UPMSP Intermediate Pali common mistakes
UPMSP Intermediate Pali · Understand the traps and how to avoid them.
When a passage is printed in Devanagari but the question asks for transliteration, candidates copy the words back in Devanagari or mix scripts mid-answer, and they confuse letters that look similar across scripts (ba/va, ta/ṭa, da/ḍa, sa/śa/ṣa).
The prescribed syllabus is delivered in Devanagari, so candidates read fluently in one script but rarely practise converting to Roman (IAST) or to another script. Under exam pressure they default to the script they read most, and the niggling visual similarity of aspirated and retroflex letters is never drilled.
What to do instead
Build a two-column conversion table of the Pali alphabet (vowels, consonants, conjuncts, anusvāra, visarga) in Devanagari and IAST, then transliterate ten short sentences daily in both directions. Before writing, mark retroflex (ṭ, ḍ, ṇ) and sibilants (ś, ṣ, s) in the source with a pencil so the conversion is deliberate.
Candidates translate a Pali sentence word-by-word in the order printed, producing Pali word order inside Hindi/English, and they leave case endings untranslated (e.g., rendering rājena as 'king' rather than 'by the king').
School translation habits from Hindi/English into Pali reverse, but candidates rarely practise Pali into the target language. Pali is heavily inflectional with free word order, and the case system (eight cases, three numbers) is memorised as a table rather than applied during reading.
What to do instead
Drill the eight cases with one noun paradigm (e.g., buddha, rāja, munī) until instrumental, dative, ablative, genitive and locative endings are automatic. Translate in three passes: (1) locate the verb and mark it, (2) tag each noun's case and number in the margin, (3) recast into target-language order. Practise ten sentences per day with the ending-tagging step written out.
Candidates answer 'who wrote this text' or 'which nikāya does this sutta belong to' from vague memory, and they attribute commentarial material to the canon or the canon to a commentator (e.g., crediting Buddhaghosa with a sutta or the Dhammapada with a commentary).
Literature study is often reduced to a list of names memorised without a framework, so canonical/non-canonical, sutta/commentary and author/text boundaries blur. Candidates also revise from summaries that omit authorship.
What to do instead
Maintain a three-column register for every prescribed text: title, basket/collection (Vinaya, Sutta, Abhidhamma; Dīgha, Majjhima, Saṃyutta, Aṅguttara, Khuddaka), and author/commentator if any. Add a one-line content tag. Recite the register weekly and test yourself by covering one column.
In prose comprehension, candidates lift an entire sentence from the passage as the answer to a specific short question, instead of stating the point in their own words, and they miss the question's scope (e.g., asking only about the cause, not the whole episode).
Fear of losing marks makes candidates copy safe-looking text; they also read the question after the passage rather than before, so they do not know what to extract. The result is an answer that is too long and does not address the exact point asked.
What to do instead
Read the questions first, underline the key interrogative (ka, kena, kasmā, kathaṃ), then read the passage hunting only for that information. Answer in one or two sentences of your own, then check that every part of the question is covered. Practise with past passages under timed conditions.
On verse passages, candidates translate each line independently and produce a garbled meaning, ignoring that Pali verse uses inverted word order, ellipsis of the verb, and metrical filler particles (eva, hi, kho, pana) that carry little meaning.
Verse is compressed and often lacks an explicit verb; the sandhi and metre obscure word boundaries. Candidates trained on prose do not know that the sentence may run across two or three lines and that the verb may appear at the very end or be implied.
What to do instead
Before translating, scan the metre and mark the caesura; then bracket the sentence boundaries across lines and supply the missing verb in brackets. Treat eva, hi, kho and similar particles as optional in translation. Practise by re-ordering a verse into prose word order before rendering it.
Candidates conflate canonical and commentarial doctrine, attributing later systematised categories (e.g., detailed citta classifications, the full list of 52 cetasikas) to the suttas themselves, or treating Abhidhamma definitions as if they were narrative sutta statements.
Prescribed textbooks often present doctrine as a single unified system, and candidates do not distinguish the strata: sutta, abhidhamma and commentary developed over centuries. The examination rewards precision about source, not just the content.
What to do instead
For each doctrinal topic (e.g., paṭiccasamuppāda, khandha, satipaṭṭhāna), note the earliest canonical locus and any commentarial elaboration separately. Keep a two-column sheet: 'sutta statement' versus 'commentarial/abhidhamma development'. Cite the text, not just the doctrine, in answers.
Candidates explain a philosophical term by giving a one-word gloss (e.g., anicca = 'impermanent') without unpacking its doctrinal import, and they confuse near-neighbours such as anattā/aniccā, vedanā/viññāṇa, samādhi/sati, or nibbāna/nirodha.
Vocabulary is learned as a bilingual list, not as a concept. Since many Pali terms have no single target-language equivalent, the gloss feels sufficient, and candidates do not practise writing explanatory paragraphs that use the term in context.
What to do instead
For each key term, write three things: literal analysis (prefix + root), a one-line definition, and one canonical example sentence with translation. Pair confusing terms in a contrast table (what each does and does not include). In answers, always give the term, its analysis, its definition, and its textual illustration.
In composition questions, candidates write grammatically impossible sentences because they copy Hindi/English syntax into Pali, use the wrong tense or voice ending, and ignore agreement between adjective and noun in gender, number and case.
Composition is the least practised skill; candidates can decode but not encode. They also do not internalise that Pali adjectives must agree with their nouns and that the verb's person and number must match the subject, since their reading practice never forces production.
What to do instead
Write five short sentences daily using a fixed frame (subject + object + verb), then check three things: case of each noun, agreement of each adjective, person/number of the verb. Keep a checklist of common endings and do a formal self-correction pass before finalising.
Candidates apply sandhi rules mechanically without reversing them: they fail to split a compound or sandhi-joined form (e.g., reading a resolved word as a single unit, or splitting where no boundary exists) and then translate the wrong stem.
Sandhi is taught as a list of rules for joining, but exam passages require the reverse operation, splitting. Candidates also do not know the common compound types (kammadhāraya, tappurisa, bahubbīhi, dvandva), so they cannot predict where a boundary should fall.
What to do instead
Practise splitting: take twenty sandhi-joined forms daily and resolve them into original stems, then rejoin to verify. Learn the four main compound types with one example each and identify the type before translating. Read the resolved form aloud to check that each stem is a real word.
Candidates give the historical and cultural background of a text as a general essay (place, date, dynasty) without connecting it to the specific passage or doctrine asked about, and they assert dates or patronage facts they cannot support.
Background material is memorised as disconnected narrative, and candidates pad answers to reach length. When the question asks how the cultural context illuminates a passage, they produce a timeline instead of an interpretation.
What to do instead
For each prescribed text, prepare three linked points: where and for whom it was composed, what doctrinal or social issue it addresses, and one line from the text that shows it. In answers, state the context in one sentence, then spend the rest of the space linking it to the passage. Omit any date or patronage claim you cannot source.