Two stems for one root: diagnosing a dhatu-rupa table and a sandhi sheet
A ten-mark grammar test has two halves: the present-tense rūpa table of the root kṛ (nine forms, 5 marks) and five sandhi items (5 marks). Arjun's table reads: karomi, karoṣi, karati / karvaḥ, karuthaḥ, kurutaḥ / kurmaḥ, karutha, karavanti. Four of the nine are correct, so he scores 2/5. His sandhi answers are: nara + indra → naraindraḥ; vana + adhipaḥ → vanaadhipaḥ; iti + api → iti api; sūrya + udayaḥ → sūryodayaḥ; muni + indra → munīndraḥ. Two of five are correct, so 2/5 again, total 4/10. In conversation he can recite 'karoti' instantly and can state the rule 'a + i becomes e', yet he did not apply it in writing. The teacher has ten minutes of the next class and must decide which half to drill, knowing that the same errors will cost more in the board paper where grammar carries a larger block.
1. Classify Arjun's nine table errors and state the single misconception behind them. Give the corrected table and the block rule that would prevent the error.
Read the worked answer
His errors are karati (for karoti), karvaḥ (kurvaḥ), karuthaḥ (kuruthaḥ), karutha (kurutha) and karavanti (kurvanti); he is correct in karomi, karoṣi, kurutaḥ and kurmaḥ.
The misconception is one stem applied everywhere. He has learned the strong grade kar- and attaches every ending to it, even inventing a -vanti on kar-. The root kṛ in the present alternates: strong forms take kar- (karomi, karoṣi, karoti), the remaining forms take the weak stem kur- (kurvaḥ, kuruthaḥ, kurutaḥ, kurmaḥ, kurutha, kurvanti).
Block rule to teach: 'mi, si, ti → kar-; everything else → kur-', and drill the third person first because karoti and kurvanti are the two forms that appear most in prose passages. A cross-check he can run alone: if a written form contains two consecutive full vowels (karauti, karavanti), it is wrong; the weak stem never shows the a of kar-.
2. Take the three wrong sandhi items. Give the correct form, name the rule, and say why 'naraindraḥ' is not merely misspelt but impossible.
Read the worked answer
- nara + indra → narendraḥ: guṇa sandhi, a + i = e (ād guṇaḥ).
- vana + adhipaḥ → vanādhipaḥ: savarṇa-dīrgha, a + a = ā (akaḥ savarṇe dīrghaḥ).
- iti + api → ityapi: yaṇ sandhi, i before a vowel becomes y (iko yaṇ aci).
- His two correct items, sūryodayaḥ (guṇa, a + u = o) and munīndraḥ (savarṇa-dīrgha, i + i = ī), show that he knows the rules in the abstract.
'naraindraḥ' is not a spelling slip: Sanskrit orthography does not write two vowels in hiatus across a word boundary in this position, so the string violates the sandhi system rather than a convention. The oral test that repairs it: say the two words fast - 'nara-indra' collapses to 'narendra' - and write what the mouth says. His pattern is recognition without production: he can recite 'a + i = e' but does not scan his own output for a hiatus, so the drill must be write-and-check, not recite.
3. You have ten minutes. Which half do you drill, in what order, and why not simply split the time equally?
Read the worked answer
Drill the dhatu table first, for about six minutes, then sandhi for four.
Reasoning: both halves are worth 5 marks, but the error types differ in cost. The dhatu errors are production errors - he cannot generate the form - and production errors need choral repetition plus writing, which takes time but fixes five marks. The sandhi errors are application errors on rules he can already state, so a four-minute checklist (scan for two adjacent vowels; if found, name the pair and apply the rule) recovers three marks cheaply.
Plan: two minutes of choral drill on karomi/karoṣi/karoti, two minutes on the kur- block, two minutes writing the whole table from memory with the book closed; then one minute on the hiatus checklist, three minutes on ten sandhi pairs in writing, marking his own work.
Equal time would be wrong because the sandhi half is nearer to mastery: he needs a checking habit, not new knowledge. The dhatu half needs a new mental organisation - two stems - and that cannot be done in four minutes.