Poetry explication is one of the trickiest assignments to grade because the strongest essays often look very different from one another. One student might focus on sound patterns, another on shifts in tone, and a third on how a single image develops across the poem. A rubric that only rewards one approach will penalize the others unfairly. The goal is a set of criteria flexible enough to recognize good reading in several forms, yet specific enough that scores are consistent.

Keats's 1817 poems are a good testing ground because they contain so many features students can seize on. The sonnets reward attention to structure and turns, the longer poems reward attention to imagery and tone, and the verse epistles reward attention to voice. A rubric that works across all of them is likely to work for most Romantic poetry, which shares an interest in nature, feeling, and the imagination.

The first decision is how to describe the central skill. Explication is not summary and it is not a list of devices; it is an argument about how the poem produces its effects. Writing that definition at the top of the rubric, in plain language, helps students understand what is expected and helps graders return to a shared idea when essays fall between levels.

Four criteria that cover most of what matters

Most explication rubrics can be built from four rows: a central interpretive claim, use of evidence, analysis of technique, and control of writing. Each row should have three or four performance levels described in observable terms. For example, a top-level claim row might say the essay makes an arguable point about the poem's effect and sustains it through every paragraph, while a lower level might say the thesis names a theme without explaining how the poem develops it.

• Interpretive claim: arguable, specific, and sustained across the essay
• Evidence: short quotations chosen for a reason and explained in detail
• Technique: imagery, form, and sound discussed for their effect on meaning
• Coherence: paragraphs organized around ideas, not around line order
• Writing: precise diction and clean mechanics that support the argument

Anchor each level with a Keats example

Descriptors become much clearer when you attach a sample sentence to each level. For a sonnet like "On the Grasshopper and Cricket," a top-level analytical sentence might explain how the poem links two different seasons through a shared sense of warmth. A middle-level sentence would notice that both insects appear and that the poem mentions the seasons, while a lower-level sentence would simply say that the poem is about nature.

These examples do more than illustrate the rubric; they reveal where it is unclear. If you cannot easily write a sample sentence for a level, the descriptor probably needs revising. Testing the rubric on a handful of real student paragraphs before the whole class submits work will show whether the levels are actually distinguishable in practice.

Watch for the common traps in Romantic poetry essays

Romantic poetry attracts a particular set of habits in student writing. Many essays lean on big abstractions like beauty, nature, and emotion without tying them to specific lines. Others assume that the speaker of the poem is the poet and fill the essay with guesses about Keats's feelings that the text does not support. Naming these traps in the rubric lets you comment on them consistently.

It also helps to decide in advance how to treat mechanical errors. In a poetry essay, an occasional comma splice matters less than a misquotation that changes the meaning of a line. Weighting accuracy of evidence above minor mechanics, and saying so on the rubric, aligns your grading with the real goals of the assignment.

Make the rubric work as a feedback tool

A rubric is at its best when students can use it before they submit. Sharing the criteria with a short example essay and asking students to score it themselves builds shared understanding and cuts the number of grade disputes later. Peer review sessions built around the rubric rows also make comments more focused than an open request to "give feedback."

When it comes time to grade, tie your comments to the rubric language so students can connect the feedback to the score. If your workflow includes an AI drafting tool, feed it the same rubric so its suggested comments follow the same standard you would apply. The result is grading that feels transparent, consistent, and easier to defend.