From the difficulty of retaining a right discrimination of tense seems to have proceeded the active auxiliary verb. It is possible that this was borrowed from the Teutonic languages of the barbarians, and accommodated both by them and by the natives to words of Latin origin. The passive auxiliary is obtained by a very ready resolution of any tense in that mood, and has not been altogether dispensed with even in Greek, while in Latin it is used much more frequently. It is not quite so easy to perceive the propriety of the active habeo or teneo, one or both of which all modern languages have adopted as their auxiliaries in conjugating the verb. But in some instances this analysis is not improper; and it may be supposed that nations, careless of etymology or correctness, applied the same verb by a rude analogy to cases where it ought not strictly to have been employed.[489] Next to the changes founded on pronunciation and to the substitution of auxiliary verbs for inflections, the usage of the definite and indefinite articles in nouns appears the most considerable step in the transmutation of Latin into its derivative languages. None but Latin, I believe, has ever wanted this part of speech; and the defect to which custom reconciled the Romans would be an insuperable stumbling-block to nations who were to translate their original idiom into that language. A coarse expedient of applying _unus_, _ipse_, or _ille_ to the purposes of an article might perhaps be no unfrequent vulgarism of the provincials; and after the Teutonic tribes brought in their own grammar, it was natural that a corruption should become universal, which in fact supplied a real and essential deficiency. [Sidenote: Pronunciation no longer regulated by quantity.] That the quantity of Latin syllables is neglected, or rather lost, in modern pronunciation, seems to be generally admitted. Whether, indeed, the ancient Romans, in their ordinary speaking, distinguished the measure of syllables with such uniform musical accuracy as we imagine, giving a certain time to those termed long, and exactly half that duration to the short, might very reasonably be questioned; though this was probably done, or attempted to be done, by every reader of poetry. Certainly, however, the laws of quantity were forgotten, and an accentual pronunciation came to predominate, before Latin had ceased to be a living language. A Christian writer named Commodianus, who lived before the end of the third century according to some, or, as others think, in the reign of Constantine, has left us a philological curiosity, in a series of attacks on the pagan superstitions, composed in what are meant to be verses, regulated by accent instead of quantity, exactly as we read Virgil at present.[490]