Sandhi — Sound Joining
Sandhi (literally 'joining') explains why Sanskrit words change when placed next to each other. Once you understand the basic rules, many mysterious spellings become logical.
WHAT IS SANDHI?
When two Sanskrit words are placed together, their boundary sounds often merge or change to flow smoothly. This happens in English too — 'an apple' not 'a apple.' In Sanskrit, these rules are systematic and predictable.
Example from BG 2.20: na + ayam = nāyam (not + this = 'this [soul] not')
Common Sandhi Rules for Gita Reading
Two 'a' vowels merge into ā
mama + ātmā = mamātmā
'my' + 'soul' = 'my soul.' The a+a contracts to ā.
Final 'a' before vowel becomes nothing
tat + eva = tateva → tattvam eva
Often 'a' simply drops before another vowel.
Final consonant softens before voiced sound
tat + bhavati = tadbhavati
'that becomes' — the t softens to d before b.
's' after 'a' becomes 'o'
yaḥ + iti = ya iti
Visarga (ḥ) rules are complex — in practice, just accept the merged form and split words by meaning.
Anusvara (ṃ) is simply nasal
saṃsāra → sam-sāra
The ṃ before s sounds like 'n'. Before m, it sounds like 'm'.
PRACTICAL TIP
You don't need to memorize all sandhi rules to read the Gita. Use a word-by-word parsing resource alongside the transliteration. Over time, you'll recognize common joins automatically. Focus on vocabulary first — sandhi will make sense as you go.