Section II is five free-response questions in 115 minutes, worth 50% of your exam score. The set: (1) a Short Answer question of 6–8 subquestions on a syllabus passage; (2) a Translation scored in roughly 15 segments; (3) a Short Essay with two subquestions; and (4) and (5) the two Project Passage Short Essays — one prose, one poetry — each worth 11 points, split into a summary (Part A, 4 points) and an analytical essay (Part B, 7 points). Graders award points against discrete, pre-set units: segments on the translation, individual point-boxes everywhere else. Fluency earns nothing on its own — every point is tied to a specific task the prompt named. Answer in English unless a question explicitly asks for Latin.
Translate every Latin word visibly. The translation is scored in ~15 segments, each rewarding a literal, word-for-word rendering. Skip or smooth over a word — an adjective, a conjunction, an enclitic — and you cap that segment, no matter how elegant the English reads.
Answer exactly what's asked, then stop. A short-answer prompt that says "identify one and only one" means one; a second guess can void the point. If it asks for the case, give the case — not the case plus a translation you weren't asked for.
Cite the Latin, not a paraphrase. Part B of each project essay requires at least two citations: quote the Latin (or give line numbers) and translate or accurately paraphrase it, referring to more than a single word. A sharp interpretation with no anchored Latin earns the claim row but loses the evidence.
Read "in context" as syntax, not dictionary. When a prompt asks you to translate a phrase or give a word's meaning "in context," graders want the sense as it functions in that sentence. A bare dictionary gloss that ignores the surrounding grammar misses the point.
Scan when told to scan. A scansion subquestion is its own point inside the 6–8-point Short Answer. Mark the meter in any accepted notation. Skipping it because it feels minor forfeits one of very few available points.
Prompt (Question 2 — Translation). Translate the passage as literally as possible: At gemini lapsu delubra ad summa dracones / effugiunt saevaeque petunt Tritonidis arcem… (Aeneid 2.225–226; Tritonis = Minerva is glossed)
Model fragment. "But the twin serpents, with a glide, flee to the highest shrines and seek the citadel of savage Minerva…"
Where the points land. The rubric asks for a literal translation that accounts for every word, checked segment by segment before any credit for style. - At gemini … dracones → "But the twin serpents": At (but), gemini (twin), dracones (serpents) all rendered → segment earns full credit. - delubra ad summa → "to the highest shrines": ad, summa, delubra all present. ✓ - saevae … arcem → "seek the citadel of savage Minerva": saevae modifies Tritonidis. Drop the adjective — "the citadel of Minerva" — and this segment is capped, because saevae is a scored word.
A polished "the two dragons slither up to the temple" reads well but loses segments: it compresses lapsu, drops saevae, and blurs ad summa. Literal and complete beats fluent and thin.
115 minutes, five questions. Targets that track the official weighting: Short Answer ~15 min, Translation ~15 min, Short Essay ~25 min, and each Project essay ~30 min (15 + 15 + 25 + 30 + 30). Bank the fast, discrete points first — do the Short Answer and Translation before the essays. Cut-losses rule: if a translation segment stalls you, write your best literal attempt for every word and move on. A partial segment beats a blank one, and the two project essays hold nearly half the section's points — never leave them short on time.