Sandhi Splits That Recombine but Do Not Restore the Original Pada
A pre-board Sanskrit paper's 5-mark संधि section asks for विच्छेद of five items, 1 mark each. The key prints: 2. नरेन्द्रः = नर + इन्द्र; 4. तच्छ्रुत्वा = तत् + श्रुत्वा; 5. जगदीशः = जगत् + ईशः. Scripts show three splits for item 2 - (a) नर + इन्द्र, (b) नर + इन्द्रः, (c) नरेन्द्र + अः - and two for item 4: (a) तत् + श्रुत्वा, (b) तच् + श्रुत्वा. One script gives जगद् + ईशः for item 5. The teacher has 58 papers and a 90-minute window. A departmental note says a विच्छेद that does not restore the original पद is wrong. Two colleagues disagree: one accepts any split that recombines to the printed form, the other insists on the pre-sandhi पद only. The head examiner has asked for a written rule before the section is marked.
1. Rule on the three splits offered for नरेन्द्रः. Which earn the mark and which do not?
Read the worked answer
Accept (a) and (b); reject (c). The practical test of a विच्छेद is recombination: नर + इन्द्र → नरेन्द्र and नर + इन्द्रः → नरेन्द्रः both reproduce the printed form. The only difference is where the nominative -ः is lodged - on the whole pada in (a), on the final member in (b). Both conventions are taught in schools, and unless the key says 'प्रातिपदिकयोः विच्छेदः' (stems only), penalising (b) is unfair. (c) fails differently: it detaches -अः from a stem that never carried it, inventing a phantom पद नरेन्द्र. Award 1 to (a) and (b), 0 to (c), and note the variant on the key so all 58 scripts are treated alike.
2. तच् + श्रुत्वा and जगद् + ईशः look phonetically plausible. Are they correct splits, and how should they be marked?
Read the worked answer
Both earn 0. In तच्छ्रुत्वा the च् is produced by श्चुत्व (त् becomes च् before श्); the pre-sandhi पद is तत्, which exists independently, whereas तच् does not. जगदीशः is the same case: the pre-sandhi form is जगत्, and जगद् is itself a sandhi variant occurring before voiced sounds, not an independent पद. The departmental note is therefore applied consistently: a split must restore an independently attested पद, not a phonetic halfway house. If a colleague wishes to reward 'awareness of the change', that is a separate half-mark the scheme does not contain; inventing it would make the section unmarkable across 58 scripts.
3. Draft the one-line rule the head examiner asked for, with the operational checks a marker should apply.
Read the worked answer
Rule: 'विच्छेदः सिद्धेन पदेन भवेत्; संधिजन्यं रूपं पदत्वेन न ग्राह्यम्; विभक्त्यन्तं पदम् अन्तिमे अंशे स्थापनीयम्।' Operationally: (1) recombine the proposed split and confirm it reproduces the printed form; (2) confirm each member is an attested, independently occurring पद; (3) allow the case ending to stand on the final member, so नर + इन्द्रः is acceptable; (4) reject splits that use a sandhi-produced consonant such as च् or द् as if it were the original; (5) record every accepted variant on the key so that the rule is visible to a moderator and to any re-check.