Sandhi and Sandhi-vichchhed

Sandhi means the joining of sounds with a visible or audible change at the boundary between two units. In Hindi grammar for competitive exams, the two practical questions are simple: what two parts joined, and what change occurred at the joining point? A sandhi word is the formed word, such as vidyalaya, sadachara, jagadisha, nissandeha, or manoranjana. Sandhi-vichchhed is the reverse analysis: vidyalaya = vidya + alaya, sadachara = sat + achara, jagadisha = jagat + isha, nissandeha = ni: + sandeha, and manoranjana = manah + ranjana. The exam does not reward merely saying that two words are joined; it rewards spotting the exact boundary and the exact rule.

According to the RPSC official syllabus for the Sub Inspector/Platoon Commander Competitive Examination-2016, Paper I Hindi carries 200 maximum marks. That exam frame matters because sandhi is not being tested as literary ornament; it is being tested as an objective grammar skill. The syllabus places sandhi and sandhi-vichchhed under word formation, so the preparation problem is not to write a long theory note. It is to recognise a boundary, apply the correct sound-change rule, and reject options that look familiar but do not reconstruct the printed word. Long theoretical lists help only after the common examples are stable; until then, the candidate should work from the printed form back to the source form.

Three terms should be kept separate. Sanyog is simple coming together, but sandhi requires a sound change. If two parts are written side by side with no change, that may be a compound or phrase, not necessarily sandhi. Sandhi-shabd is the output after transformation. Sandhi-vichchhed is not any meaningful division; it is the division that recreates the output by rule. For example, sadachara may look like sad + achara after formation, but the tested explanation is sat + achara, because final t becomes d before the vowel-led voiced environment. Similarly, sajjana is not sa + jana; it is sat + jana with consonant change. A correct split must therefore be both meaningful and rule-producing; meaning alone is not enough.

In exam work, begin from the visible join. If the middle of the word contains a long vowel such as a, i, u, e, ai, o, or au in the transliterated form, test swar sandhi first. If the middle contains doubled consonants, voiced stops, nasal markers, or changes around t and d, test vyanjan sandhi. If the source part may end in visarg, test visarg sandhi. Do not assume that every Sanskrit-looking word has sandhi. A word can be tatsam or formal without being sandhi-derived for the purpose of the option. Conversely, common Hindi words such as upadesha, sadachara, namaskara, and manoranjana are testable because their traditional splits are rule-bound.

A useful way to remember the scope is this: sandhi is about a sound change at contact, while samasa is about compound formation and pratyaya or upasarga questions are about affixes. These topics often sit close to one another in Hindi grammar, so a wrong option may tempt the candidate into treating a compound-like word as sandhi or treating a sandhi word as a plain compound. The boundary test prevents that confusion. If no sound change is needed to explain the word, it is not a sandhi answer merely because two meaningful parts can be imagined.

The cleanest working habit is to treat each option as a reconstruction claim. The proposed split is the evidence, the rule is the bridge, and the final printed word is the result to be tested. If any one of those elements fails, the option fails. This prevents two common mistakes: accepting a split because both parts sound familiar, and rejecting a correct traditional split because the intermediate change is not visible in the formed word. In sandhi, the original boundary often has to be restored from behind the final spelling.

A reliable option-comparison routine is: identify the proposed parts, join their boundary sounds, apply the relevant rule, and see whether the exact printed word returns. If the answer option says jagadisha = jag + disha, it fails because disha is not the real second unit and the t-to-d change is ignored. If it says jagadisha = jagat + isha, it succeeds because t changes into d before the vowel-led second unit. Sandhi questions are compact, but they are not guesswork; every correct option must survive this reverse reconstruction.