Poetry reviewing has a reputation for being either gushing or savage, with very little in between. The truth is duller and more useful: a good review is a close reading made public. You look hard at the poems, work out what they are trying to do, and tell a reader — precisely and with evidence — whether they manage it. What follows is a working method for doing that with a collection you have just finished reading.
Read It Twice Before You Write a Word
The first reading should be quick and slightly greedy. Read the whole book in one or two sittings and take no notes. You are testing the reading experience, not annotating it. Where does your attention lift? Where does it sag?
Then go back with a pencil. Mark the first and last lines of any poem that stays with you — openings set terms, endings deliver or withhold. Circle repeated words, place names, colours and images. Underline the poem you would read aloud to somebody. If you got bored on page thirty, treat that as data rather than a personal failure, and try to say why.
Scale matters, too. A twenty-page pamphlet and a hundred-page New and Selected ask different things of a reviewer, and your reader needs to know which one you are holding.
Form: What Shape Is the Poem Making?
Form is not the box a poem gets poured into; it is part of the meaning. Look first at what happens on the page. Long unbroken lines or a narrow column? Neat couplets or solid blocks of text? Are the stanzas regular, or do they loosen as the poem goes on?
Then look at where sentences end. A sentence finishing mid-line reads differently from one that runs three lines past its stop. Enjambment is a form of suspense, and a poem that never uses it is making a claim about steadiness or control.
Received forms are choices, not housekeeping. A sonnet brings argument and a turn; a ghazal brings leaps and refrains; a pantoum brings obsession; a sestina brings grinding circularity; a prose poem brings pressure without the relief of line breaks. You do not have to enjoy tight metre to review it fairly, but if a poet adopts a form, test whether the form earns its keep. And watch for the exception: a single sonnet in a free-verse collection can land like a raised voice.
Three questions to ask of form
- Where does the poem breathe? Find the stanza that holds one idea, and the one that does not.
- What happens at the line break? Every line ending on a full stop is a very different poem from one that never does.
- What is the visual shape saying? A poem one line wide reads faster and tighter than one that fills the margin.
Voice: Who Is Speaking, and to Whom?
Voice is the hardest thing to describe and the thing readers most want described. Start with the plain facts. Who is speaking? To whom? In what register? Is the first person intimate or performing? Is it addressing a lover, a dead parent, a stranger, a version of itself? When the poem says you, does it mean the reader, or somebody inside the poem?
Do not collapse the speaker into the poet. The I of a poem is a construction. If a collection reads like confession, that is a style, and you can say so without pretending to know anything about the author's life.
Listen for register as well as tone: Latinate formality against short Anglo-Saxon words, slang, hymn, liturgy, dialect, technical vocabulary. Note the tense. Present-tense poems feel immediate; past-tense poems often feel like elegy. If the voice shifts across the book, ask whether that is range or inconsistency — naming that difference is often the spine of the whole review.
Imagery: Load-Bearing or Merely Decorative?
In poetry, images carry the argument. Ask of each one: is this doing a job, or is it standing about looking pretty? Decorative imagery tends to be general — moons, blood, dust, mirrors, bones — and it arrives unearned. Load-bearing imagery is specific, and it returns.
Track the returns. Water might begin as a bath and end as a flood. A house might be dismantled room by room across forty pages. When an image develops like that, say so; that is the difference between noticing and reviewing.
Be wary of using cliché as a verdict. A familiar image handled with precision can be renewed by its context, and a strange image can still be inert. The test is whether the image changes how you read the line around it.
Structure: The Collection as a Whole
A collection is not a shuffled file. Look at the order. What comes first? Opening poems set the terms of the book. What comes last? The final poem is usually what the poet wants you to close on. Are there named sections, and do those titles mean anything, or are they intervals?
Watch for changes of gear: a long poem arriving after ten short ones, a prose piece placed exactly where the reader needs a breath. In a pamphlet the arc is usually tighter, so one weak poem can sag the whole thing.
The useful test is simple. Could these poems be rearranged without loss? If yes, say so plainly. If no, work out what is holding them in place — recurrence, argument, chronology, grief, geography. That answer is usually your review's central claim.
From Notes to Verdict
By now you should have several pages of noticing. The review is a compression of them, not a summary of the plot. Work in this order:
- Write your central claim in one sentence. If you cannot, you have not decided what you think yet.
- Choose two or three poems as evidence. Quote from those, not from the whole book.
- Quote briefly and accurately, from the correct edition, with line breaks preserved.
- Describe before you judge. Say that a poem leans hard on anaphora, then say whether it works.
- Judge what is on the page, not what the poet probably meant. Intention is not available to you.
- Deliver a verdict with reasons. Say who would love this book, and what might put a reader off.
Honesty is not the same as cruelty. A review that risks nothing helps nobody. If the book is uneven, say where it is strong and where it drifts, and be specific about both.
Filing the Review: A Last Look
Before you file, run through this short check. It catches most of the things that make poetry reviews unreadable.
- Does the review tell a reader what the book is like, as well as how good it is?
- Have you said something specific about at least two poems?
- Have you quoted accurately and checked the line breaks survive?
- Have you reviewed the book rather than the poet's reputation or persona?
- Is your verdict consistent with the reasons you have given?
If the answer to all five is yes, you have written something a reader can actually use — and something the poet, however grudgingly, can learn from.
Photo: josealbafotos / Pixabay
