The Critical Questions Method — How to Audit Any Argument

Most people, faced with an argument that feels wrong, can only say "something is off." They sense the manipulation but cannot name it. The Critical Questions Method — developed by argumentation theorist Douglas Walton across four decades of work — turns that vague unease into a precise diagnostic procedure. For every recognizable pattern of reasoning, there is a small, fixed set of questions that, when asked aloud, exposes exactly where the argument holds and where it breaks. This article explains how the method works, why it succeeds where pure logic fails, and how to apply it to the arguments you encounter every day.

This is the fourth deep-dive in TellDear's coverage of D5: Argumentation Schemes. Where The Anatomy of Argumentation Schemes, The Hidden Engines of Persuasion, and When Evidence Runs Out mapped the schemes themselves, this article zooms in on the method that makes the whole framework usable. Schemes without critical questions are just labels. Critical questions turn them into tools.

I. Why Logic Alone Is Not Enough

Formal logic asks one question of any argument: does the conclusion follow necessarily from the premises? If yes, the argument is valid. If no, it is invalid. This is a beautifully clean standard — and almost useless in everyday discourse.

The reason is simple: most real arguments are not deductive. A doctor diagnosing a patient, a juror weighing testimony, a manager deciding a budget, a citizen evaluating a politician's claim — none of these reasoners are extracting necessary conclusions from airtight premises. They are working with presumptive reasoning: drawing conclusions that are reasonable to accept provisionally, given the available information, until something better comes along.

Classical logic has nothing to say about presumptive reasoning. By its standards, almost every real argument is "invalid" — yet some are clearly stronger than others. We need a way to distinguish a careful diagnosis from a hasty one, a thorough analogy from a forced one, an honest appeal to authority from a corrupt one. That is what argumentation schemes plus critical questions provide.

Walton's central insight was this: different patterns of reasoning fail in different ways. Each scheme has its own characteristic vulnerabilities, and each set of vulnerabilities can be encoded as a short list of questions. Pass the questions, and the conclusion stands as a working hypothesis. Fail any of them, and the argument loses its presumptive force until the failure is repaired.

II. The Origin: Aristotle, Topics, and the Walton Project

The idea that arguments come in recognizable patterns is ancient. Aristotle's Topics — written in the fourth century BCE — was already a catalogue of common argument forms (topoi, "places"), each with characteristic uses and abuses. Cicero and the Roman rhetoricians extended the catalogue. Medieval logicians refined it further. But the project of treating these patterns as full reasoning templates, complete with their own evaluation procedures, fell out of fashion as formal logic rose to dominance in the twentieth century.

Walton, working with Erik Krabbe, Chris Reed, Fabrizio Macagno, and others, revived and modernized the project beginning in the 1980s. The result was a systematic catalogue of roughly sixty argumentation schemes, each paired with a short list of critical questions tailored to its specific structure. Where formal logic had treated everyday reasoning as a degraded form of proof, Walton treated it as a different but equally legitimate enterprise — one with its own standards, its own failure modes, and its own diagnostic methods.

The result is closer in spirit to medical diagnostics or engineering inspection than to mathematical proof. You do not ask whether the heart is "valid." You run a checklist of tests, each designed to detect a specific kind of failure. Critical questions are the same idea applied to arguments.

III. The Anatomy of a Critical Question

A critical question is not just any objection. It has a precise role in the structure of presumptive reasoning, and it carries a particular kind of weight. Three features distinguish it from ordinary skepticism.

It is scheme-specific. Each question targets a specific weak point in a specific reasoning pattern. The question "Is the expert actually qualified in the relevant field?" makes sense only when the argument in front of you is an Argument from Expert Opinion. Asking it of an analogy or a sign argument would be a category error. The critical questions are not generic doubts; they are surgical instruments matched to the structure being examined.

It shifts the burden of proof. This is the most important property and the one most often missed. Once a critical question has been raised, the original arguer must answer it, or their conclusion loses its presumptive standing. The dialectical balance shifts. The arguer no longer needs to prove the conclusion in any deductive sense — they only need to address the question well enough that a reasonable interlocutor can let it pass. But until that response comes, the argument is on hold.

It is finite. Each scheme has only a handful of critical questions — typically three to seven. This is what makes the method practical. You are not trying to think of every possible objection; you are running through a fixed checklist that competent reasoners have agreed captures the most common failures of that scheme. The list is closed enough to memorize and apply quickly, and open enough to catch the failures that matter.

IV. The Seven Patterns of Critical Question

Although every scheme has its own questions, certain question types recur across many schemes. Recognizing these recurring patterns helps you internalize the method without memorizing sixty separate checklists.

1. The qualification question. Does the source actually possess what the scheme assumes? An expert must have expertise in the relevant field. A witness must have been in a position to know. A statistic must come from an appropriate sample. The qualification question challenges the implicit credentials behind the argument. It is the first line of defense against arguments built on borrowed authority.

2. The relevance question. Does the supporting material actually bear on the conclusion? An analogy may be vivid but irrelevant. An expert may be qualified but speaking outside their domain. A precedent may exist but not apply. Relevance questions cut through arguments that pile up impressive-sounding support without ever connecting it to the point in dispute. They share territory with Red Herring from D1 and Whataboutism from D6.

3. The consistency question. Does the argument fit with what else is known? Does the expert agree with other experts? Does the analogy hold across additional cases? Does the witness account match independent evidence? Consistency questions look outward, comparing the argument's claims against the surrounding body of knowledge. Inconsistency does not refute an argument outright, but it raises the bar that must be cleared.

4. The exhaustiveness question. Have alternatives been considered? Is this the only explanation for the sign? Are there other analogies pointing the other way? Could the consequences play out differently? Exhaustiveness questions guard against the very common move of treating a single plausible story as if it were the only possible one. They connect to False Dilemma and to Inference to the Only Explanation in D4.

5. The bias question. Does the source have an interest in the conclusion? Was the expert paid by a stakeholder? Was the witness threatened? Was the data collected by someone who needed a particular result? Bias questions do not assume dishonesty — they ask whether the structure of incentives could distort the reasoning, regardless of intent. They overlap with Conflict of Interest arguments.

6. The defeater question. Is there evidence that would specifically defeat this argument? Defeaters are not just any contrary evidence; they are facts that, if true, would undermine the scheme itself. For an analogy, a defeater is a relevant disanalogy. For an authority, a defeater is a documented case where the same authority was decisively wrong. Defeater questions force the arguer to acknowledge what would change their mind.

7. The acceptability question. Are the premises themselves acceptable to the audience? Would a reasonable person, having heard the argument, agree to the starting points? This is the most pragmatic of the questions, and the most easily overlooked by those trained in formal logic. An argument from premises no one believes is rhetorically dead, no matter how elegant its structure.

These seven types are not a rigid taxonomy — Walton himself never enumerated them this way. They are a heuristic for noticing that the critical questions for a new scheme will usually fall into familiar shapes. Once you have these patterns in mind, you can often generate the critical questions for an unfamiliar scheme on the fly.

V. Critical Questions in Practice — A Worked Example

Suppose a public health official says: "Studies show that this new screening test catches ninety-five percent of cases. We should make it mandatory for everyone over fifty." This looks like a confident, evidence-based recommendation. Apply the critical questions and the picture changes.

The argument combines at least two schemes — an Argument from Expert Opinion (the studies) and an Argument from Consequences (mandatory screening will save lives). Each scheme brings its own questions.

Qualification: Were the studies conducted by qualified researchers, peer-reviewed, and replicated? Or are we relying on a single industry-funded paper?

Relevance: Does the ninety-five percent figure refer to sensitivity (catching true cases) or to overall accuracy? The two are easily confused, and the difference matters enormously when the disease is rare. This is the territory of the Base Rate Fallacy from D4.

Consistency: Do other studies of the same test reach similar numbers? Are there populations where the test performs much worse?

Exhaustiveness: Have alternative interventions been considered — targeted screening, lifestyle programs, earlier symptom education? Is universal screening the only option, or merely the most visible one?

Bias: Who manufactures the test? Did they fund the studies? Will mandating it create a captive market?

Defeater: What about false positives? In a population of low-prevalence cases, a ninety-five percent sensitive test will produce vastly more false alarms than true detections, leading to unnecessary biopsies, anxiety, and downstream harm. This single defeater can flip the argument's conclusion from "obviously good" to "almost certainly net negative."

Acceptability: Would the affected population, fully informed, accept the trade-off being made on their behalf?

Notice what just happened. We did not need to be epidemiologists to dismantle this argument. We did not need to know whether the screening test is actually effective. We only needed to know which questions to ask. The argument may well survive — perhaps every question has a satisfying answer — but it cannot survive in the form originally presented. The burden of proof has shifted.

VI. The Method as Defense Against Manipulation

Most manipulation works not by lying but by presenting partial reasoning as complete reasoning. The manipulator gives you the premises, the scheme, and the conclusion, and trusts that you will not pause to ask the critical questions. The faster the conversation moves, the less likely any individual question is to surface. This is why Gish Gallop from D6 is so effective: by piling argument upon argument, the manipulator outruns the audience's ability to question any single one.

The Critical Questions Method is precisely the defense against this kind of pace-driven manipulation. It does not require you to have a counter-argument. It does not require you to know the truth. It only requires you to say: before I accept this, here is the question that has not been answered. Once you have asked, the burden is no longer yours.

This shifts the asymmetry that manipulators exploit. Normally, the person making a claim has the easy job — they only need to assert — while the person doubting has the hard job of disproving. Critical questions invert this. The doubter only needs to ask; the asserter has the work of answering. This is structurally similar to the legal principle that the prosecution bears the burden of proof, and for the same reason: it protects the audience from being talked into things by sheer volume.

In the modern attention economy, where Firehose of Falsehood tactics from D2 produce more claims than any individual could ever investigate, critical questions are not optional. They are the only scalable defense. You cannot fact-check everything. You can ask the right question about anything.

VII. The Discipline: When to Ask, When to Act

A common misuse of the Critical Questions Method is to treat it as a tool for endless deferral. Every conclusion can be questioned. Every answer raises new questions. If you take this seriously, you will never decide anything — and indecision is itself a decision, often the worst one available.

Walton was clear on this point. Critical questions are not a license for permanent skepticism. They are a tool for calibrated provisional belief. The point is not to keep asking questions until certainty arrives — certainty rarely arrives — but to ask enough questions that the remaining uncertainty is acknowledged, manageable, and roughly proportional to the stakes of the decision.

For low-stakes decisions, run through the critical questions quickly and act. For high-stakes decisions, run through them carefully and demand satisfying answers. For irreversible decisions — those where you cannot easily undo the consequences if the argument turns out wrong — demand answers to every question and treat unanswered ones as veto-strength objections.

The discipline is to match the depth of inquiry to the cost of error. This connects to Asymmetric Risk reasoning from D4. Not every argument deserves the same level of scrutiny, but every argument deserves some, and the most consequential arguments deserve the most.

VIII. Connecting to the Broader Landscape

The Critical Questions Method does not exist in isolation from the rest of the dimensions TellDear maps. It is, in a sense, the application layer through which all other forms of error become detectable.

D1 (Logical Fallacies): Many fallacies are exactly what happens when an argumentation scheme is deployed without its critical questions. Appeal to Authority becomes fallacious when the qualification and bias questions are not asked. Hasty Generalization is what an argument from example looks like without the exhaustiveness question. The scheme/fallacy distinction is, in Walton's framework, almost entirely a matter of whether critical questions have been engaged.

D2 (Manipulation and Propaganda): Propaganda is, structurally, the systematic suppression of critical questions. Loaded Language embeds conclusions in premises so that questioning the premises feels rude. Emotional Flooding raises the social cost of stopping to ask questions. Manufactured Consensus creates the illusion that the questions have already been answered by everyone else.

D3 (Cognitive Biases): Biases are the internal reasons we fail to ask critical questions even when no one is stopping us. Confirmation Bias makes the questions feel unnecessary when the conclusion already pleases us. Cognitive Ease makes the asking feel like effort we would rather avoid. Illusion of Explanatory Depth makes us feel we already know the answers without checking.

D4 (Statistical Errors): Many of the critical questions for empirical arguments are statistical in form. The base rate question, the sample question, the measurement question — these are the critical questions that quantitative claims must pass. Statistical literacy is, from this angle, just the ability to ask the right critical questions of numerical arguments.

D6 (Discourse Mechanics): The dialectical machinery of Burden Shifting, Goalpost Moving, and the sabotage tactics covered elsewhere are all techniques for preventing critical questions from doing their job — by changing the subject before the question can be answered, by redefining terms after the fact, or by treating the asking itself as bad faith.

Across every dimension, the same pattern repeats: errors of reasoning are, almost without exception, errors of unasked questions. The Critical Questions Method names them so they cannot stay invisible.

IX. Building the Habit

The method is simple to describe and surprisingly hard to internalize. The obstacle is not intellectual but social. Asking critical questions in real conversations is slow, mildly confrontational, and exposes you to the charge of being pedantic, contrarian, or a "concern troll." Most people, most of the time, would rather let an argument pass than risk these costs. Manipulators rely on exactly this reluctance.

Three practices help build the habit without becoming insufferable. First, ask the questions silently first. Before deciding whether to raise a question aloud, run through the relevant critical questions in your head. Often the answer is obvious and the question can stay private. The discipline of asking — even silently — is most of the value.

Second, frame critical questions as collaborative. "Help me understand — how does this fit with…" is structurally identical to a hostile cross-examination, but socially completely different. The same question asked in good faith and asked in bad faith produces wildly different conversations.

Third, practice on arguments you already accept. The hardest critical questions to ask are the ones aimed at conclusions you want to be true. If you can run the full method on an argument you agree with, and find that some questions remain unanswered even there, you have begun the discipline that makes the method work.

X. Conclusion: The Audit Habit

The Critical Questions Method is, at its core, a checklist culture for arguments. Aviation made its catastrophic accident rate plummet by adopting checklists; medicine made surgical error rates plummet by adopting checklists; the same can happen to everyday reasoning if the appropriate checklists become habitual. The schemes provide the patterns. The critical questions provide the checklists. What remains is the willingness to use them.

This is not a method for becoming smarter. It is a method for becoming more honest — first with yourself, then with others. The critical questions force the arguer (whether someone else or your own past self) to do the work that should have been done before the conclusion was offered. They are the discipline that converts opinion into reasoned belief, and reasoned belief into the kind of conclusion you can act on without later regret.

With this article, TellDear's coverage of D5: Argumentation Schemes moves beyond cataloguing the schemes themselves to the methodology that makes them usable. Together with The Anatomy of Argumentation Schemes, The Hidden Engines of Persuasion, and When Evidence Runs Out, this article completes the picture: the schemes give us the patterns, and the critical questions give us the audit. The combination is the most practical thing critical thinking has produced in the last fifty years — and the most underused.