1

had at school, and we can hardly understand how at Rome, grammar- pure and simple grammar-should have formed a subject of general interest, and a topic of fashionable conversation. When one of the first grammarians of the day, Crates of Pergamus, was sent to Rome as ambassador of King Attalus, he was received with the greatest distinction by all the literary statesmen of the capital. It so happened that when walking one day on the Palatian hill, Crates caught his foot in the grating of a sewer, fell and broke his leg. Being thereby detained at Rome longer than he intended, he was persuaded to give some public lectures, or akroaseis, on grammar; and from these lectures, says Suetonius, dates the study of grammar at Rome. This took place about 159 B. C., between the second and third Punic wars, shortly after the death of Ennius, and two years after the famous expulsion of the Greek rhetors and philosophers (161). Four years later Carneades, likewise sent to Rome as ambassador, was prohibited from lecturing by Cato. After these lectures of Crates, grammatical and philological studies became extremely popular at Rome. We hear of Lucius Elius Stilo,1 who lectured on Latin as Crates had lectured on Greek. Among his pupils were Varro, Lucilius, and Cicero. Varro composed twenty-four books on the Latin language, four of which were dedicated to Cicero. Cicero, himself, is quoted as an authority on grammatical questions, though we know of no special work of his on grammar. Lucilius devoted the ninth book of his satires to the reform of spelling.2 But nothing shows

1 Mommsen, ii. 413, 426, 445, 457. Lucius Ælius Stilo wrote a work on etymology, and an index to Plautus. - Lersch, Die Sprachphilosophie der Alten, ii. 111.

2 Lersch, ii. 113, 114, 143.

more clearly the wide interest which grammatical studies had then excited in the foremost ranks of Roman society than Caesar's work on Latin grammar. It was composed by him during the Gallic war, and dedicated to Cicero, who might well be proud of the compliment thus paid him by the great general and statesman. Most of these works are lost to us, and we can judge of them only by means of casual quotations. Thus we learn from a fragment of Cæsar's work, De analogia, that he was the inventor of the term ablative in Latin. The word never occurs before, and, of course, could not be borrowed, like the names of the other cases, from Greek grammarians, as they admitted no ablative in Greek. To think of Cæsar fighting the barbarians of Gaul and Germany, and watching from a distance the political complications at Rome, ready to grasp the sceptre of the world, and at the same time carrying on his philological and grammatical studies together with his secretary, the Greek Didymus,1 gives us a new view both of that extraordinary man, and of the time in which he lived. After Cæsar had triumphed, one of his favorite plans was to found a Greek and Latin library at Rome, and he offered the librarianship to the best scholar of the day, to Varro, though Varro had fought against him on the side of Pompey.2

We have thus arrived at the time when, as we saw in an earlier part of this lecture, Dionysius Thrax published the first elementary grammar of Greek at Rome. Empirical grammar had thus been transplanted to Rome, the Greek grammatical terminology was translated into Latin, and in this new Latin garb it has travelled now for nearly two thousand years over the whole civilized world. 1 Lersch, iii. 144.

2 Mommsen, iii. 557. 48 B. C.

Even in India, where a different terminology had grown up in the grammatical schools of the Brahmans, a terminology in some respects more perfect than that of Alexandria and Rome, we may now hear such words - as case, and gender, and active and passive, explained by European teachers to their native pupils. The fates of words are curious indeed, and when I looked the other day at some of the examination papers of the government schools in India, such questions as" Write the genitive case of Siva," seemed to reduce whole volumes of history into a single sentence. How did these words, genitive case, come to India? They came from England, they had come to England from Rome, to Rome from Alexandria, to Alexandria from Athens. At Athens, the term case, or ptōsis, had a philosophical meaning; at Rome, casus was merely a literal translation; the original meaning of fall was lost, and the word dwindled down to a mere technical term. At Athens, the philosophy of language was a counterpart of the philosophy of the mind. The terminology of ▾ formal logic and formal grammar was the same. The logic of the Stoics was divided into two parts,1 called rhetoric and dialectic, and the latter treated, first, "On that which signifies, or language ;" secondly, "On that which is signified, or things." In their philosophical language ptosis, which the Romans translated by casus, really meant fall; that is to say, the inclination or relation of one idea to another, the falling or resting of one word on another. Long and angry discussions were carried on as to whether the name of ptōsis, or fall, was applicable to the nominative; and every true Stoic

1 Lersch, ii. 25. Περὶ σημαινόντων, or περὶ φώνης; and περὶ σημαινομένων, οι περὶ πραγμάτων.

E

would have scouted the expression of casus rectus, be-
cause the subject or the nominative, as they argued, did
not fall or rest on anything else, but stood erect, the
other words of a sentence leaning or depending on it.
All this is lost to us when we speak of cases.

And how are the dark scholars in the government
schools of India to guess the meaning of genitive? The
Latin genitivus is a mere blunder, for the Greek word
genike could never mean genitivus. Genitivus, if it is
meant to express the case of origin or birth, would in
Greek have been called gennētikē, not genike. Nor does
the genitive express the relation of son to father. For
though we may say, "the son of the father," we may
likewise say, "the father of the son." Genikē, in Greek,
had a much wider, a much more philosophical mean-
ing. It meant casus generalis, the general case, or

rather the case which expresses the genus ór kind.
This is the real power of the genitive. If I say, "a
bird of the water," "of the water" defines the genus
to which a certain bird belongs; it refers it to the genus
of water-birds. "Man of the mountains," means a
mountaineer. In phrases such as "son of the father,"
or "father of the son," the genitives have the same
effect. They predicate something of the son or of the
father; and if we distinguished between the sons of
the father, and the sons of the mother, the genitives
would mark the class or genus to which the sons re-
spectively belonged. They would answer the same pur-
pose as the adjectives, paternal and maternal. It can
be proved etymologically that the termination of the
genitive is, in most cases, identical with those derivative

1 Beiträge zur Geschichte der Grammatik, von Dr. K. E. A. Schmidt.
Halle, 1859. Uber den Begriff der yεviêǹ πτwσɩ, s. 320.

7 suffixes by which substantives are changed into adjectives.1

7

It is hardly necessary to trace the history of what I call the empirical study, or the grammatical analysis of language, beyond Rome. With Dionysius Thrax the

1 In the Tibetan languages the rule is, “Adjectives are formed from substantives by the addition of the genitive sign," which might be inverted into, "The genitive is formed from the nominative by the addition of the adjective sign." For instance, shing, wood; shing gi, of wood, or wooden: ser, gold; ser-gyi, of gold, or golden: mi, man; mi-yi, of man, or human. The same in Garo, where the sign of the genitive is ni, we have; mánde-ní jak, the hand of man, or the human hand; ambal-ní ketháli, a wooden knife, or a knife of wood. In Hindustání the genitive is so clearly an adjective, that it actually takes the marks of gender according to the words to which it refers. But how is it in Sanskrit and Greek? In Sanskrit we may form adjectives by the addition of tya. (Turanian Languages, p. 41, seq.; Essay on Bengálí, p. 333.) For instance, dakshinâ, south; dakshinâ-tya, southern. This tya is clearly a demonstrative pronoun, the same as the Sanskrit syas, syâ, tyad, this or that. Tya is a pronominal base, and therefore such adjectives as dakshinâ-tya, southern, or âp-tya, aquatic, from âp, water, must have been conceived originally as "water-there," or "south-there." Followed by the terminations of the nominative singular, which was again an original pronoun, âpty is would mean âp-tya-s, i. e., water-there-he. Now, it 7 makes little difference whether I say an aquatic bird or a bird of the water. In Sanskrit the genitive of water would be, if we take udaka, udaka-sya, This sya is the same pronominal base as the adjective termination tya, only that the former takes no sign for the gender, like the adjective. The genitive udakusya is therefore the same as an adjective without gender. Now let us look to Greek. We there form adjectives by ouoc, which is the same as the Sanskrit tya or sya. For instance, from diuos, people, the Greeks formed dŋuócios, belonging to the people. Here oç, a, ov, mark the gender. Leave the gender out, and you get dnμooio. Now, there is a rule in Greek that an ç between two vowels, in grammatical terminations, is elided. Thus the genitive of γένος is not γένεσος, but γένεος, or γένους ; hence δημόσιο would necessarily become duoto. And what is doo but the regular Homeric genitive of δῆμος, which in later Greek was replaced by δήμου ? Thus we see that the same principles which governed the formation of adjectives and genitives in Tibetan, in Garo, and Hindustání, were at work in the primitive stages of Sanskrit and Greek; and we perceive how accurately the real power of the genitive was determined by the ancient Greek grammarians, who called it the general or predicative case, whereas the Romans spoiled the term by wrongly translating it into genitivus.