How to Write a Rhetorical Analysis: Appeals, Evidence and the Claim That Scores

What a rhetorical analysis actually claims

The essay's claim is never 'the text is persuasive' or 'I agree with the author.' It is a statement about strategy: 'Kennedy's concession-to-commitment structure lets the program sound prudent while committing the country to an expensive one' — a claim about how the text works on its audience. That claim answers the 'how does it persuade' question with specifics, names the audience, and previews the two or three rhetorical moves the body will examine. A thesis that merely identifies the topic or the appeal types scores no better than a summary with a colon in it.

Audience is the load-bearing element students skip. The same ad reads as sincere to one audience and manipulative to another, so every effect claim needs its audience named — voters watching a crisis speech, parents scrolling a fundraising email, commuters hearing a radio spot. Course prompts usually supply the audience; when they do not, choose the most plausible one and state it in the introduction, because 'the audience' as an abstraction cannot anchor any judgment about effect. Effect plus audience plus named moves is the complete thesis.

The numbers at a glance

Rhetorical moves to name, with the evidence each requires — the range, and where each row actually shows up in coursework. Skim it first; the sections below explain why the range exists and which row applies to you.

The evidence rule: the artifact is your source

A rhetorical analysis runs on quotation at close range — the two-word phrase, the parallel structure, the sentence that pivots — and the artifact itself is the primary source, cited in the style the course assigns. MLA 9 dominates English courses; APA 7 dominates communications ones, where a speech is cited as a lecture or media object and an advertisement as the campaign material it is. The discipline that separates analysis from impression is quoting the exact language and then saying what it does: the quote is the datum, the next sentence is the interpretation.

Secondary sources are a course-by-course variable. Most introductory versions want none — the essay is you, the text, and the audience — while upper-level rhetoric courses expect two or three scholarly sources to frame the situation, situate the artifact in its moment, or support claims about audience reception. Check the prompt before adding criticism, because importing a scholar's reading of the text can crowd out your own analysis, which is the thing being graded. When secondary sources are required, they support the context and the framework, and your claim stays yours.

The structure graders recognize

The recognizable shape runs introduction, context paragraph, two to four body sections organized by rhetorical move, and a conclusion that judges overall effectiveness. The introduction supplies the artifact's basics — author, venue, date, audience — and ends in the strategy claim. Body sections each examine one move (the appeal to fear, the data-heavy middle section, the closing appeal to shared identity), quote the language that executes it, and explain the effect on the named audience. Organizing by move beats organizing by chronology, which drifts into summary by paragraph three.

Paragraph economics keep the essay in band. At 750-1,200 words, the budget runs roughly 100-150 words of introduction, 100-150 of context if the prompt wants it, 150-250 per body section, and 100-150 of conclusion. Each body paragraph quotes at least once — an unquoted paragraph in a rhetorical analysis is assertion — and each quotation earns its place by generating the sentence that follows it. The conclusion answers the so-what: did the strategy work for this audience, at what cost, and what does that reveal about the genre the artifact lives in.

How the essay loses points

Four failures account for most deductions. Summary instead of analysis: two paragraphs retelling the text before any claim about how it works. The label fallacy: 'the author uses pathos' — a category name is not an analysis; the sentence must say what the emotional appeal does and how. Agreement disguised as analysis: judging whether the author is right about the topic instead of whether the strategy fits the audience. And audience amnesia: effect claims with no audience attached, which cannot be wrong and therefore cannot be scored.

Quotation discipline is the fifth. Dropped quotes — a sentence of the text pasted in with no comment — spend word count without earning it; every quotation needs the interpretive sentence it exists to support. So does quote inflation, where 40% of the essay is the artifact speaking. The 15-25% guideline that applies to research essays applies here too, and the fix is mechanical: after drafting, read each paragraph and make sure the quote is the evidence while your claim about its effect is the sentence doing the arguing.

Analyzing non-print artifacts

Ads, speeches, music videos and social campaigns follow the same essay logic with adjusted evidence. Print and video ads supply visual choices as data — color, framing, casting, typography, the placement of the product relative to the person — and each needs the same quote-and-interpret treatment, with panel numbers or timecodes as locators: 'at 0:12, the cut from the empty crib to the car seat.' Speeches add delivery to the evidence base: pause, repetition, the shift in register that marks the pivot from problem to commitment.

Transcription and citation solve the retrieval problem the artifact creates. Quote a speech from a published transcript when one exists and cite it as such; otherwise describe the moment precisely and cite the recording with the timecode. Ads and campaign material cite the campaign, outlet and date, and a screenshot appendix earns its keep in courses that allow one. The analysis never changes: visual and aural choices are rhetorical choices, and the essay that names the choice, quotes or locates it, and states its effect on the audience is doing the same work a print analysis does.

A five-minute audit of your own source list

Before submitting, run the one-sentence test on every reference: can you state, without opening the PDF, what this source argues and where it supports your paper? Sources that fail the test are collected, not used — cut them or promote them by actually integrating their evidence. Count what remains; that number, not the folder in your reference manager, is what satisfies the requirement.

Two composition checks finish the audit. First, no more than a third of your list should be non-scholarly — news pages, organizational posts, encyclopedias. Second, your three or four load-bearing sources should appear in the text two or more times each; if every source appears exactly once, your paper is a relay of summaries, and the fastest fix is deepening the analysis around your best two sources. The citation generator below formats whatever survives the audit in APA, MLA and Chicago.