HookThe exam where a Cambridge professor proved cleverness isn't enough
In the 1920s the Cambridge critic I.A. Richards ran an experiment that still shapes your exam. He handed his undergraduates poems with the author, title and date stripped away — no reputation to lean on, no period to place them in — and asked them to write down what they thought. The results, published as Practical Criticism in 1929, were humbling: intelligent students at one of the best universities in the world produced confident nonsense, praising doggerel and dismissing masterpieces, because without the crutches of author and context they had never actually learned to read the words on the page. Richards had invented 'close reading', and he had also, by accident, invented the unseen exam.
Section B of Paper 1 is Richards' experiment with a mark scheme. You get two poems you have never met, both on some aspect of love, and one compulsory question asking you to compare how the poets present it. There is no author to research, no context to recite, no critic to borrow — Section B assesses AO1 and AO2 only. It is just you and the words, under time pressure, worth 25 marks. That sounds frightening until you realise the flip side: because nobody can revise the poems, the whole test is a single transferable skill, and skills can be trained. Learn to read form, image and sound quickly, and to hold two poems in one argument, and the surprise stops mattering.
ModelForm and structure — the shape is an argument
Before you read a single image, read the shape on the page, because form is the poet's first decision and AO2 rewards it heavily. Is it a sonnet — fourteen lines pretending to argue their way to a resolution — or loose free verse that refuses the neatness love poems traditionally promise? Count the stanzas; regular stanzas suggest control, a poem straining against irregular ones suggests feeling that will not sit still. Above all, hunt for the volta, the turn — the 'but', 'yet' or 'and now' where a poem pivots from problem to resolution, or from certainty to doubt. Where the turn falls, and whether the poem earns it or fakes it, is often the whole reading.
Structure is movement across the poem: where it starts, where its focus shifts, how it ends. Does it circle back to its opening image, leave you on an unresolved line, or end mid-thought with enjambment spilling past the final full stop? Name the shape, then say what the shape does to the feeling — that is structural AO2, and most candidates never claim it.
ModelImagery, diction and tone — where the meaning hides
The heart of AO2 is the poet's choice of words and images. Diction is word choice: is the register tender, clinical, playful, bitter? A love poem that reaches for the vocabulary of accountancy or warfare is telling you something the surface may deny. Imagery is the picture a word builds — and love poetry runs on it, from the extended metaphor or conceit that stretches a single comparison across a whole poem, to the small telling detail that carries the emotional weight. Tone is the attitude the poem takes to its subject, and it is the thing weak answers miss entirely: a poem can say 'I love you' in a voice that is adoring, resigned, ironic or afraid, and the tone, not the statement, is the meaning.
The technique is the same chain every time: one exact word, what it connotes, the effect it creates. Do not write 'the poet uses imagery.' Write about this word and what it makes you feel, and why the poet reached for it instead of the obvious alternative.
Take an original demonstration poem written for this lesson (so no set text is reproduced): 'Come, let me count you in the dark: / one for the hollow of your throat, / two for the pulse behind your wrist, / three for the word you did not say.' A model AO2 paragraph: 'The speaker turns intimacy into arithmetic — count you, one, two, three — and the cold diction of counting sits strangely against the tender act of touch, as though love could be tallied and kept. Yet each number lands on a fragment of the beloved's body rather than a whole person: a hollow, a pulse, a wrist. The imagery is anatomical, almost inventory-like, and the effect is unsettling — this is desire that possesses by dismantling. The poem then withholds its own climax: the count reaches three only to arrive at the word you did not say, so the final image is an absence, and the tenderness curdles into something closer to loss.' Notice the moving parts: an exact word quoted, its connotation unpacked, the tonal shift named, and the structure's final turn read for effect — depth on four short lines, not a list of devices.
MechanismMetre, rhythm and sound — the marks you claim in sixty seconds
Under exam pressure most candidates abandon sound analysis, which is exactly why it lifts you above them. You do not need to scan every line. You need a fast ear for where the sound and the sense pull together or apart. Is there a regular iambic pulse suggesting control, and does it stumble at the emotional crux? Is there heavy enjambment making the voice breathless and headlong, or end-stopped lines making it measured and final? A caesura — a hard pause mid-line — can enact hesitation or a change of heart. Listen for half-rhyme (a rhyme that almost lands but doesn't): in a love poem it can suggest a relationship that not quite works, harmony withheld.
A sixty-second sound checklist: tap the rhythm of two or three key lines, mark the loudest pause, note whether rhyme is full, absent or slightly off, and ask one question — does the sound reinforce the feeling or undercut it? One sentence connecting a specific sound effect to meaning is worth more than a full, unlinked scansion.
CaseConfident inference — read the gaps, take a position (AO1)
AO2 handles method; AO1 handles the reading you build from it, and at the top it demands a confident, personal interpretation rather than a cautious paraphrase. Unseen poems reward inference because they so often withhold: the situation is implied, the relationship half-sketched, the ending deliberately open. The strong candidate leans into the ambiguity and commits to a reading, while acknowledging the alternative — 'the poem can be read as tender, but its clipped syntax lets a colder interpretation stand.' That is not fence-sitting; it is control.
Trust the detail over the vibe. A first impression that a poem is 'romantic' is worth nothing until it is tested against the words, and the words often resist it. Some of the best love poems are angry, ironic, grieving or possessive, and a reader who notices the poem's real temperature — not the temperature love poems are 'supposed' to have — is already writing in the upper bands.
ModelComparing two poems in one argument, not two (AO1)
The commonest structural failure in Section B is the 'slab' essay: everything about Poem A, then everything about Poem B, with a hopeful sentence of comparison at the end. It caps your AO1 mark because the objective is a balanced, integrated comparison, and integration means the two poems meet in every paragraph. Organise by idea, not by poem. Each paragraph takes one point of comparison — how each opens, how each handles the beloved's body, how each ends — and moves between the poems using real comparative connectives: similarly, whereas, by contrast, where the first insists, the second doubts.
Aim for genuine balance. If one poem gets three paragraphs and the other gets one, the comparison is not balanced and the mark reflects it. Plan the link first — the single sentence about how the two poems relate — and build outward from it, so the comparison is the spine of the essay rather than an afterthought stapled to the end.
Pair the counting poem above with a second original demonstration poem: 'She keeps my letters in a tin / that once held toffees, on the shelf; / she says she'll read them all again / when she has time. She never will.' A model integrated sentence: 'Where the first poem possesses through touch, counting the beloved's body in the present tense, the second measures love by its neglect — the letters shut in a sweet-tin, promised a rereading that the flat final clause, She never will, quietly refuses; both poems, then, locate love in a small physical object, but the first makes the object a site of desire and the second a site of quiet abandonment.' See the AO1 machinery: both poems in one sentence, a precise connective (where… the second…), a shared idea (love housed in an object) and a sharp divergence (desire versus abandonment) — one thread, two poems, no slab.
ModelExpression and terminology — precise, not decorated (AO1)
AO1 also credits the quality of your writing and the accuracy of your terms. Use the vocabulary of poetry — volta, enjambment, caesura, conceit, half-rhyme, free verse — but only as tools that let you say something exact. 'The poet uses enjambment' is inert; 'the enjambment spills the vow across the line-break, so the promise runs on faster than the speaker can control it' is analysis. The test before writing any technical word is whether you can immediately say what it does.
Expression itself is assessed. Write in continuous, argued prose with clear topic sentences, embed short quotations inside your own sentences rather than dropping them alone, and keep one line of argument visible throughout. The band descriptors reward 'coherent' and 'accurate' expression at every level and 'sophisticated' control at the top — which in practice means a reader never has to work out what you mean.
VocabularyKey terms the mark scheme pays for
TrapsMisconceptions that cost marks
ExamWhat examiners want
Section B is a single compulsory 25-mark essay comparing two unseen poems, and it is marked on AO1 and AO2 only — no AO3 context, no AO5 critics. Do not waste time inventing a biography or a period for poems that have neither; every mark is in the reading and the comparison. The band descriptors climb from 'general', 'descriptive' work through 'clear' and 'coherent' analysis to 'perceptive', 'assured' and 'sophisticated' comparison at the top, and the ceiling is set by two things: the depth of your method analysis and the genuineness of the comparison.
Manage the pressure with a routine. Spend five minutes reading both poems twice and jotting the one-sentence link between them before you write a word — candidates who plan the connection outscore those who discover it in their final paragraph. Then build the essay by idea, not by poem, so the two texts meet in every paragraph with real comparative connectives. Keep the poems balanced; an essay that neglects one poem cannot reach the top band however brilliant it is on the other. Quote briefly and embed, analyse method for effect rather than naming it, and commit to a confident personal reading while acknowledging the alternative the words allow. Precision and balance win here — two poems, one argument.