Sie sind auf Seite 1von 36

MATHEMATICS AND OTHER STRAS:

PINI AND PIGALA



M.D.SRINIVAS
CENTRE FOR POLICY STUDIES
mdsrinivas50@gmail.com


DEVELOPMENT OF VYKARAA OR ABDASTRA
Pre-Pinian: Yskas Nirukta, Prtikhya Texts
piali, Indra, Kaktsna, kayana, Vydi, etc
Pini (c. 500 BCE): Adhyy Strapha, Dhtupha, Gaapha
Ktyyana: Vrttika, Pli-vykaraa
Patajali (c. 100 BCE): Mahbhya
arvavarman: Ktantra-vykaraa
Candragomin (c.450 CE): Cndra-vykaraa
Devanandin (c.450): Jainendra-vykaraa
Bharthari (c. 450): Vkyapadya, Mahbhya-dpik
Jayditya, Vmana (c. 600): Kikvtti
Jinendrabuddhi (c.900): Kikvivaraa-pajik or Nysa
Kaiyaa (c. 900): Mahbhya-pradpa
Haradatta (c. 1000): Padamajar
Dharmakrti (c.1000): Rpvatra
Hemacandra (c. 1100): Siddhahaimacandra, etc
Vopadeva (c.1250): Mudgdhabodha
DEVELOPMENT OF VYKARAA OR ABDASTRA
Rmacandra (c.1350): Prakriykaumud
Nryaa Bhatiri (c.1600): Prakriysarvasva
Bhaoji Dkita (c.1625): Siddhntakaumud, Prauhamanoram,
abdakaustubha
Kauabhaa (c.1650): Vaiykaraabhaa
Varadarja (c.1650): Laghu-siddhntakaumud, Sra-siddhntakaumud
Ngeabhaa (c.1700): Mahbhya-pradipodyota, Bhacchabdendu-
ekhara,Vaiykaraa-siddhntamaj, Paribhenduekhara

Grammars of Other Languages
Tamil: Tolkppiyam (c.200 BCE), Vrasolyam (c.1200), Nannl (c.1300)
Kannada: Karnaka-bhbhaa (c.1100), abdamaidarpaa
(c.1200), Karnaka-abdnusana (c.1600)
Telugu: ndhra-abdacintmai (c.1100), ndhrabhbhaa (c.1250),
Triliga-abdnusana (c.1300)
Pali: Kaccyana-vykaraa, Saddalakkhaa (c. 1150)
Prkta: Prkta-praka, Prkta-abdnusana (c.1200)
Persian: Praspraka (c.1575)
STRAS: PRESENT SYSTEMATIC PROCEDURES
Most of the canonical texts on different disciplines (stras) in Indian
tradition do not present a series of propositions; instead they present a
series of rules, which serve to characterize and carry out systematic
procedures to accomplish various ends.
These systematic procedures are generally referred to as vidhi, kriy or
prakriy, sdhana, karma or parikarma, karaa, upya etc in different
disciplines.
The rules are often formulated in the form of stras: According to
Vinudharmottarapura (3.5.1): A stra has to be concise, unambiguous,
pithy, comprehensive, shorn of irrelevancies and blemish-less.
~ U(
Pinis Adhyy is acknowledged to be the paradigmatic example of a
canonical text in Indian tradition. All other disciplines, especially
mathematics, have been deeply influenced by its ingenious symbolic and
technical devices, recursive and generative formalism and the system of
conventions governing rule application and rule interaction.
PINI AND EUCLID
In Euclids geometry, propositions are derived from axioms with the help
of logical rules which are accepted as true. In Pinis grammar, linguistic
forms are derived from grammatical elements with the help of rules which
were framed ad hoc (i.e. stras)....
Historically speaking, Pinis method has occupied a place comparable to
that held by Euclids method in western thought. Scientific developments
have therefore taken different directions in India and in the West.... In
India, Pinis perfection and ingenuity have rarely been matched outside
the realm of linguistics. Just as Plato reserved admission to his Academy
for geometricians, Indian scholars and philosophers are expected to have
first undergone a training in scientific linguistics....
[J.F.Staal, Euclid and Pini, Philosophy East and West, 15, 1965, 99-116]
Note: The word derived means demonstrated in the case of Euclidean
Geometry; it means generated in the case of Pinis Grammar (upapatti
and nipatti)
ABDNUSANA: PINIS ADHYY

u u ~ j +
u [ uU~]
u u ~ ~
Jg ~
u u " ~
u u
[ ]
Now, the instruction of utterances
Instruction, namely generation (of utterances) by using prakti, pratyaya
and other components, this is done by grammar, and that it is its direct
purpose.
Valid utterances cannot be taught by pratipada-ptha (stating each of them
individually). Bhaspati tried to teach Indra valid utterances by pratipada-
pha for thousand divine years, but reached nowhere near the end.
ABDNUSANA: PINIS ADHYY
u~ ( u
uU
> ~>
~ ~ U
( ) U ( )
[ ]
How are these utterances to be known?
Some characterisation with what is general and particular is to be
provided, by which, with little effort, great amount of utterances are
known.
What is that characterisation? Utsarga (general) and Apavda
(special/exceptional) rules...


INDOLOGISTS ON PINIS ADHYY
Of particular interest is the stress laid on the small number of primitive
elements, themselves not used (i.e., themselves abstract) from which the
Sanskrit grammarians are said to derive the infinite variety of actual
forms in use.
[J.F.Staal on Francois Pons letter of 1740 (published 1743) in, A Reader on the
Sanskrit Grammarians, MIT Press, 1972, p.30]

The Descriptive Grammar of Sanskrit, which Pini brought to its
perfection, is one of the greatest monuments of human intelligence and an
indispensible model for the description of languages.
[L. Bloomfield, Review of Liebich, Konkordanz das Pini-Candra, Language, 5, 267-276,
1929]
INDOLOGISTS ON PINIS ADHYY
The idea that a language is based on a system of rules determining the
interpretation of its infinitely many sentences is by no means novel. Well
over a century ago, it was expressed with reasonable clarity by Wilhelm
von Humboldt in his famous but rarely studied introduction to general
linguistics (Humboldt 1836). His view that a language makes infinite use
of finite means and that a grammar must describe the process that makes
this possible.. Pinis grammar can be interpreted as a fragment of such a
generative grammar in essentially the contemporary sense of this term.
[N. Chomsky, Aspects of the Theory of Syntax, MIT Press, 1964, p.v]

Modern linguistics acknowledges it as the most complete generative
grammar of any language yet written and continues to adopt technical
ideas from it .
[P.Kiparsky, Pinian Linguistics, in Encyclopaedia of Language and Linguistics, VI, 1994]
INDOLOGISTS ON PINIS ADHYY
The algebraic formulation of Pinis rules was not appreciated by the
first Western students; they regarded the work as abstruse or artificial. ...
The Western critique was muted and eventually turned into praise when
modern schools of linguistics developed sophisticated notation systems of
their own. Grammars that derive words and sentences from basic elements
by a string of rules are obviously in greater need of symbolic code than
paradigmatic or direct method practical grammars....
It is a sad observation that we did not learn more from Pini than we did,
that we recognised the value and the spirit of his "artificial" and "abstruse"
formulations only when we had independently constructed comparable
systems. The Indian New Logic (navya-nyya) had the same fate: only
after Western mathematicians had developed a formal logic of their own
and after this knowledge had reached a few Indologists, did the attitude
towards the navya-nyya school change from ridicule to respect.


H. Scharfe, Grammatical Literature, Wiesbaden 1977, p.112, 115.

INDOLOGISTS ON PINIS ADHYY
Pini has composed a list of formulae called stra...serving to form
words and sentences from a given material of minimal elements...It
comprises both lists of primary elements, and a program for the
combination of these elements. These elements are the phonemes, the
roots, group of words sharing a grammatical feature, morphemes
(suffixes) having a meaning ...
The program is made up of operating rules as well as conventions
necessary for the application of the rules. It is composed in a true meta-
language very apt to its purpose, achieving the maximum brevity, which
makes it easy to memorize, and is the first and foremost example of the
formalization of the technical exposition in the universal history of
sciences. Because of its practical objective and form, it cannot be
compared with a systematic grammar of a European type. By contrast, its
resemblance to a modern computer program is striking.
[P S Filliozat: The Sanskrit Language: An Overview, Indica Books, Varanasi 2000
(French Edition 1992), p. 24]
INDOLOGISTS ON PINIS ADHYY
Pini's grammar is universally admired for its insightful analysis of
Sanskrit...Generative linguists for their part have marvelled especially at
its ingenious technical devices, and at intricate system of conventions
governing rule application and rule interaction that it presupposes, which
seem to uncannily anticipate ideas of modern linguistic theory (if only
because many of them were originally borrowed from Pini in the first
place.)...
The grammar has four distinct components:
1. Adhyy: a system of about 4,000 grammatical rules
2. ivastras: the inventory of phonological segments...
3. Dhtupha: a list of about 2,000 verbal roots...
4. Gaapha: a list of 261 lists of lexical items ...
The grammar is a device that starts from meaning information... and
incrementally builds up a completely interpreted sentence.
[P.Kiparsky, On the Architecture of Pini's Grammar, 2002]
IVA-STRAS AND PRATYHRAS



Each stra has a set of varas followed by a marker (, K, , C, etc)
called the called the it vara [ ]
Pratyhras are formed by any of the varas and an it which follows it.
The pratyhra then stands for the class of varas enclosed by them except
for the intervening it-varas.
aK stands for {a, i, u, , }. iK stands for { i, u, , }
aC stands for all the vowels. haL stands for all the consonants
In this way about 300 pratyhras are possible; Pini uses 42 of them.
Recent studies show that the iva-stras give an optimal encoding for
these 42 partially ordered subsets of Sanskrit sounds.
METHOD OF ADHYY
Pinis Stras are mainly of the following types:
Vidhi-stra: Operational rules
Saj-stra: Rules which introduce class names and establish
conventions regarding the use of terms
Adhikra-stra: Headings
Paribh-stra: Metarules, which serve to interpret and regulate other
rules. They regulate the operations specified in the vidhi-stras:
ahsthneyog (1.1.49): Genitive designates in place of
Tasminnitinirdie prvasya (1.1.66): Locative defines the right
context
Yathsakhyamanudea samnm (1.3.10): For groups with the
same number of elements, the corresponding elements are to be
related in order.
Prvatrsiddham (8.2.1): (From now on every rule is regarded as)
not having taken effect with reference to preceding ones.
CONTEXT SENSITIVE RULES OF ADHYY



Example: Ikoyaaci (6.1.77)

CONTEXT SENSITIVE RULES OF ADHYY
ikoyaaci (6.1.77)
iK stands for {i, u, , },
ya stands for {y, v, r, l}
aC stands for all the vowels.
From 6.1.72, sahitym is carried forward. Thus the stra provides that:
i, u, , y, v, r, l before a vowel, in close contact
This gives
i + a y + a, u+ a v + a, and so on.
Aka savare drgha (6.1.101) is an apavda- stra to the above, and
gives:
i + i = , u+ u = , and so on.

PINI AND ZERO
Panini introduces the notion of zero-replacement (zero-phoneme, zero-
morpheme etc)
Adarana lopa (1.1.60) Non-appearance is zero.
There are about fifty stras where lopa appears explicitly and more than
hundred if we take into account anuvtti.
There are several other kinds of zeroes in Panini.
For instance, there are the it varas in pratyhras. [Tasya lopah (1.3.39)]
There are also luk, lu and lup which correspond to non-appearance of a
pratyaya or suffix.

VKYAPADYA ON STRA AS UPYA


" ~
: ( .)
Worldly activities are accomplished on the basis of different theories and
philosophies. What is important in one theory may not be so in another.
u
( S
( : (uU ( .-)
Upyas (procedures taught in stras) are to be discarded, even though
they are to be used for accomplishing an objective. There is no necessary
limitation on such means. One accomplishes objectives by one means or
the other.
As noted by the commentator Puyarja:
> ~ >
:

DEVELOPMENT OF CHADA-STRA
In his Chanda-stra (c.300 BCE), Pigala introduces some
combinatorial tools called pratyayas which can be employed to study the
various possible meters in Sanskrit prosody. Following are some of the
important texts which include a discussion of various pratyayas:
Pigala (c.300 BCE): Chanda-stra
Bharata (c.100 BCE): Nyastra
Brahmagupta (c.628): Brhmasphuasiddhnta
Virahka (c.650): Vttajtisamuccaya
Mahvra (c.850): Gaitasrasagraha
Halyudha (c.950): Mtasajvan Comm. on Pigalas Chanda-stra
Kedrabhaa (c.1000): Vttaratnkara
Ydavapraka (c.1000): Comm. on Pigalas Chandah-stra
Hemacandra (c.1200): Chandonusana
Prkta-paigala (c.1300)
Nryaapaita (c.1350): Gaitakaumud
Damodara (C.1500): Vabhaa
Nryaabhaa (c.1550): Nrya Comm. on Vttaratnkara
VARA-VTTA

A syllable (akara) is a vowel or a vowel with one more
consonants preceding it.
A syllable is laghu (light) if it has a short vowel. Even a
short syllable will be a guru if what follows is a conjunct
consonant, an anusvra or a visarga.
Otherwise the syllable is guru (heavy).
The last syllable of a foot of a metre is taken to be a guru
optionally.
? ? U
z " ~ ~
u u u
u u ?
GGG GLG GLL LLL LGG LGG LGG
THE EIGHT GAAS


Ya: LGG Ra: GLG Ta: GGL
Bha: GLL Ja: LGL Sa: LLG
Ma: GGG Na: LLL
The pattern of a metre is usually characterised in term of these gaas.
For instance the verse of Klidsa cited earlier is in Sragdhar metre:
9
Thus Sragdhar is characterised by the pattern: MaRaBhaNaYaYaYa,
with a break (yati) after seven syllables each.
GGGGLGG LLLLLLG GLGGLGG

A MNEMONIC FOR THE GAAS
There is the mnemonic attributed to Pini
YaMTRJaBhNaSaLaGam
L G G G L G L L L G
If we replace G by 0 and L by 1, we have a binary sequence of length 10
1 0 0 0 1 0 1 1 1 0
The above linear binary sequence generates all the 8 binary sequences of
length 3. We can remove the last pair 1, 0 and view the rest as a cyclic
binary sequence of length eight. In modern mathematics such sequences
are referred to as de Bruijn cycles.
[, stand for L, G or 1,0]
PRATYAYAS IN PIGALAS CHADA-STRA
In Chapter Eight of Chanda-stra, Pigala introduces the following six
pratyayas:
Prastra: A procedure by which all the possible metrical patterns with a
given number of syllables are laid out sequentially as an array.
Sakhy: The process of finding total number of metrical patterns (or
rows) in the prastra.
Naa: The process of finding for any row, with a given number, the
corresponding metrical pattern in the prastra.
Uddia: The process for finding, for any given metrical pattern, the
corresponding row number in the prastra.
Lagakriy: The process of finding the number of metrical forms with a
given number of Laghus (or Gurus).
Adhvayoga: The process of finding the space occupied by the prastra.

PRASTRA
z " " ~ (~ .-)
Form a G, L pair. Write them one below the other.
Insert on the right Gs and Ls.
[Repeating the process] we have eight (vasava) metric forms in the 3-
syllable-prastra
Single syllable prastra


Two syllable prastra



1 G
2 L
1 G G
2 L G
3 G L
4 L L
PRASTRA
Three syllable prastra





Another method of generating the prastra
U

U ( .-)
Start with a row of Gs. Scan from the left to identify the first G. Place
an L below that. The elements to the right are brought down as they are.
All the places to the left are filled up by Gs. Go on till a row of only Ls
is reached.
1 G G G
2 L G G
3 G L G
4 L L G
5 G G L
6 L G L
7 G L L
8 L L L
SAKHY
z z (~ .-)
The number of metres of n-syllables is S
n
= 2
n
. Pigala gives an optimal
algorithm for finding 2
n
by means of multiplication and squaring operations
much less than n in number.
Halve the number and mark "2"
If the number cannot be halved deduct one and mark "0"
[Proceed till you reach zero. Start with 1 and scan the sequence from the right]
If "0", multiply by 2
If "2", square
Example: Six-syllable metres n = 6
6/2 = 3 and mark "2"
3 cannot be halved (3-1) and mark "0"
2/2 = 1 and mark "2"
1-1 = 0 and mark "0"
Sequence 2, 0, 2, 0 yields 1x2, (1x2)
2
, (1x2)
2
x2, ((1x2)
2
x2)
2
= 2
6

Pigalas algorithm became the standard method for computing powers in
Indian mathematics.
SAKHY
Next stra of Pingala gives the sum of all the sakhys S
r
for r = 1, 2, ...n.
zU (~ .)
S
1
+ S
2
+ S
3
+ ... + S
n
= 2S
n
-1
Then comes the stra
(~ .)
S
n+1
= 2S
n

Together, the two sutras imply
S
n
= 2
n

and
1 + 2 + 2
2
+ ... + 2
n
= 2
n+1
-1
The latter clearly is the formula for the sum of a geometric series.
SAKHY
The Sakhya 2
n
discussed above is for the case of syllabic metres of n-
syllables which are sama-vttas metres which have the same pattern in
all the four pdas or quarters.
Ardhasama-vttas are those metres which are not sama, but whose halves
are the same. Viama- vttas are those whose pdas are all different.
In the fifth Chapter of Chanda-stra, Pigala has dealt with the sakhy
of Ardhasama and Viama- vttas.
((~ .-)
The number of Ardhasama-vttas with n-syllables in each pda is (2
n
)
2
-2
n

In the same way, the number of Viama-vttas with n-syllables in each
pda is [(2
n
)
2
-2
n
]
2
- [(2
n
)
2
-2
n
]

NAA
(~ .-)
To find the metric pattern in a row of the prastra, start with the row-
number
Halve it (if possible) and write an L
If it cannot be halved, add one and halve and write a G
Proceed till all the syllables of the metre are found.
Example: Find the 7
th
metrical form in a 4-syllable prastra
(7+1)/2 = 4 Hence G
4/2 = 2 Hence GL
2/2 = 1 Hence GLL
(1+1)/2 = 1 Hence GLLG

If we set G=0 and L = 1, we can see that Pigalas naa process leads to
the desired metric form via the binary expansion
7 = 0 + 1x2 + 1 x 2
2
+ 0 x 2
3

FOUR-SYLLABLE PRASTRA
1 G G G G
2 L G G G
3 G L G G
4 L L G G
5 G G L G
6 L G L G
7 G L L G
8 L L L G
9 G G G L
10 L G G L
11 G L G L
12 L L G L
13 G G L L
14 L G L L
15 G L L L
16 L L L L

If we set G=0 and L=1, then we see that each metric pattern is the mirror
reflection of the binary representation of the associated row-number-1.
UDDIA
u zU j (~ .-)
To find the row number of a given metric pattern, scan the pattern from
the right.
Start with number 1
Double it when an L is encountered.
Double and reduce by 1 when a G is encountered
Example: To find the row-number of the pattern GLLG in a 4-syllable
prastra:
Start with 1. It is unchanged with first G
Then we find L. So we get 1x2 = 2
Then we find L. So we get 2x2 = 4
Finally we have G. We get 4x2-1 = 7


UDDIA
Another Method:
(? zU
" ( .)
Place 1 on top of the left-most syllable of the given metrical pattern
Double it at each step while moving right.
Sum the numbers above L and add 1 to get the row-number
Example: To find the row-number of the pattern GLLG
1 2 2
2
2
3

G L L G
Row-Number = 0 x 1 + 1 x 2 + 1 x 2
2
+ 0 x 2
3
+ 1= 7
Both the naa and uddia process of Pigala are essentially based on the
fact that every natural number has a unique binary representation: It can be
uniquely represented as a sum of different sakhy S
n
or the powers of 2.
LAGAKRIY
(~ . )
Pigalas stra on lagakriy process is too brief. Halyudha, the tenth
century commentator explains it as giving the basic rule for the
construction of the table of numbers which he refers to as the Meru-
prastra.
? S + Sz
q? u
u S ~ u S
S S U S
Sz
S S Sz

VARA-MERU OF PIGALA

Clearly the number of metrical forms with r Gurus (or Laghus) in the
prastra of metres of n-syllables is the binomial coefficient
n
C
r

The above passage of Halyudha shows that the basic rule for the
construction of the above table, is the recurrence relation
n
C
r

=
n-1
C
r-1
+
n-1
C
r


PASCAL TRIANGLE
The above Vara-Meru is actually a rotated version of the so called Pascal
Triangle (c. 1655) shown below:





REFERENCES
1. G.Cardona, Pini A Survey of Research, Mouton, The Hague 1976.
2. R.N.Sharma, The Adhyy of Pini, 6 Volumes, Munshiram
Manoharlal, Delhi 1990-2003.
3. G.Cardona, Pini His Work and its Traditions, 2ed, Motilal
Banarsidas, New Delhi 1997
4. G.Huet and A.Kulkarni and P.Scharf Eds., Sanskrit Computational
Linguistics, Springer, New York 2009
5. Chandastra of Pigala with Comm. Mtasajvan of Halyudha
Bhaa, Ed. Kedarnath, 3ed. Bombay 1938.
6. Vttaratnkara of Kedra with Comms. Nrya and Setu, Ed.
Madhusudana Sastri, Chaukhambha, Varanasi 1994.
7. B. van Nooten, Binary Numbers in Indian Antiquity, Jour. Ind. Phil.
21, 1993, pp.31-50.
8. R.Sridharan, Sanskrit Prosody, Pigala Stras and Binary Arithmetic, in
G.G.Emch et al Eds., Contributions to the History of Indian
Mathematics, Hindustan Book Agency, Delhi 2005, pp. 33-62.