Canonical substitution rule

The Ashtadhyayi may be considered a meta-language that describes the Sanskrit language, having its own special syntax and semantics. Every word in a sutra is significant, as also are the preceding 'relevant contexts'.

The canonical substitution rule is composed from terms in a sutra along with its 'relevant contexts'. The 'relevant contexts' for a sutra could include preceding sutras, or even overarching sutras, that must be read together with a particular sutra in order to complete the rule. The objective of a substitution rule is to unambiguously identify some specified pattern in the input and substitute it with some pattern, yielding the transformed output. For e.g., in BG1.42, a rule may perform the following transformation:

• identify the pattern '...i\e...' in the input 'hi\eSHAm' (the '\' indicates the end of the preceding term in the input)
• substitute the 'i' with 'y'
• yielding the transformed output 'hyeSHAm'

We must first obtain the underlying terms in a sutra after an analysis of 'euphonic combinations' (sandhi analysis). After that, we must identify the Case of each underlying term. These underlying terms are usually declined forms of special words (for e.g., 'ikaHa-(GEN-S)' refers to the 'ik' class) that are used in the Ashtadhyayi to denote various classes, and must not be confused with similar words, if any, found in a Sanskrit dictionary. The important Case suffixes used are:

GEN = apply rule to (substitute) this character ('in place of:sthAne') {1.1.49: SHaSHThi sthAneyogA}
LOC = apply rule to the immediately preceding character {1.1.66: tasminniti nirdiSHTe pUrvasya}
ABL = apply rule to the immediately succeeding character {1.1.67: tasmAdityuttarasya}
INS = apply rule to either the immediately preceding or succeeding character
NOM = apply this as the substitute character(s)

Hence, a rule that has the form

X-GEN Y-NOM Z-LOC

tells us that we must substitute 'X' by 'Y' if 'Z' follows the 'X'. This is the rule obtained from the sutra {6.1.77: iko yaNNachi} (i.e. ik-GEN yaNN-NOM ach-LOC).

NOTES:

• The order of these Cases may vary in each sutra, and additional terms may need to be included from the 'relevant contexts' (that are not explicit in the sutra).
• The manner in which the substitution must be done may be specified by a different sutra.
• For example, {6.1.87: AdguNNaHa} and {6.1.88:vRuddhirechi} tell us to replace the combination 'X' and 'Z' with 'Y' (and not just 'X' alone).
• This is indicated by the overarching sutra {6.1.84: ekaHa pUrvaparayoHa} that applies to all the sutras between {6.1.84} and {6.1.112}.
• Since X, Y, and Z usually refer to classes with multiple elements, other rules may be required to figure out which element of the class 'Y' to select.
• For example, in the sutra {6.1.77: iko yaNNachi}, the specific element chosen from the 'yaNN' class is decided by the sutra {1.1.50: sthAne'ntaratamaHa}, which tells us to choose the element 'most similar' to the 'ik' class element found in the input.
• The sutra {1.1.9: tulyAsyaprayatnaM savarNNam} tells us which letters have the same 'place of origin' (throat, palate, teeth, etc.) as well as 'effort' (consonants, semi-vowels, sibilants, vowels, etc. combined with aspiration, accent, etc.).
• Hence, if the 'ik' class element is the vowel 'i', then we can choose the substitution 'y' from the class 'yaNN' because the letter 'y' and 'i' in the two classes are similar in that both share the feature '+palatal'.
• Similarly, the letter 'v' (+labial, +dental) in the 'yaNN' class is similar to the vowel 'u' (+labial) in the 'ik' class because they share the feature '+labial'.

EXAMPLES:

• NOTE: See the section marked 'RULE:' in each sutra.
• NOTE: Underlying terms shown in square brackets indicate terms obtained from the 'relevant contexts' for that sutra.
• {6.1.77: iko yaNNachi}.
• {6.1.78: echo'yavAyAvaHa}, where the 'ach' is read from the immediately preceding sutra to complete the rule.
• {6.1.87: AdguNNaHa}, where the 'ach' is read from a preceding sutra (but not the immediately preceding sutra) to complete the rule.

Information on rule construction will be provided only for a few sutras, because the additional information that may be required about the 'relevant contexts' for each sutra is much too detailed for us to provide on this site.