iko yan.aci. (12)As it happens, a language consists of sentences, not just of words. From the
perspective of the history of science, therefore, Pa ̄n.ini’s step is momentous
because it resulted not just in the creation of artificial expressions but of an
artificial metalanguage. Rule (12) is not merely an artificial expression but an
element of an artificial language like the artificial languages of algebra that
came into being much later through Arab and European efforts. Pa ̄n.ini demon-
strates in passing that artificial languages need not be written.
Patañjali refers to Pa ̄n.ini’s first step of generalization by declaring: “this
science pertains to all the Vedas” or “this science is a Par s.ad=Pra ̄tis ́a ̄khya for
all the Vedas” (sarvavedapa ̄ris.adam idam s ́a ̄stram). Actually, Pa ̄n.ini’s Vedic rules
are haphazard and incomplete while his rules for the spoken language are almost
perfect, if not in syntax, at least in phonology and morphology. Claims like these
sound like cheap commercials but are substantiated by comparing his grammar
with linguistic usage as we know it. It is sometimes said that “Pa ̄n.ini was, of
course, aided in his analysis by the extraordinary clarity of the Sanskrit lan-
guage,” but John Brough observed: “We are apt to overlook the possibility that
this structure might not have seemed so clear and obvious to us if Pa ̄n.ini had
not analysed it for us.”
The perfection of Pa ̄n.ini’s grammar is not only due to its high degree of for-
malization. His science is as empirical as it is formal. The locus classicus on the
importance of empirical description is the laconic expression lokatah.“on account
of (the usage) of the people.” “The people” are the native speakers as illustrated
by Patañjali: “He who needs a pot for some purpose goes to the house of a potter
and says: ‘You make a pot. I need a pot for some purpose.’ No one who wants to
use words goes to the house of a grammarian and says: ‘You make words. I want
to use them.’ (On the contrary,) having brought something to mind, without
further ado, he uses words.”
4 Later Schools
Patañjali’s “Great Commentary” of around 150 bcseems to have been followed
by a lull of several centuries. In the fifth century ad, the great grammatical
philosopher Bhartr.hari broke the silence by composing an original subcom-
mentary on Patañjali. By that time, Jainas and Buddhists had also begun to use
Sanskrit for their canonic writings and this resulted in several new grammars,
at first practical manuals, but increasingly systematical and scientific treatises.
Devanandin and Candragomin, Jaina and Buddhist grammarians, respectively,
also of the fifth century, adopted Pa ̄n.ini’s system and metalanguage and sought
to simplify or abbreviate his expressions, sometimes introducing innovations.
Candragomin refers to Middle-Indic forms that had been used in earlier Buddhist
356 frits staal