Lecture: Argument, Reasoning & Logical Fallacies
Lecture script · about 15–20 minutes spoken · Week 13
Opening
Picture a shift supervisor at a warehouse. Late arrivals dropped the month a new schedule started, and everyone in the break room says the schedule fixed it. Now picture a neighbor at a town meeting about a community garden. One speaker says everybody on the block wants it. Another says that if the garden goes in, the whole lot will be covered in trash within a year. Both rooms are full of confident arguments. And in both rooms, somebody needs to ask a quiet question: does the reasoning actually hold?
That question is this lecture.
Here's where we are. Week 4 gave you evidence and the oral citation, the who, what, and when you say before the evidence. Week 12 turned the running speech into a persuasive speech with a policy claim, built on ethos, pathos, and logos. Logos is the appeal to evidence and reasoning. This week we open up the reasoning part and look at how it's built and how it breaks. Next week the course turns to special-occasion speaking and working in small groups.
In this lecture I'll take you through five ideas. First, the four types of reasoning. Second, Toulmin's map of an argument. Third and fourth, the fallacies that break an argument's support and the ones that dodge or rig it. Fifth, two drills plus a check on the machine.
1. Four types of reasoning
Think about how you decide a new coffee shop is good. Maybe you went three times and every drink was great. Maybe a friend told you every shop that roasts its own beans is good, and this one does. Maybe your afternoons got better after you switched shops. Or maybe you loved the shop's other location, so you figure this one's similar. Four different paths from evidence to a conclusion.
That's reasoning: the process of moving from evidence or principles to a conclusion. This course names four types. Other treatments group them differently, so you won't be asked to recite a count.
Inductive reasoning moves from specific cases to a general conclusion. Three great drinks, so the shop is good. Its strength depends on how many cases there are and how typical they are.
Deductive reasoning runs the other way, from a general principle to a specific conclusion. Its classic form is the syllogism: two premises, one general and one specific, and the conclusion they lead to.
Causal reasoning claims one thing produces another. And here's the rule to tattoo on your notes: sequence and correlation are not cause. If B came after A, that alone doesn't show A caused B.
Analogical reasoning says two cases alike in known ways are likely alike in another way. It holds only when they're alike in the ways that matter.
Now let's find all four inside the running speech. Its claim is that the dining hall should run a two-week plate-waste audit before deciding whether to remove trays.
Deductive first. Premise one, the general principle: a campus should measure a change's effect before adopting it when the best available study found no significant effect. Premise two, the specific case: the best available study on trays found no significant reduction in food waste. Conclusion: measure before removing trays. That's 2 premises and 1 conclusion, so 3 statements.
Causal next. The speech's second oral citation is a study by Wenhao Zhang and Junehee Kwon, published in the journal Sustainability in December 2022. In a university dining hall, a week without trays reduced how much food diners took. But it did not show a significant reduction in food waste. So the speech doesn't claim a cause the study didn't find.
Analogical: copying another campus's decision assumes the two dining halls match in the ways that matter. The qualifier "in dining halls like ours" says that out loud.
And inductive: one study's two weeks in one dining hall are specific cases. Stretching them into "trayless never works anywhere" would generalize too far. That's 4 types of reasoning, each found once.
Pause and try this. "Every shop that roasts its own beans is good, this shop roasts its own, so this shop is good." Inductive or deductive?
Think about which way the argument travels.
Here's the answer. It's deductive, because it starts from a general principle and lands on one specific case.
2. Toulmin's map of an argument
Imagine a friend says you should take the early bus. You ask why. They say it's never crowded. You both nod, because you both assume something nobody said out loud: a less crowded bus is a better bus. That unspoken assumption is the hinge of the whole argument.
Stephen Toulmin gave that hinge a name. Toulmin's model, from Toulmin's 1958 book The Uses of Argument, lays out an argument's parts. It names 6 parts, and this course keys 3 of them.
The claim is the statement you want the audience to accept. The evidence, which Toulmin called grounds, is the supporting material you offer, cited aloud with who, what, and when. And the warrant is the usually unstated principle that makes the evidence count as proof of the claim. It's the link, not a fact you cite.
The other three are named but not keyed. Backing supports the warrant itself. A qualifier limits how strongly or widely the claim applies. A rebuttal names the condition under which the claim wouldn't hold. So 6 parts, minus the 3 keyed ones, leaves 3 named but not keyed.
Let's map the running speech. The claim: run a two-week plate-waste audit before deciding on trays. The evidence: oral citation 1, the U.S. Department of Agriculture's Food Waste FAQ page, which puts food waste at between 30 and 40 percent of the food supply in the United States. And oral citation 2, the Zhang and Kwon study.
The warrant: a campus should measure a change's effect before adopting it when the best available study found no significant effect. The qualifiers: "for a two-week trial" and "in dining halls like ours." The rebuttal: "unless the audit costs more than the trial."
Count it up. One claim, 2 pieces of evidence, one warrant, 2 qualifiers, and one rebuttal. That's 7 statements across 5 of the 6 parts. Backing is the one the course's map leaves unstated.
Now the distinction that matters most: warrant versus evidence. Ask one question. Could a listener look it up? If yes, it's evidence. The Zhang and Kwon study is evidence. If it's the reason the evidence matters, it's the warrant. "Measure before adopting" is the warrant. When a listener rejects an argument with solid evidence, the warrant is usually where they got off the bus.
Pause and try this. In the running speech, is "measure a change's effect before adopting it" the evidence or the warrant?
Try the look-it-up test on it.
Here's the answer. It's the warrant, because nobody can look it up at a source; it's the principle that turns the study into a reason for the audit.
3. Fallacies that break the support
Think of a chair with a cracked leg. From across the room it looks fine. Sit on it and you find out. A fallacy is a flaw in reasoning that makes an argument look stronger than it is, usually by breaking the link between the evidence and the claim.
This course names 12 fallacies. Other treatments list more or fewer, so the count is taught in words. For study, I'll split them into two groups of six. The first six break an argument's support. Every example is made up for teaching and names no real person.
Hasty generalization draws a general conclusion from too few cases. "I asked three friends who eat at their desks, and all three keep them spotless, so every student will keep the library clean." It's an error about sample size.
False cause treats one thing as the cause of another because it came first or because the two occur together. "Sales at the café went up the month the new manager started, so the new manager caused the rise." It's an error about causation.
Weak analogy compares two cases that differ in ways that matter. A pizza shop guarantees its delivery time, so campus security should guarantee its response time too. But an emergency isn't a pizza order.
Appeal to ignorance treats a lack of proof as proof. Nobody proved the helmet loaner program made riders safer, so it made no difference.
False authority offers someone's opinion as proof when they have no expertise in the field. A popular gaming streamer says a bike helmet is the safest on the market. The fix comes from Week 4: expert testimony from someone qualified in helmet safety, cited aloud with who, what, and when.
Bandwagon treats a claim as true because many people believe it. Everyone on my floor snacks in the library, so it must be fine.
Here's the mix-up that trips almost everyone. False cause and hasty generalization both feel like jumping to conclusions, but they jump in different directions. Ask one sorting question: is this about how many cases, or about what caused what? Three friends standing in for every student is a sample problem. A manager and rising sales is a cause problem.
Pause and try this. "Sales at the café went up the month the new manager started, so the new manager caused the rise." Hasty generalization or false cause?
Ask the sorting question before you choose.
Here's the answer. It's false cause, because the error is about what caused what, not about how many cases were counted.
4. Fallacies that dodge or rig the argument
Picture a neighborhood meeting about a later closing time for the public library. One neighbor makes a careful case. The next speaker never answers it. Instead they mock the first neighbor's job, or describe a sillier version of the idea, or start talking about parking. The argument didn't lose. It got dodged.
The second six fallacies dodge the real argument or rig the choice.
Ad hominem attacks the person and offers that attack as a reason to reject the argument. Don't listen to his plan for security response times, he got a parking ticket last month.
Straw man attacks a distorted version of the position. Someone proposes covered drinks in the library, and the reply is, so you think spills on the rare books don't matter.
Red herring changes the subject. Asked why security responses are slow, a speaker talks about the new parking garage.
False dilemma offers two options when more exist. Either we ban all food in the library, or the carpets get ruined. Week 12 previewed this one as false urgency, decide now or lose everything.
Slippery slope claims one step will set off a chain of worse steps, with no evidence for each step. Allow coffee, then pizza, then a food court.
Begging the question uses the conclusion as one of its own premises. The plan will work because it's a plan that works. And notice, begging the question doesn't mean raising a question.
Two more mix-ups. Ad hominem isn't any insult; it's an attack on the person offered as a reason to reject the argument. A straw man stays on the topic but bends the position, while a red herring leaves the topic. And not every emotional appeal is a fallacy. Pathos is legitimate when it's relevant and proportionate. It becomes a fallacy when it replaces the evidence.
Pause and try this. Someone proposes covered drinks, and the reply is, "So she wants the library to be a café." Straw man or red herring?
Ask whether the reply stayed on the topic.
Here's the answer. It's a straw man, because the reply stays on the proposal but attacks a bent version of it instead of the one she made.
5. The drills, and a check on the machine
A mechanic who spots a worn brake pad is useful. A mechanic who can also replace it is the one you call back. So this week has two drills.
The spot-the-fallacy drill has 4 items. Campus security's response time, answered with talk about a parking garage: red herring. The new manager and rising sales: false cause. Food in the library, either ban it all or ruin the carpets: false dilemma. And the gaming streamer's helmet: false authority. Add the 2 fallacies from the opening meeting, the three friends and the coffee chain, and you've labelled 6 fallacy examples.
The build-the-argument drill turns it around. Take one main point of your own persuasive speech, on a topic from the course roster. Write five lines: your claim, your evidence as a complete oral citation, your warrant, a qualifier, and a rebuttal. Then read your warrant out loud and ask whether a skeptical listener would accept it.
Now the machine. Say you ask a chatbot to label the new-manager item. It answers that this is a hasty generalization, because it draws a conclusion from limited evidence and assumes the manager caused the increase. Then you ask it to check the running speech, and it calls the compost-bin scene an appeal-to-emotion fallacy and tells you to cut it.
Check each line against the definitions. The label is wrong: the error is about causation, so it's false cause. The explanation blends two fallacies, the "limited evidence" piece of one and the "caused" piece of the other. And the compost-bin scene is relevant, proportionate, and labelled illustrative, and it sits beside a real study, so it's legitimate pathos. That's 3 problems caught. A confident label isn't a correct one.
Pause and try this. The chatbot calls the compost-bin scene a fallacy. What makes it legitimate pathos instead?
Think about what sits right beside the scene in the speech.
Here's the answer. It's relevant and proportionate, and it sits beside real evidence instead of replacing it, so it's pathos doing its honest job.
Closing
Let me pull it together in the order we built it. Reasoning comes in four types this course names: inductive, deductive, causal, and analogical, and sequence and correlation are never cause by themselves. Toulmin's model keys the claim, the evidence, and the warrant, and the warrant is the link you usually don't say. The first six fallacies break an argument's support, and the sorting question tells false cause from hasty generalization. The second six dodge or rig the argument, and pathos is fair when it stands beside evidence. And every chatbot label gets checked against the definition.
Next in this module come the slide deck and the Readings & Resources page, then the lecture tutorial, the practice set, and the quiz. Next week, in Special-Occasion & Small-Group / Team Communication, you'll fit a speech to a moment and help a group reason its way to a decision.
Built with The Course Maker · samples.thecoursemaker.com