The interface is not live yet. The framework that powers it is fully specified and runs inside the full Ora system today; the standalone browser version is downstream infrastructure and ships when ready. This page describes what the tool does and the methodology it wraps.

What it does

The Propaganda Analyzer treats an input artifact as a piece of operational evidence and asks structural questions of it. Which conceptual metaphors carry the argument, in Lakoff's sense — and where do those metaphors do work the surface claims do not justify? Which Goffman-style framing operations are active — what is being keyed as serious versus playful, sincere versus performative, in-frame versus out-of-frame? Which Entman frame-functions are present — what is the piece defining as the problem, diagnosing as cause, evaluating morally, prescribing as remedy? What is being selected for attention and what is being made salient — and what is being structurally omitted? Where does the artifact deploy a motte-and-bailey: an indefensible bailey claim that retreats to a defensible motte under pressure?

The output is a structured audit rather than a verdict. It names the devices, quotes the textual cues that ground each identification, and explains why the move qualifies. The visitor reads the audit and forms their own judgment about whether the devices are deployed honestly or as engineering.

Why this exists

Existing media-literacy tools focus on fact-checking — they audit the claims and report which are accurate, which are misleading, which are false. That work matters and has its place. The Propaganda Analyzer covers a different dimension. A piece can be entirely truthful at the claim level and still carry an argument through metaphor selection, frame imports, and salience choices the surface claims do not declare. A frame can be deployed honestly or as engineering; the artifact-level structural analysis is what tells you which.

The strategic premise is that persuasion-engineering devices are most effectively counter-acted when readers learn to recognise them on first encounter. Single-shot rebuttal of individual claims is rarely effective at the population level — the device returns in the next artifact, against the next claim, and the reader is unprepared again. Pattern-recognition fluency, by contrast, accumulates. The Analyzer is a pattern-recognition tutorial wrapped around the specific artifact the visitor submitted; learning the device by seeing it named in something they would otherwise have read at face value is the load-bearing pedagogy.

How it works, operationally

The visitor pastes the artifact's text. The tool runs a structural pass — metaphor inventory, framing operations, frame-functions, selection-and-salience analysis, motte-and-bailey check, cross-reference against the named bad-faith techniques catalog. It returns a structured document organised by device type with quoted textual cues and brief explanatory notes. Privacy disclosure and rate limits operate invisibly to legitimate users; the tool does not retain the submitted artifact after the analysis is returned.

The audit is calibrated for op-eds, editorials, advertisements, manifestos, and longer opinion pieces. Very short artifacts (a single tweet, a billboard) do not carry enough structural surface for the full audit; the tool will still return what it finds but flags the limitation.

The underlying framework

The tool is the user-facing wrapper around the Propaganda Analyzer framework — a specialised variant of the broader Argument Audit Analysis the Ora system carries. The framework's analytical vocabulary draws on the propaganda-studies and framing-theory literatures: Lakoff on conceptual metaphor, Goffman on framing, Entman on frame-functions, Bandura's mechanisms of moral disengagement at the cognitive layer, Bernays / Lippmann / Schmitt as the foundational lineage the contemporary operations are deployments of.

The framework distinguishes propaganda from argument-with-frame-imports as a discipline. All argument imports frames; that is what arguments do. Propaganda is the deliberate engineering of frame selection, metaphor choice, and salience structure to move the audience without their noticing the engineering. The line between the two is not always clean, and the framework's output names what it sees rather than rendering a categorical verdict. The reader does the categorisation.

See the framework library for the full collection the Analyzer is drawn from, including the related Argument Audit Analysis (general argumentative coherence) and Argumentative Artifact Examination (structural and rhetorical evaluation of an argument's overall integrity).

What the tool will not do

The Analyzer is not a content moderator and not a verdict engine. It will not tell the visitor whether to trust the piece, whether the piece's underlying position is correct, or whether the piece's author is acting in good faith. Those judgments belong to the visitor. The Analyzer surfaces the structural devices; what to do with the audit is the reader's work.

The Analyzer is also not a fact-checker. Whether a claim in the artifact is empirically true is a different question from whether the artifact carries that claim through metaphor or frame imports. The Analyzer is calibrated for the second question. Pair it with a fact-checking source when both questions matter.

Related

  • Framework Classifier — for routing other kinds of questions to the right analytical framework.
  • The framework library — including Argument Audit Analysis and Argumentative Artifact Examination, which the Propaganda Analyzer specialises from.
  • Main Street Independent — the publication that operates from the same methodology, with the Propaganda Analyzer in production on its editorial work.