Lecture: Arguments and Thought Experiments
Lecture script · about 15–20 minutes spoken · Week 2
Opening
Picture a jury room late in the afternoon. One juror says, "If the defendant was in the building, the camera would show someone in a red coat. The camera shows someone in a red coat. So the defendant was in the building." It sounds tidy. Something about it still bothers you. By the end of this lecture, you'll be able to say exactly what.
That skill travels. A nursing student checks whether a care plan's reasons support its recommendation. A business major reads a manager's case for a new policy. A juror weighs what the evidence shows. Each of them is asking the same question: do these reasons actually get me to this claim?
In Week 1 you met Socrates, who tested people's definitions by finding a case they got wrong. In this lecture we turn that habit into tools. You'll learn to lay out an argument line by line, to judge whether its shape works, and to test ideas with counterexamples and thought experiments. Next week, in Week 3, you'll use these tools on Plato's theory of Forms.
1. Standard form
Here's the plain idea. When people argue, they mix reasons and claims together in one breath. To check an argument, you first pull its parts apart.
An argument is a set of statements, the premises, offered as reasons for another statement, the conclusion. A premise is a reason. The conclusion is the claim the reasons are meant to support.
Standard form gives each part its own line. You number the premises P1, P2, and so on, and each one is a single sentence that can be true or false. The conclusion comes last. You mark it C, and you start it with the word "So". If you add a premise the speaker left unsaid, you mark it "added", so nobody mistakes your words for theirs.
Let's do the juror. P1: if the defendant was in the building, the camera would show someone in a red coat. P2: the camera shows someone in a red coat. C: so the defendant was in the building. Now each line sits still, and you can look at it.
Pause and try this. A hospital volunteer says, "The waiting room is full, so we should open the second check-in desk." Which part is the conclusion?
Find the claim the volunteer wants you to accept.
Here's the answer. The conclusion is "we should open the second check-in desk," because that's what the volunteer wants accepted. The waiting room being full is the premise, and the unsaid link, that a full room means the second desk should open, is the premise you'd add and mark.
2. Validity and soundness
Think about a vending machine that never fails. Put in the right coins and the snack drops, every time. The machine doesn't care what you wanted. It only promises that the right input gives the output. A valid argument works the same way.
A deductive argument aims to guarantee its conclusion. It's valid when its conclusion cannot be false if its premises are true. Notice that validity depends on the argument's form, not on whether its premises are in fact true. It's sound when it's valid and all its premises are in fact true.
So you ask two separate questions. First: if the premises were true, would the conclusion have to be true? That's validity. Second: are the premises actually true? Only two yeses make a sound argument.
Here's the chapter's fish example. Every fish can fly. Every salmon is a fish. So every salmon can fly. If both premises were true, the conclusion would have to be true. So it's valid. But the first premise is false. So it's not sound.
Pause and try this. A friend says, "That argument is valid, so its premises must be true." What's wrong with that?
Think back to the salmon.
Here's the answer. Validity says nothing about whether the premises are true. The salmon argument is valid with a false premise, because validity is a promise about form: if the premises are true, the conclusion is too.
3. The valid forms
Some shapes work every time, whatever the sentences are about. Learn them once and you'll spot them everywhere. I'll use letters. P, Q, and R each stand for a whole statement. In "if P then Q", P is the antecedent, the part after "if", and Q is the consequent, the part after "then".
The first valid form is modus ponens: if P then Q; P; so Q. The second is denying the consequent: if P then Q; not Q; so not P. The third is hypothetical syllogism, which chains two conditionals: if P then Q; if Q then R; so if P then R. The fourth is disjunctive syllogism: P or Q; not P; so Q. The fifth is the categorical syllogism, which reasons about groups: every A is B; every C is A; so every C is B.
The week's examples come from Dataset B, the argument set (illustrative). Its premises are written for teaching, and the course doesn't endorse them. Take argument 2. If the mind is the brain, then mental states have locations. Mental states have no locations. So the mind is not the brain. P is "the mind is the brain". Q is "mental states have locations". The second premise denies Q, and the conclusion denies P.
Pause and try this. Argument 4 says: either free will is an illusion or determinism is false; free will is not an illusion; so determinism is false. Which form is that?
Look for the word "or".
Here's the answer. It's disjunctive syllogism. The first premise says one of two things is true, the second rules one out, and the conclusion keeps the other.
One caution. A valid form doesn't settle the big question. Whether mental states have locations is still open, and Week 6 takes it up.
4. Invalid forms and counterexamples
Now the shapes that leak. They look almost right, but the premises can all be true while the conclusion is false.
Denying the antecedent runs: if P then Q; not P; so not Q. Reasoning from the consequent back to the antecedent runs: if P then Q; Q; so P. And the invalid categorical pattern runs: every A is B; this C is B; so this C is A.
How do you show a form is broken? You build a counterexample to a form: a case in which premises of the same form are true and the conclusion is false. One such case shows that the form is invalid.
Let's go back to the juror. The juror's argument has the backward shape: if P then Q; Q; so P. Keep that shape and swap in something anyone can check. If an animal is a dog, it has four legs. A cat has four legs. So a cat is a dog. Both premises are true. The conclusion is false. So the shape can fail, and the juror's argument, with the same shape, is invalid. Someone else in a red coat would also show up on the camera.
Here's a trap. An invalid argument can still have a true conclusion. Every dog is a mammal. This beagle is a mammal. So this beagle is a dog. The conclusion's true, but the shape is the invalid categorical pattern, and it doesn't guarantee anything.
Pause and try this. Argument 6 says: if the universe was designed, it shows order; the universe was not designed; so it shows no order. Valid or invalid?
Check what the second premise denies.
Here's the answer. It's invalid, because it denies the antecedent. Even if the universe was not designed, it could still show order for some other reason.
5. Inductive strength
Some arguments never try to guarantee anything. A commuter says, "The bus on this route has run late on every trip I've taken, so the next one will run late too." That's a sensible bet. It isn't a promise.
An inductive argument is one whose premises are offered to make its conclusion probable rather than to guarantee it. Its strength is how much its premises make the conclusion probable, without guaranteeing it. Strength comes in degrees. Validity doesn't. An argument is valid or it isn't.
So with the commuter, you ask different questions. Could the premise be true and the conclusion false? Yes, so it's not valid, and it was never trying to be. Do many late trips make another late trip likely? If the trips were many and typical, yes. Learning that the route was fixed yesterday would weaken it.
Pause and try this. Someone calls the commuter's argument "valid" because it's convincing. Which word fits better, and why?
Ask whether it's trying to guarantee its conclusion.
Here's the answer. "Strong" fits better. The argument is inductive, so it's judged by strength, which comes in degrees, and "valid" belongs to deductive arguments.
6. Counterexamples to general claims
Here's the Socratic move from Week 1, in a new setting. A hiking club's notice board says, "Every member has paid this year's dues." Then a new member joins this morning, before paying.
That one person is a counterexample to a definition or a general claim: a case the claim gets wrong. For a claim about every member of a group, one such case is enough to refute it.
Why only one? Because "every" makes a promise with no exceptions. One genuine exception breaks the promise, and more good cases can't repair it. That's different from an inductive claim about what usually happens, where one odd case proves little. The same move works on definitions. If someone defines a game as an activity with a winner, two friends tossing a ball back and forth are a case the definition gets wrong.
Pause and try this. A classmate says, "One unpaid member isn't enough to show the notice is false. You'd need a lot of them." What do you say?
Look at the word "every".
Here's the answer. One is enough. The notice claims every member has paid, so a single member who hasn't paid makes that claim false.
7. Thought experiments and the barber
Not every test case has to happen. Sometimes you imagine one. A thought experiment is an imagined scenario used to test a claim, a definition, or an intuition. It doesn't need to be real. It needs to be clear enough to follow.
This course has a roster of 15 thought experiments, one at the center of each instructional week. This week's is the barber puzzle, a standard logic puzzle usually associated with Bertrand Russell. Can there be a barber who shaves all and only those who do not shave themselves?
We'll work it as a reductio: you assume a claim, show the assumption leads to a contradiction, and conclude the claim is false. So assume the barber exists. Does he shave himself? Suppose he does. Then he's someone who shaves himself, so by his rule he doesn't. Suppose he doesn't. Then he's someone who doesn't shave himself, so by his rule he does. Both cases end in contradiction. So the assumption is false. There can be no such barber.
Pause and try this. A friend says, "Easy, the barber grows a beard." Does that solve the puzzle?
Ask whether the case has changed.
Here's the answer. No, it changes the case. That's a variation, and the course labels variations as variations. The puzzle as stated is about a barber covered by his own rule, and for that barber the contradiction stands.
8. Reading charitably
Picture a group project. One teammate says, "We should split the work evenly." Another snaps back, "So the person who did all the research should do more? That's unfair." That reply attacks a version the first teammate never meant.
Reading an argument charitably means stating it in its strongest form, the way its best defenders would, before criticizing it. You pick the most reasonable reading the words allow. If a premise is missing, you add the best one and mark it added.
Read charitably, the teammate probably means: each of us has an equal stake in the project; people with an equal stake should share the remaining work equally, counting work already done; so we should split the remaining work evenly. Now an objection has a fair target. Someone can push back on that middle premise, maybe because people's free time differs.
Pause and try this. Does reading an argument charitably mean you agree with it?
Think about what the strong version is for.
Here's the answer. No. You build the strongest version so your criticism hits the real argument. An objection that beats only a weak version has done no real work.
9. Keying the argument set
Let's put it all together. Dataset B, the argument set (illustrative), holds 10 arguments, and the course keys each one for validity. It never keys soundness, because the premises are open questions.
Arguments 1, 2, 3, 4, 7, and 9 are valid. That's 6. Arguments 1 and 7 are modus ponens. Argument 2 denies the consequent. Argument 3 is hypothetical syllogism. Argument 4 is disjunctive syllogism. Argument 9 is the categorical syllogism. Arguments 5, 6, 8, and 10 are invalid. That's 4. Argument 5 has the invalid categorical pattern, because an unjust law could also protect the vulnerable. Argument 6 denies the antecedent. Argument 8 runs the conditional backward. Argument 10 moves from what most people believe to what is right.
So 6 are valid and 4 are invalid, and together they make all 10.
Pause and try this. Argument 4 is keyed valid. Does that mean the course says determinism is false?
Remember what a key covers.
Here's the answer. No. The key covers form only, so someone who thinks free will is an illusion can agree the argument is valid and still reject its second premise.
Closing
Let's gather it up. Standard form puts numbered premises first and the conclusion last, marked C and started with "So". Valid means the conclusion can't be false if the premises are true. Sound means valid with true premises. Five valid forms are worth knowing by heart. The invalid ones leak, and a counterexample to the form shows it. Inductive arguments are judged by strength, in degrees. One counterexample refutes a claim about every case. A thought experiment tests an idea with an imagined case, and the barber puzzle works as a reductio. Reading charitably means the strongest version first. And Dataset B gives you 6 valid arguments and 4 invalid ones, keyed for form, never for soundness.
Next in this module come the slide deck, the Readings & Resources page, the lecture tutorial, the practice set, the quiz, the discussion, and the assignment. In Week 3, Plato: Reality and the Forms, you'll put these tools to work on Plato's arguments about what is real.
Built with The Course Maker · samples.thecoursemaker.com