwardly the two forms appear to be identical, as early as the 12th century.36 Still later, towards the end of the 14th century, the terminations were entirely lost, though Spenser and Shakespeare have occasionally to killen, passen, delven, when they wished to impart an archaic character to their language. In modern English the infinitive with to is used as a verbal substantive. When we say, “I wish you to do this,” “you are able to do this,” we can still perceive the datival function of the infinitive. Likewise in such phrases, “it is time,” “it is proper,” “it is wrong to do that,” to do may still be felt as an oblique case. But we have only to invert these sentences, and say, “to do this is wrong,” and we have a new substantive in the nom. sing., just as in the Greek τὸ λέγειν. Expressions like for to do, show that the simple to was not always felt to be sufficiently expressive to convey the meaning of an original dative.
Works on the Infinitive.
The infinitive has formed the subject of many learned treatises. I divide them into two classes, those which appeared before and after Wilhelm’s excellent essay, written in Latin, “De Infinitivi Vi et Natura,” 1868; and in a new and improved edition, “De Infinitivo Linguarum Sanscritæ, Bactricæ, Persicæ, Græcæ, Oscæ, Umbricæ, Latinæ, Goticæ, forma et usu,” Isenaci, 1873. In this essay the evidence supplied by the Veda was for the first time fully collected, and the whole question of the nature of the infinitive placed in its true historical light. Before Wilhelm the more important works were Hofer’s book, “Vom Infinitiv, besonders im Sanskrit,” Berlin, 1840; Bopp’s paragraphs in his “Comparative Grammar;” Humboldt’s paper, in Schlegel’s “Indische Bibliothek” (II. 74), 1824; and his posthumous paper in Kuhn’s “Zeitschrift” (II. 245), 1853; some dissertations by L. Meyer, Merguet, and Golenski. Benfey’s “Sanskrit Grammar” (1852), too, ought to be mentioned, as having laid the first solid foundations for this and all other branches of grammatical research, as far as Sanskrit is concerned. After Wilhelm the same subject has been treated with great independence by Ludwig, “Der Infinitif im Veda,” 1871, and again “Agglutination oder Adaptation,” 1873; and also by Jolly, “Geschichte des Infinitivs,” 1873.
I had myself discussed some questions connected with the nature of the infinitive in my “Lectures on the Science of Language,” vol. ii. p. 15 seq., and I had pointed out in Kuhn’s “Zeitschrift,” XV. 215 (1866) the great importance of the Vedic vayodhai for unraveling the formation of Greek infinitives in σ-θαι.
The Infinitive in Bengali.
At a still earlier time, in 1847, in my “Essay on Bengali,” I said: “As the infinitives of the Indo-Germanic languages must be regarded as the absolute cases of a verbal noun, it is probable that in Bengali the infinitive in ite was also originally a locative, which expressed not only local situation, but also movement towards some object, as an end, whether real or imaginary. Thus the Bengali infinitive corresponds exactly with the English, where the relation of case is expressed by the preposition to. Ex. tâhâke mârite âmi âsiyâchi, means, I came to the state of beating him, or, I came to beat him; âmâke mârite deo, give me (permission), let me (go) to the action of beating, i.e., allow me to beat. Now as the form of the participle is the same as that of the infinitive, it may be doubted if there is really a distinction between these two forms as to their origin. For instance, the phrase âpan putrake mârite âmi tâhâka dekhilâm, can be translated, I saw him beating his own son; but it can be explained also as, what they nonsensically call in Latin grammar accusativus cum infinitivo, that is to say, the infinitive can be taken for a locative of the verbal noun, and the whole phrase be translated, I saw him in the action of beating his own son, (vidi patrem cædere ipsius filium). As in every Bengali phrase the participle in ite can be understood in this manner, I think it admissible to ascribe this origin to it, and instead of taking it for a nominative of a verbal adjective, to consider it as a locative of a verbal noun.”
The Infinitive in the Dravidian Languages.
I also tried to show that the infinitive in the Dravidian languages is a verbal noun with or without a case suffix. This view has been confirmed by Dr. Caldwell, but, in deference to him, I gladly withdraw the explanation which I proposed in reference to the infinitive in Tamil. I quote from Dr. Caldwell’s “Comparative Grammar of the Dravidian Languages,” 2d ed. p. 423: “Professor Max Müller, noticing that the majority of Tamil infinitives terminate in ka, supposed this ka to be identical in origin with kô, the dative-accusative case-sign of the Hindi, and concluded that the Dravidian infinitive was the accusative of a verbal noun. It is true that the Sanskrit infinitive and Latin supine in tum is correctly regarded as an accusative, and that our English infinitive to do, is the dative of a verbal noun; it is also true that the Dravidian infinitive is a verbal noun in origin, and never altogether loses that character; nevertheless, the supposition that the final ka of most Tamil infinitives is in any manner connected with ku, the sign of the Dravidian dative, or of kô, the Hindi dative-accusative, is inadmissible. A comparison of various classes of verbs and of the various dialects shows that the kâ in question proceeds from a totally different source.”
On Labialized and Unlabialized Gutturals.
As in my article on Vayodhai, published in Kuhn’s “Zeitschrift,” 1866, p. 215, I had entered a caveat against identifying Greek β with Sanskrit ज, I take this opportunity of frankly withdrawing it. Phonetically, no doubt, these two letters represent totally distinct powers, and to say that Sanskrit ज ever became Greek β is as irrational to-day as it was ten years ago. But historically I was entirely wrong, as will be seen from the last edition of Curtius’ “Grundzüge.” The guttural sonant check was palatalized in the Southeastern Branch, and there became j and z, while in the Northwestern Branch the same g was frequently labialized and became gv, v, and b. Hence, where we have ज in Sanskrit, we may and do find β in Greek.
But after withdrawing my former caveat, I make bold to propose another, namely, that the original palatal sonant flatus, which in Sanskrit is graphically represented by j, can never be represented in Greek by β. Whether j in Sanskrit represents an original palatal sonant check or an original palatal sonant flatus can generally be determined by a reference to Zend, which represents the former by j, the latter by z. We may therefore formulate this phonetic law:—
“When Sanskrit j is represented by Zend z, it cannot be represented by Greek β.”
In this manner it is possible, I believe, to utilize Ascoli’s and Fick’s brilliant discovery as to a twofold, or even threefold, distinction of the Aryan k, as applied to the Aryan g. They have proved that all Aryan languages show traces of an original distinction between a guttural surd check, k, frequently palatalized in the Southeastern Branch (Sk. c, Zend c) and liable to labialization, in Latin, Greek, Cymric, and Gothic; and another k, never liable to labialization, but changed into a flatus, palatal or otherwise, in Sanskrit, Lithuanian, and Old Slavonic. They showed, in fact,—
| Sanskrit. | Lith. | Slav. | Gadh. & Cym. | Lat. | Greek. | Gothic. | |
| क (च) | = k | = k, č, c | = c | = p | = c, qu, v | = κ, κϝ, κκ, π, ππ, τ, ττ, = | hv, h. |
| श | = sz | = s | = c | = c | = κ | = h | |
In the same manner we ought in future to distinguish between a guttural sonant check, g, frequently palatalized in the Southeastern Branch (Sk. j, Zend j), and liable to labialization, like k; and another g, never liable to labialization, but changed into a flatus, palatal or otherwise, in Zend, Lithuanian, and Old Slavonic. As we never have π = श we never have β = ज, if ज in Zend is z.
The evidence will be found under Sk. jan, jabh, jar (to decay, and to praise), jush, jñâ, jñu, jâmâtar; aj, bhrâj, marj, yaj, raj(atam).
Gothic quinô, Gadh. ben, Bœot. βάνα depend on Zend jeni; Gadh. baith-is on Zend jaf-ra. It is wrong to connect σβεσ with jas, on account of Zend zas, and gyâ-ni with βία, on account of Zend zyâ-ni.