which stands nearest to it” (Rik Pra ̄tis ́a ̄khya1.14); and later simply as: “a change
is expressed as an Accusative” (Va ̄jasaneyi Pra ̄tis ́a ̄khya 1.133).
The Padapa ̄t.ha discovered the differences between sentences, words, stems,
pre- and suffixes, roots, etc., but the Pra ̄tis ́a ̄khya added an almost perfect ana-
lysis of the sounds of language into vowels, consonants, semi-vowels, stops,
dentals, velars, nasals, etc. They placed these sounds in a two-dimensional con-
figuration, developed from the square or varga offive-by-five series (also called
varga) of syllables that begin with a stop followed by a short a:
ka kha ga gha n.a
ca cha ja jha ña
t.at.ha d.ad.ha n.a
ta tha da dha na
pa pha ba bha ma (1)These sounds are marked in our modern transcription by diacritics which we
need because our alphabet cannot express all the sounds of Sanskrit. It is not
necessary for the reader to know how exactly they are pronounced (but see
below); however, we need to know that all the vowels ain the syllables of the
table are short(like the first aofapra ̄h.). A bar above a vowel (like the second aof
apra ̄h.) indicates that it is long, that is, lasts twice as long as the short.
The 5 ¥ 5 varga“square” was extended and completed with fricatives or sibi-
lants, semi-vowels and vowels. One should not look upon the resulting inventory
as the beginningof linguistics or compare it to our haphazard “ABC’s” to which
there is no rhyme or reason. Like Mendelejev’s Periodic System of Elements, the
varga system was the result of centuries of analysis. In the course of that devel-
opment the basic concepts of phonology were discovered and defined.
In the mentalrepresentation of a such a square, it does not matter what is its
spatial orientation and whether vargadenotes a row or a column. The configura-
tion remains the same when the square is rotated around its diagonal. Proximity
and distance are also the same whatever the direction. Since two directions are
met with, this shows that these squares were composed and transmitted orally.
Renou and Filliozat (1953: 668) had already made a more significant general
observation: “One is forced to observe in this context that a Semitic type of writing
would have hindered phonetic studies if it had existed at the time in India, because
it would have provided a model of analysis of the sounds of language that was
practical but not scientific.” The Indian science of language, in other words, did
not originate in spite ofthe absence of writing but becauseof it.
3Pa ̄n.ini
The “Eight Chapters” (As.t.a ̄dhya ̄yı ̄) of Pa ̄n.ini’s grammar of Sanskrit (early to mid
fourth century bc) consist of rules, metarules, and defining rules. In the area of
352 frits staal