Algorithms used for identifying the syllables in a Sanskrit word

Question

Could anyone give a reference to the best book or website for learning the algorithms used for identifying the syllables in a Sanskrit word, in a completely unambiguous way, just from a piece of text? If it can be explained in a post then please feel free to do so.

Thank-you in advance

Below is the information I have obtained so far:

---

SOURCE: Devavāṇīpraveśikā: An Introduction to the Sanskrit Language 3rd Ed, Page 18, section 2.23

A syllable is generally considered to be either a single vowel, or a consonant (or consonant cluster) followed by a vowel. “tat tvam asi, "you are that," would be syllabified as “ta-, ttva-, ma-, -si.”

---

SOURCE: The Cambridge Introduction to Sanskrit, Page 25, Chapter 2A, Section LIGHT AND HEAVY SYLLABLES

"Śakuntalā" is split into syllables in the following way "Śa-kun-ta-lā"

---

SOURCE: https://learnsanskritlanguage.com/grammar/sounds/syllabels/

It is very important that a syllable should always start with a consonant. However, it can start with a vowel if a syllable is at the beginning of the line then it starts with a vowel.

A syllable can end with a various number of consonants, but it must stop when a nasal and a stop appears.

---

SOURCE: http://prakrit.info/sanskrit/adhyayanavidhi.html

A syllable, akṣaram, is a unit of speech that contains the following elements:

1. an optional onset, which consists of one or more consonants;
2. an obligatory rime, which consists of:

• an obligatory nucleus, which consists of a vowel; and
• an optional coda, which consists of one or more consonants.

A syllable therefore has the pattern CVC (where C means “consonant,” V means “vowel,” and * means “zero or more”). A syllable can be thought of as a vowel and the consonants that are “attracted” to it. A word will always have as many syllables as it has vowels. To parse a word, or a larger phrase, into syllables, one must decide whether a given consonant goes with the preceding vowel (as a coda) or with the following vowel (as an onset); the general principle is to associate a consonant with the vowel that immediately follows it, if possible, and otherwise to associate it with the vowel that precedes it.

Answer 1

Part of Sanskrit syllabification is clear. Every vowel or diphthong is in a separate syllable. There are bisyllabic vowel sequences, which are typically indicated in transliteration with a space between the vowel letters in, but these arise from sentence sandhi (e.g. āi#e → ā e) and may not be of concern for your purposes (anyhow, look for the blanks between vowels). Also please note that r,l can function as vowels, hence ṛṣī is two syllables, not one. Syllables can be light or heavy: if the vowel is short (a i u ṛ ḷ) it may be light, and other vowels make a syllable heavy (ā ī ū e o ai au ṝ ḹ – macrons may be added non-distinctively to e o ai au).

Consonants have not figured into this, yet. The evidence for what happens to consonants is mainly based on poetic metrics, and the general rule is that a short vowel followed by at most one consonant then a vowel defines a light syllable, and everything else is heavy. In computing light syllables, it is important to know that the orthographic symbols anusvara and visarga "count" as consonants, even though they are written as vocalic annotations. In transliterations, these come out as "ṃ" and "ḥ" so there is little likelihood of confusion, if you are dealing with transliterations.

The facts of the language don't tell us how to syllabify an arbitrary V̆CCCCV sequence, except that at least the first consonant is likely to be a coda (making the first syllable heavy).

There are potentially esoteric linguistic arguments to be made based on aspects of Sanskrit phonology, for example /s/ deletes between /gh/ and /t/, where one can spin a tale about s not being syllabified, which would then result in automatic deletion (rather than an actual rule of Sanskrit). Theories of reduplication are relevant to this possibility, since Sanskrit has CV- reduplications, and some people take the choice of copied consonant (ti-ṣṭha, sa-smar) to say something about syllable structure.