Back to the on-screen lesson ·
When a claim rests on somebody's word, who they are is part of the evidence; when it rests on reasons, it is not.
Paper packet. Every task here also exists on screen, where it is checked automatically; answers written on paper are not assessed by Nydus. When you are back at a device, enter your answers there.
You will decide whether a remark about the person speaking belongs in the argument, by first listing what the claim is actually resting on. You will mark the reply that leaves checkable reasons standing untouched, leave alone the challenge that bears on somebody's testimony, and say what a fair challenge to a source is entitled to conclude.
Bel proposed moving the two taps to the middle of the site.
She gave three reasons: the longest carry is halved, the old pipe is the one that froze, and the middle plots have no tap at all.
Ade replied that Bel keeps the untidiest plot on the site and has been asked twice about her bonfires.
A roofer told the committee the shed roof will last another thirty years.
He showed no measurements and would be paid for the work he recommended instead.
Ade replied that the committee should get a second estimate before believing that one.
You have just done the same shape of thinking about experts: whether somebody's word counts depends on what the word is being asked to do. This lesson is the other half of it — when something bad about a person is a reason, and when it is a way of not answering.
| Term | What it means |
|---|---|
| Ad hominem | A personal attack used in place of addressing the relevant support for a claim. |
| Source challenge | An objection connected to a source's opportunity, method, competence, or reliability. |
| Tu quoque | Dismissing a claim merely because its speaker behaves inconsistently with it. |
| Genetic fallacy | Judging a claim solely from an origin that does not establish its merits. |
| Provenance | The traceable origin and history of a piece of evidence. |
Read the two objections in the minutes. Both are about a person rather than about a claim, both are true so far as anybody knows, and Ade made both of them. Somebody at that meeting will now say that Ade attacks people instead of answering them, and somebody else will say that Ade is the only one asking who is being believed and why.
Both of those are worth something, which is what makes this a case rather than an exercise. Rule out every remark about the speaker and the committee has to take a paid roofer's unsupported estimate at face value. Allow them all and Bel's three reasons about the pipe never get answered by anybody.
So the question is not did Ade talk about a person. He did, twice. It is what was each claim standing on when he did it — and that is a question about the two proposals, not about Ade.
Somebody gives you a conclusion. Before you decide whether a remark about them is fair, ask what the conclusion is standing on.
Case one: reasons you can check. Bel gives three reasons for emptying the bins twice a week. Those reasons are now on the table, and they are as good or as bad as they are. You could learn that Bel is careless, untidy and generally disliked and not one of the three reasons would be worth a penny less. Replying with any of that is changing the subject.
Case two: the person's word. A supplier tells you the roof will last thirty years, with no working shown. Here the claim is resting on his judgment and on nothing else — so how his judgment has turned out before, and whether he profits from your believing it, are directly to the point.
And notice what the fair version concludes. It does not conclude the roof will fail. It concludes his estimate is not worth much, which leaves the roof exactly as unknown as it was.
Another way: picture
Picture the claim standing on legs. If its legs are reasons, sawing at the speaker does nothing: the legs are not attached to them. If its only leg is because I say so, then the speaker is the leg, and everything about their reliability is load-bearing.
Another way: steps
Three steps:
| The question | Why it bears on the report |
|---|---|
| Do they gain if you believe it? | A stake is a reason to want more than a word |
| How have their past judgments turned out? | An estimate is a judgment, and judgments have records |
| Were they in a position to know? | A report from somebody who was not there is not a report |
All three are about how much weight one person's say-so will bear. None of them is an insult, and none of them shows the claim false. A supplier with a terrible record may be right about this roof; you simply have no reason from him to think so.
The unfair version always has the same shape: reasons on the page, and a reply about the person, and the reasons still sitting there at the end of it.
A personal detail becomes relevant to evaluating an argument when it bears on the evidence being offered. Suppose a witness reports seeing a gate open at six, but a reliable location record places the witness elsewhere at that time. The detail matters because it challenges the witness's opportunity to observe the gate. By contrast, a remark about the witness's untidy desk does not normally bear on that opportunity. Both remarks concern the person, but only one supplies a relevant connection to this report.
Write the connection out explicitly. 'The witness was elsewhere, so this cannot be a firsthand observation from the stated location' is a limited conclusion about the report's basis. 'The witness was elsewhere, so the gate was closed' goes too far. The gate might have been open even though this witness did not see it. Removing one source of support does not establish the opposite claim. A careful objection states what evidence has been weakened and what remains unknown.
The same care applies when evidence is presented in a checkable form. A photograph, table, or set of figures is not automatically independent of its producer. Someone chose the camera angle, selected the measurements, and decided which cases to include. A financial stake may be relevant if it gives a concrete reason to inspect those choices. However, merely naming a stake does not show that a photograph is fake or a calculation wrong. The next step is to check the evidence path rather than replace examination with suspicion.
For a photograph of a rotten fence, distinguish several claims. The image may show that one section has rot. It may not establish how much of the entire fence is affected. The proposal to replace all 40 meters requires information about the rest of the fence and the alternatives. Asking which sections were photographed is an evidential question. Calling the photographer unpleasant is not. Asking whether original images and inspection notes are available can help assess selection without assuming dishonesty.
A relevant track record also needs a proper comparison. If a contractor repeatedly underestimated completion time for similar jobs under similar conditions, that history may weaken a bare promise about the next job. A complaint about an unrelated personal disagreement does not have the same force. Ask what past errors concern, how many relevant cases are known, and whether conditions or methods have changed. A single old mistake does not make every later statement false, just as a long record of success does not guarantee the next result.
Keep the tone separate from the logic. A rude statement can raise a relevant evidential issue, and a courteous statement can evade the argument. That does not make rudeness desirable. In a useful discussion, aim for both relevance and respect: name the information needed, explain why it matters, and avoid claims about character that the evidence does not support. 'Please show how the sample was selected' advances the inquiry more directly than 'You are the kind of person who would select a biased sample'.
A familiar reply to advice is 'You do it too.' This is called tu quoque, a Latin phrase commonly translated as 'you too', when it is used to dismiss a claim merely because the speaker behaves inconsistently with it. Suppose a student gives accurate measurements showing that leaving the tap running wastes water. Another student replies that the speaker left a tap running yesterday. That reply may identify inconsistent behavior, but it does not change the measurements or show that running taps conserve water.
Write two separate propositions: the advice is well supported; the adviser follows the advice. Either can be true without the other. A person can understand a good rule and fail to follow it. A person can also follow a bad rule consistently. The argument about water use therefore needs an answer about the measurements and the proposed action. The question about hypocrisy is a further discussion about behavior, sincerity, or fairness of enforcement.
There are contexts where inconsistent behavior is relevant. If a candidate's claim is 'I always follow this rule', a documented counterexample directly challenges that claim. If a policy grants one group exemptions while penalizing another under otherwise similar conditions, the difference may bear on fairness. The mistake is not mentioning inconsistency anywhere. It is using that inconsistency to refute an unrelated factual or practical claim without supplying the connection. State exactly which claim the behavior is supposed to undermine.
The same discipline helps with the genetic fallacy: accepting or rejecting a claim solely because of its origin when that origin does not establish its present merits. Imagine rejecting a correct calculation because the first person to suggest it belonged to a disliked club. The club affiliation does not alter the addition. Or imagine accepting a planning proposal only because it originated in a prestigious organization. Prestige does not replace examining the assumptions, costs, and expected effects of this proposal.
Origins can nevertheless be evidentially important. An anonymous screenshot with no original document is weaker evidence about what a city council decided than the city council's authenticated minutes. That is a point about provenance and verification, not a claim that every sentence from an anonymous person is false. Similarly, the origin of a rumor may reveal that five reports copy a single unverified story. The relevant question is how the origin affects the route to knowledge. Avoid the slogan that origins never matter; identify the particular connection that makes them matter here.
Use a three-line response when these moves arise. First state the original claim. Second state what the personal or historical detail actually establishes. Third say whether that established fact bears on the original support. In the tap case: running water increases measured use; the adviser also wasted water yesterday; the inconsistency does not refute the measurement. This response acknowledges a potentially true criticism without allowing it to stand in for an answer to a different argument.
A fictional neighborhood committee receives a 2,400-credit quotation to replace 40 meters of fencing. The volunteer recommending replacement supplies photographs of rot in two sections, each 2 meters long. Her brother works for the company that supplied the quotation. One committee member calls her careless. A second asks whether the remaining 36 meters were inspected. A third notes the family connection and asks for another quotation using the same specification.
The careless label does not answer the photographs, the measurements, or the proposed price. It gives no stated reason to think that the two photographed sections lack rot. The question about the other 36 meters directly addresses the step from evidence about 4 meters to a proposal covering all 40. Even if the photographs are entirely authentic, they do not by themselves establish the condition of the unpictured sections. That objection has identified a missing evidential link.
The family connection does not establish fraud or an excessive price. It makes transparent comparison particularly useful because the volunteer may have an interest in the supplier's selection. A second quotation would need the same length, materials, disposal work, and installation terms to answer the price question fairly. Otherwise differences in scope could explain different totals.
The committee can record a limited conclusion: there is evidence of rot in at least the photographed sections, but the full replacement scope and comparable price need checking. It should not record 'the fence is sound' merely because the proposer has a family connection. Nor should it record 'the whole quotation is proved' simply because photographs exist. Careful source criticism follows the evidence through each separate claim.
Thinking any mention of the person is out of order. It is not. Where a claim rests on testimony, the teller is part of the evidence.
Thinking a true remark must be a relevant one. Wen is always late can be perfectly true and still leave the bus schedule exactly where it was.
Letting a fair challenge conclude too much. Showing that somebody's word is worth little shows that you do not know, not that the opposite is true. A liar who says it is raining has not made the sun come out.
Identify the claim.
The gate was open at six.
This is the fact the report asks us to accept.
Identify its stated basis.
Mira says she personally saw it.
The report depends on an opportunity to observe.
State the relevant objection.
A reliable location record places Mira two streets away at six.
The record bears directly on the claimed observation point.
Draw the limited conclusion.
This account is not supported as the stated firsthand observation.
The challenge weakens the report's basis, not the gate's actual state.
Reject an excessive conclusion.
The gate is not thereby proved closed.
A fact may be true even when this person's testimony fails to establish it.
State the water-use claim.
Leaving the tap running for longer uses more water at the same flow rate.
The claim concerns a measurable relation.
Identify the offered support.
A timed collection measures the additional liters.
The observation can be checked independently of the adviser's habits.
Identify the personal reply.
The adviser left a tap running yesterday.
This concerns the adviser's conduct rather than the collection result.
Separate the two conclusions.
The adviser was inconsistent; the measurement may still be correct.
Inconsistency does not alter the quantity collected.
Name the misstep and the appropriate next question.
Tu quoque if used as refutation; check the measurement to challenge the factual claim.
A relevant criticism must bear on the support or the claim being discussed.
Record the proposed work and cost.
Replace 40 meters for 2,400 credits.
The proposal concerns the whole fence and a particular price.
Record the actual photographic scope.
Two 2-meter sections: 4 meters pictured.
The photographs support claims about the displayed sections.
Identify the missing coverage.
40 − 4 = 36 meters not pictured.
Whole-fence replacement requires evidence beyond the photographed portion.
Assess the family connection proportionately.
A possible interest supports asking for transparent comparison.
It does not by itself prove the quotation false or the photographs fabricated.
Identify a useful check.
Inspect the remaining sections and compare an equivalent specification.
Those actions address the scope and price claims directly.
State what has and has not been established.
Some rot is evidenced; full replacement and price remain to be justified.
The conclusion tracks the available evidence without turning suspicion into a contrary fact.
State the calculation offered.
Eight packs at 12 credits each cost 96 credits.
The price total rests on multiplication.
Identify the objection.
A disliked club member first suggested this calculation.
The objection supplies an origin, not an arithmetic error.
Check whether the origin changes the result.
Four replies, each of them about the person rather than about the world. Match each to whether it is a fair point here.
| A fair point about the source | An attack that leaves the reasons untouched | |
|---|---|---|
| He says the pipe is sound and has not looked at it since March, so his word is worth little. | ||
| She measured the damp in four places, but her coat is a mess, so ignore her. | ||
| She is the only witness, and she was in another town that evening, so her account is worth little. | ||
| He costed the job line by line, but his cousin is on the committee, so the costing is wrong. |
A proposal replaces a 40-meter fence. Photographs show two different 2-meter sections with rot. Complete the evidence-scope check.
Find the length shown in the photographs.
2 × 2 = shown meters.
The two non-overlapping sections each cover two meters.
Find the length not shown.
Total length minus photographed length = unshown meters.
The whole-fence proposal extends beyond the photographed evidence.
Compare photographed scope with proposed scope.
Unpictured length divided by total length is share percent.
This gives the share of the proposed fence for which the photographs supply no direct view.
*Bel argues that the bins should be emptied twice a week in summer, and gives three reasons. Sam replies that Bel's handwriting is unreadable. So Bel is wrong.* Is the point about the person a fair one here?
Construct only the supported connections. A witness says they saw a gate open at six. Verified travel records put the witness in another town at six. Someone also says the witness has an untidy desk. Connect the evidence to the limited conclusion it supports; do not infer the gate's actual state from a failed report.
This task has no paper form; do it on a device.
A budget gives eight packs at 12 credits each. A verified multiplication gives 96. The proposal first came from an unpopular club. Construct only evidential support links to the claims shown; club popularity supplies no arithmetic evidence.
This task has no paper form; do it on a device.
A disagreement about a second-hand van rather than a parking lot. Mark the one remark about a person that answers nothing.
This task has no paper form; do it on a device.
Lesson test: one question per skill, one attempt each, no hints. Your answers are checked when you submit.
A student says running a tap for longer at the same rate uses more water. A checked collection experiment shows 2 liters in one minute and 4 liters in two. A reliable video also shows the student needlessly leaving a tap running yesterday. Draw only supported links: evaluate the water claim separately from the student's conduct.
This task has no paper form; do it on a device.
You can tell an attack that answers nothing from a fair question about how much somebody's word is worth, and you can say why the fair version concludes something about the report rather than about the world. Explain why a true remark about a person can still be beside the point. Next: what a count of people can and cannot settle.
15. A calculation with an unpopular origin, step 3
8 × 12 = 96 regardless of who first wrote it.
Rejecting the calculation solely for this origin would commit the genetic fallacy.