A teacher's guide to literary analysis: the claim, evidence, commentary paragraph engine, a SOAPStone / TPCASTT / rhetorical triangle crosswalk for pre-writing, and a full worked example on a Frost poem. Includes a free AI literary analysis generator.
Literary analysis is where high school English tends to hit the wall. Everyone agrees students should write it. Almost nobody agrees on what a good one looks like, and the rubrics teachers inherit often list six criteria that sound identical to a fifteen year old: "cite evidence," "develop analysis," "explain significance," "connect to theme." A student who can write a solid summary can spend a semester producing paragraphs that check all six boxes and still land in the C range, with feedback that says "go deeper" and no clear instruction on how.
This post is the working teacher's version of the two things that reliably move student writing from summary to argument: a paragraph engine (claim, evidence, commentary) that gives every body paragraph a shape, and a pre-writing crosswalk (SOAPStone, TPCASTT, and the rhetorical triangle) that helps students discover something worth arguing before they draft. A full worked example on a Robert Frost poem sits at the end so you can steal it for tomorrow's lesson.
Generate literary analysis prompts on any text — Paste a passage or upload a PDF. AssessmentWiz produces a set of guided prompts (SOAPStone or TPCASTT walkthrough, a menu of interpretive questions, and thesis starters) plus a rubric aligned to CCSS ELA anchors. Free, no signup to try.
A literary analysis is a defensible interpretation of a text, backed by evidence from the text itself. Three words carry the definition. Defensible means someone could reasonably disagree; a thesis nobody would contest ("Macbeth is ambitious") is not analysis, it is observation. Interpretation means the writer is making a claim about what the passage does or means, not what happens in it. Textual evidence means the argument stands on quoted words and phrases, not plot recap or personal reaction.
| Literary analysis is | Literary analysis is not |
|---|---|
| A defensible claim about what the text does or means | A summary of what happens in the passage |
| Built on quoted words and phrases the writer explains | Built on plot points paraphrased in the writer's voice |
| Arguable (a reasonable reader could disagree) | A restatement of the obvious ("the poem uses imagery") |
| Focused on one interpretive question with a clear answer | A tour through everything interesting in the text |
| Commentary that explains the "so what" of each quote | A quote sandwich with no explanation after the quote |
| A response to the text | A response to the writer's own life experience |
The single most useful diagnostic on a student draft is to ask: could I write the opposite thesis and defend it from the same text? If the answer is no, the thesis is too safe. If the answer is yes and the student's draft is still convincing, they have written analysis.
Every body paragraph of a literary analysis is a mini argument with the same three moves: a claim that advances the thesis, evidence that grounds the claim in the text, and commentary that explains how the evidence proves the claim. Different textbooks call this different names (CEC, PIE, TEA, PEEL, MEAL). The names change; the moves do not.
| Move | What it does | Sample sentence | Common failure mode |
|---|---|---|---|
| Claim (topic sentence) | States a mini argument that supports the thesis | "Frost frames the two choices as equivalent, undercutting the poem's later insistence that one was different." | The claim is plot summary or a fact ("The speaker faces two roads") |
| Evidence (quotation) | Grounds the claim in the text with an exact phrase | "The speaker admits the second road was 'just as fair' and that both 'had worn them really about the same.'" | The quote is a long block with no lead in, or the student paraphrases instead of quoting |
| Commentary (analysis) | Explains how the specific words prove the claim | "The intensifier 'just as fair' and the phrase 'about the same' insist on equivalence. The word 'really' is the tell: the speaker is arguing with himself about whether the roads were the same, which suggests he already knows they were not." | The commentary restates the quote ("This shows the roads were the same") instead of interpreting it |
Two rules make CEC paragraphs work in practice. First, commentary should be at least twice as long as the evidence. If a student quotes eight words and comments on them in six, they are still summarizing. Second, commentary should point at specific words inside the quote, not at the quote as a whole. "The word 'really' is the tell" is commentary. "This shows the speaker was conflicted" is paraphrase with a fancier verb.
A strong CEC paragraph often runs two evidence and commentary cycles inside one topic sentence. Claim, evidence one, commentary one, evidence two, commentary two, closing sentence that returns to the claim. The two evidence pieces let the writer triangulate a subtler argument than one quote can carry alone.
Students cannot write a defensible thesis about a text they have not yet noticed anything about. Pre-writing frameworks are structured noticing routines that surface the patterns a thesis can then organize. Three of them earn their spot on a classroom wall. Each fits a different kind of text.
| Framework | What each letter is | Best for | Weakness |
|---|---|---|---|
| SOAPStone | Speaker, Occasion, Audience, Purpose, Subject, Tone | Rhetoric: speeches, essays, letters, opinion writing, nonfiction | Under-serves imagery-heavy poetry and fiction; there is nowhere to note figurative language |
| TPCASTT | Title, Paraphrase, Connotation, Attitude, Shifts, Title (revisit), Theme | Poetry and lyric-driven prose | The two Title steps can feel redundant; students often skip the second one and lose the strongest thesis prompt |
| Rhetorical triangle | Ethos (credibility), Pathos (emotion), Logos (logic) | Argumentative writing: op-eds, speeches, ad copy, court opinions | Only three lenses; misses tone, structure, and audience choices that matter for full analysis |
The crosswalk between SOAPStone and TPCASTT is closer than it looks. Both begin with orientation (Speaker/Occasion, Title/Paraphrase). Both push toward tone (Tone, Attitude). Both end at meaning (Purpose/Subject, Theme). The difference is where they linger: SOAPStone lingers on the rhetorical situation, TPCASTT lingers on the words themselves.
| Function | SOAPStone step | TPCASTT step |
|---|---|---|
| Situate the text | Speaker, Occasion, Audience | Title, Paraphrase |
| Investigate the words | Subject | Connotation |
| Name the tone | Tone | Attitude |
| Track how the text moves | (not covered directly) | Shifts |
| Land on meaning | Purpose | Title (revisit), Theme |
For a hybrid text (a Baldwin essay that reads like a personal narrative, an August Wilson monologue, a Toni Morrison paragraph) use both. Start with SOAPStone to lock in the rhetorical situation, then move through Connotation and Shifts from TPCASTT to catch the craft moves SOAPStone would miss. The rhetorical triangle is the fastest lens for a debate-style unit or an SAT or AP argument prompt.
The path from "we read the poem" to "I have a thesis I can defend" is the step most students skip. It fits inside a single 45 to 55 minute block if the passage is short.
A defensible thesis usually names three things: what the writer does, where you see it, and what it means. "In the last stanza of 'The Road Not Taken,' Frost stages a moment of self-deception in which the speaker rewrites a coin-flip choice into a heroic one, revealing more about the human need to explain our lives than about the roads themselves." That thesis names the move (self-deception), the location (last stanza), and the meaning (the human need to explain our lives).
Robert Frost, "The Road Not Taken" (1916). Public domain, four stanzas, fits on a half page. Below is a compressed TPCASTT, an interpretive question, a thesis, and a single body paragraph following the CEC engine.
| TPCASTT step | Notes |
|---|---|
| Title (before reading) | Suggests regret; the road NOT taken implies the speaker chose one and is haunted by the other. |
| Paraphrase | A speaker faces a fork, picks one road, admits both looked the same, and predicts that later he will claim the choice mattered. |
| Connotation | "Yellow wood," "fair," "just as fair," "wanted wear," "really about the same," "sigh," "difference." The color and season words feel autumnal and melancholy; the equivalence words insist the roads were identical. |
| Attitude | Ambivalent, then rueful. The speaker is not confident he chose better; he is rehearsing the story he will tell. |
| Shifts | Time shift in the last stanza ("I shall be telling this with a sigh / Somewhere ages and ages hence"). Tone shifts from reflective to performative. |
| Title (after) | The title lies. The poem is not about a road not taken; it is about the story the speaker will invent about a road that was basically the same as the other. |
| Theme | The human tendency to retrofit meaning onto arbitrary choices. |
Interpretive question: Why does Frost keep insisting the roads were equivalent when the poem is titled "The Road Not Taken" and ends by claiming one road made all the difference?
Thesis: In "The Road Not Taken," Frost stages the gap between a choice as it happens and a choice as it is remembered, using the poem's repeated equivalence claims to expose the speaker's later "difference" as a story he tells himself rather than a truth the roads support.
Sample CEC paragraph:
Frost frames the two roads as functionally identical, and the poem's obsession with proving that equivalence is what makes the final stanza's about-face so telling. The speaker admits the second road was "just as fair" as the first, then doubles down: the two "had worn them really about the same," and "both that morning equally lay / In leaves no step had trodden black." The intensifier "just as," the qualifier "really," and the adverb "equally" pile on top of each other, as if the speaker is arguing with himself. The word "really" is the tell. Nobody uses "really" to describe something obviously true; you use it to insist on something that could be doubted. Frost has the speaker insist, three times over, that the roads were the same, so that when the last stanza flips ("I took the one less traveled by, / And that has made all the difference") the reader sees the sleight of hand. The poem is not about the road; it is about the sentence the speaker is preparing to say "with a sigh / Somewhere ages and ages hence." The equivalence claims set up the last stanza to land as self-deception rather than revelation, which is why the poem's most quoted lines are also its least trustworthy.
That paragraph runs two evidence and commentary cycles under one claim. Every quote is short (four to nine words), lead-ins carry meaning, and the commentary points at specific words ("really," "just as," "equally") rather than the quote as a whole. Total length: about 200 words for one body paragraph, which is the right density for a five-paragraph literary analysis at the high school level.
CEC is a three-part paragraph shape used in literary analysis. The Claim is a topic sentence that advances the thesis with a mini argument. The Evidence is a short quoted phrase from the text. The Commentary is at least two sentences explaining how the specific words in the quote prove the claim. Different textbooks call the same pattern PIE, TEA, PEEL, or MEAL; the moves are the same. The rule of thumb is that commentary should be at least twice as long as the evidence and should point at specific words inside the quote, not at the quote as a whole.
Use SOAPStone (Speaker, Occasion, Audience, Purpose, Subject, Tone) for rhetorical texts: speeches, essays, letters, opinion writing, and nonfiction. Use TPCASTT (Title, Paraphrase, Connotation, Attitude, Shifts, Title, Theme) for poetry and lyric-driven prose. For hybrid texts like a Baldwin essay that reads like a memoir, start with SOAPStone for the rhetorical situation and move to TPCASTT for the craft moves. Neither framework is the goal; both are noticing routines that make patterns visible so students have something worth arguing about.
A defensible thesis is arguable, specific, and answerable from the text. Arguable means a reasonable reader could disagree; if nobody would contest the claim ("Macbeth is ambitious"), it is not a thesis. Specific means it names what the writer does, where you see it, and what it means, not just a topic ("Frost uses imagery"). Answerable from the text means the argument can be won or lost with textual evidence, not personal reaction or biographical trivia. A quick diagnostic: could you write the opposite thesis and defend it from the same passage? If yes and your original still holds, you have a strong thesis.
For a standard high school five-paragraph literary analysis, body paragraphs typically run 150 to 250 words. That length gives room for one topic sentence (claim), two evidence and commentary cycles, and a closing sentence that returns to the thesis. Shorter than 150 usually means the commentary is missing or thin; longer than 250 often means the writer is quoting too much or drifting into a second claim that belongs in its own paragraph. Length is a symptom, not a target; if the commentary is doing real work, the paragraph will land in this range on its own.
A book report summarizes what happens in the text and often evaluates whether the writer liked it. A personal response describes how the text made the reader feel or what it reminded them of. Literary analysis does neither. It makes a defensible claim about what the text is doing or means and supports that claim with quoted words the writer explains. The writer's opinion appears only as the argument itself; the writer's life, tastes, and reactions stay off the page.