Episode 5: Flaw Questions and the Flaw Taxonomy

Adrian Calloway

Here's an argument. Towns with more firefighters suffer more fire damage. Therefore the firefighters are causing the damage. The question asks what's wrong with that reasoning, and one answer choice reads, the argument relies on a sample too small to support a general conclusion.

Nora Ashford

And that sounds like a real flaw. Small sample, that's a thing people get dinged for.

Adrian Calloway

It is a real flaw. It's also dead wrong here, and the test-writer is betting you'll grab it anyway. Sit with that, because the whole episode lives in why a true-sounding answer can still be the wrong answer.

Nora Ashford

Before we go further, quick retrieval, all the way back to where we started. In Episode 1 you said the LSAT lives in the gap. In one sentence, what was the gap?

Adrian Calloway

The gap is the unstated assumption. It's the space between what the premises actually prove and what the conclusion claims. Premises down here, conclusion up there, bridge missing.

Nora Ashford

So a Flaw question that asks what's wrong with the argument is just asking me to name that gap.

Adrian Calloway

That's the whole payoff of Episode 1. A Flaw question is the gap, wearing a costume. And here's what makes it learnable. The gap only shows up in a handful of recurring shapes. Today's objective, one skill. By the end you'll hear an abstract answer choice, something like takes for granted that a correlation establishes causation, and match it to the concrete argument in under fifteen seconds.

Nora Ashford

So the skill isn't memorizing flaws, it's recognizing the shape of the gap from the answer-choice wording.

Adrian Calloway

Right. Three moves today. Move one, the four-step method. Move two, the two-part test that kills wrong answers. Move three, the flaw taxonomy, the named gaps, taught in clusters so they survive in your ears. Method first, vocabulary second. The names are useless without the method.

Adrian Calloway

Picture this. You're handed a stimulus, and the stem says, the reasoning is most vulnerable to criticism on the grounds that. Vulnerable to criticism, flawed because, questionable because, commits which error in reasoning. Those stems all mean one thing. Find the broken reasoning, accepting the facts as given.

Nora Ashford

Step one is just confirming from the stem that it's a Flaw question.

Adrian Calloway

Confirm the stem. Step two, find the conclusion and the premises, then say out loud, in your own words, why the premises don't prove the conclusion. That's the gap from Episode 1. Step three, prephrase the flaw before you read a single answer. Step four, match your prephrase to the abstract answer.

Nora Ashford

Wait, step two has a constraint I keep forgetting. You said accept the premises as true.

Adrian Calloway

That's the anchor for this entire question type, and it reverses most people's instinct. You think your job is to find what's false. It isn't. You never doubt the facts. If the stimulus says firefighter towns have more damage, you swallow it whole and attack only the inference, the jump to the conclusion. So any answer that says, but that premise might not be true, is automatically wrong on a Flaw question. It's solving a problem the question didn't ask.

Nora Ashford

So the prephrase in step three is me predicting the flaw before the answers get a chance to seduce me.

Adrian Calloway

Here's the key move, and it's the test-writer's intent laid bare. The answer choices are engineered to be tempting. If you walk in with your own words already formed, the temptation has nothing to grab onto. Students who skip the prephrase get hypnotized by smart-sounding language. The prephrase is your immunity.

Adrian Calloway

Now the single most important tool in the episode. Every correct flaw answer has to clear two questions. One, did the author actually do the thing the answer describes? Two, is that thing actually a problem in the reasoning? Both, not one. A wrong answer fails at least one.

Nora Ashford

Let me say it back in my own words. A wrong answer can name a real, famous fallacy that this particular argument simply never committed.

Adrian Calloway

That's the trap, and here's the cruel part. The better you know your fallacy names, the more attractive a false one becomes. Knowledge works against you unless you test it. Let's run the cold open through it. Firefighters cause damage, and the answer was, relies on too small a sample.

Nora Ashford

Part one, did the author do that? Did they generalize from a tiny sample? No. They never sampled anything. They saw two things move together and called it cause and effect.

Adrian Calloway

Fails part one. The author never committed that error, so the answer is dead, no matter how true sample bias is in the abstract. Now pause and try this one before I answer. What is the real flaw in the firefighter argument? Take a beat and prephrase it.

Nora Ashford

It's correlation mistaken for causation. The size of the fire is the third thing driving both. A bigger blaze summons more trucks and burns more building.

Adrian Calloway

That's the prephrase. And there's a second alternative the author ignored, reverse causation. Maybe heavy damage is what makes a town hire more firefighters in the first place. The correct answer will read something like, takes for granted that because two phenomena are correlated, one must have caused the other. Run it. Part one, the author did exactly that. Part two, it's a genuine problem. Both pass.

Adrian Calloway

Now the vocabulary. And a warning. A flat list of a dozen names won't survive in your ears, so we cluster. Five clusters. One, conditional. Two, causal. Three, source and relevance. Four, language. Five, quantity. One more thing. This taxonomy isn't an official LSAC list, it's a naming system the prep world built. So labels vary, and the names matter less than spotting the shape.

Nora Ashford

Start with the most common one.

Adrian Calloway

Conditional cluster, and this is the single most common flaw on the test. Sufficient and necessary confusion. Picture this. To get into the honors program, you need a three-point-eight. Maya has a three-point-eight, so she's getting in.

Nora Ashford

Feels right for a second. But a three-point-eight is required, it isn't enough by itself. There could be an essay, an interview, a cap on spots.

Adrian Calloway

That's the tell. Treating a necessary condition as if it were sufficient. The answer sound is, treats a condition that is required for an outcome as one that guarantees it. Now the causal cluster, which you already met. The tell, two things move together and the author claims one caused the other. The answer sound, takes for granted that a correlation establishes a causal relationship. Lead with these two clusters, conditional and causal. They're the two most-tested and the two best disguised, because their abstract wording is long and dry.

Nora Ashford

And that's exactly why I skim past them. The right answer is buried in a boring sentence, so my eye slides off it.

Adrian Calloway

So slow down on the boring ones. Cluster three, source and relevance. Ad hominem, attack the person instead of the claim. Answer sound, rejects a position by questioning the motives of those who hold it. Appeal to authority, but read this carefully. It's only a flaw when the expertise is irrelevant, a physicist settling a cancer-biology question, or when an expert's say-so is treated as proof. A qualified, relevant expert is not a flaw.

Nora Ashford

That one used to burn me. I flagged every expert citation as appeal to authority, like the word expert was a trap by itself.

Adrian Calloway

Relevance is the hinge, not the citation. Same cluster, appeal to popularity, lots of people believe it so it must be true. And straw man, distort your opponent's position and then knock down the weaker version. It needs real misrepresentation, not mere disagreement, and it loves two-speaker stimuli. Cluster four, language. Equivocation, a word shifts meaning mid-argument. Circular reasoning, a premise just restates the conclusion.

Nora Ashford

And on circular reasoning, you flagged that begs the question on the LSAT isn't the everyday meaning.

Adrian Calloway

Critical, because everyday usage will mislead you. On the test, begs the question means assumes the conclusion. Not raises a question. If an answer says, presupposes the truth of what it sets out to prove, that's circular. Cluster five, quantity. Percentage versus raw number. Part-to-whole and whole-to-part, also called composition and division. Unrepresentative sample or hasty generalization. And false dilemma, treating two options as the only options when more exist.

Nora Ashford

Five clusters. Conditional, causal, source, language, quantity. And the famous-fallacy names are the bait whenever I skip the two-part test.

Adrian Calloway

Two contrasting examples now, so you abstract the move they share. First, fully worked. A feather is light. Anything light cannot be dark. So a feather cannot be dark-colored.

Nora Ashford

Something's slippery in there. Light means low weight in the first sentence. Then light means pale in the second. One word, two jobs.

Adrian Calloway

That's equivocation. Run the two-part test. Part one, did the author shift a term's meaning? Yes, light slides from weight to color. Part two, is that a problem? Yes, the premises stop connecting the instant the word changes jobs. The answer sound, relies on the term light being used in two different senses.

Nora Ashford

Say why that step works. The chain only holds if light means one thing the whole way through. The minute it changes, the two premises aren't even talking about the same property, so the bridge collapses.

Adrian Calloway

That self-explanation is what locks it in. Now your turn, faded. Second example, and notice it's a different cluster, so you can't just pattern-match the last one. A larger percentage of our small Tuesday class earned A's than our huge Monday class. So more students got A's on Tuesday. Before you solve, which cluster?

Nora Ashford

Quantity cluster. Percentage versus raw number. A bigger slice of a small pie can still be fewer actual people than a smaller slice of a giant pie.

Adrian Calloway

Finish the two-part test. Part one.

Nora Ashford

Part one, did the author jump from a percentage to a count? Yes, the premise is a rate, the conclusion is a number of students. Part two, is that a problem? Yes, because Monday is huge, so even a smaller percentage there could be more people. Both parts pass. The answer sound would be, infers a conclusion about an amount from evidence about a proportion.

Adrian Calloway

Real Flaw questions never hand you the cluster. The whole LSAT skill is discriminating between them under pressure. So before you solve this next one, name which flaw it is. I'm mixing in everything so far. Ready? The mayor says we should fund the new library. But the mayor only wants votes, so we should reject the library plan.

Nora Ashford

First instinct, that's circular. The conclusion is hidden inside the premise.

Adrian Calloway

Defend it. Walk me through the circle.

Nora Ashford

Okay, the premise is the mayor wants votes, and the conclusion is reject the plan, and those are not the same claim restated. There's no circle. I just talked myself out of my own answer.

Adrian Calloway

Good, you caught it mid-defense. That confident wrong turn is worth more than a lucky guess, because now the boundary of circular reasoning is burned into you. So what actually happened to the library plan's merits?

Nora Ashford

Nobody touched them. They went after the mayor's motive instead of the plan. That's ad hominem. Source cluster.

Adrian Calloway

Now use the two-part test to kill the decoy. The answer that says, presupposes the very thing it sets out to prove. Does it pass?

Nora Ashford

No. Fails part one. The author never assumed the conclusion inside a premise. The right answer is, rejects a claim merely by questioning the motives of the person who advanced it. Whether the library is actually a good idea just gets left unaddressed.

Adrian Calloway

That's the move. Identify the type first, then solve. And notice the wrong answer was a real, famous flaw, circular reasoning, that this argument never committed. Same trap as the cold open, new costume.

Adrian Calloway

Compress it into one heuristic you can actually use. Prephrase the gap in your own words, then make every answer pass two questions. Did the author do it, and is it a problem. Say it with me. Did they do it, is it a problem.

Nora Ashford

Did they do it, is it a problem. And the famous-fallacy answer the argument never actually committed is the trap, every single time.

Adrian Calloway

Callback to the cold open. The small-sample answer was a true fallacy the firefighter argument never made. It failed part one, so it died, no matter how smart it sounded. That's the whole skill. A flaw answer has to describe what this argument really did, and that thing has to really be broken, while you accept the premises as true the entire time.

Nora Ashford

And the two heavyweights to master first are sufficient-necessary confusion and correlation-to-causation. If I only learn two, those two.

Adrian Calloway

The quotable button. Don't pick the flaw that's famous, pick the flaw that's there. Next episode we go deep on the single most tested pattern on the whole exam, causal reasoning. Today we only named correlation-causation. Episode 6 takes it apart, the third cause, the reversal, the coincidence, and how to both attack it and defend it.

Nora Ashford

And the tiny action before next time?

Adrian Calloway

Between now and Episode 6, do five Flaw questions from your bank. On each one, write your prephrase before you read the answers, then run the two-part test out loud on the choice you pick. Five questions, two questions each. That's it.