Alex decides that Policy X is correct, then gathers and presents five strong supporting arguments. Beth evaluates the same five arguments before reaching any conclusion, and also ends up supporting Policy X. Whose arguments provide stronger evidence that Policy X is correct?
AAlex's arguments are stronger because he has more conviction behind them
BBoth provide equal evidence — the logical validity of an argument doesn't depend on the order in which it was found
CBeth's arguments provide stronger evidence, because Alex's were selected to support a conclusion already written, making them zero net evidence
DNeither provides evidence — only empirical data, not arguments, can support a policy conclusion
The key insight is that evidential weight depends on the process by which arguments were found, not just their logical form. Alex wrote the bottom line first, then filled in supporting arguments above it. No matter how valid those arguments appear, they were chosen from a space of possible arguments precisely because they support X — so finding them tells us nothing about whether X is true. Beth's reasoning process could have led her anywhere; Alex's could not. The same argument, encountered honestly, carries information; encountered via rationalization, it carries none.
Question 2 Multiple Choice
Which of the following is the clearest example of 'writing the bottom line'?
AA scientist forms a hypothesis and then runs an experiment to test it, finding support for her hypothesis
BA Bayesian reasoner starts with a strong prior that a drug is effective, then updates on clinical trial data that confirms it
CA manager concludes before reviewing the data that the new product launch was a success, then selects metrics that show positive results to include in the report
DA debater is assigned to argue one side of a topic and constructs the strongest possible case for that side
The manager wrote the bottom line — 'the launch was a success' — before examining evidence, then filtered the evidence through that conclusion. The hypothesis-testing scientist and Bayesian reasoner both have prior beliefs, but their process genuinely exposes them to contrary evidence that could change their conclusion. The debater is explicitly performing advocacy, not reasoning — both parties know the argument is one-sided, so no one is deceived about its evidential status.
Question 3 True / False
An argument that is logically valid and supported by true premises carries the same evidential weight regardless of whether it was found before or after the reasoner committed to its conclusion.
TTrue
FFalse
Answer: False
The bottom-line insight is precisely that evidential weight is process-dependent, not just structure-dependent. A valid argument found after committing to a conclusion was selected from all possible arguments because it supports that conclusion — the act of selection tells us nothing new about the world. If you looked for arguments on both sides and found this one was strongest, that's evidence. If you only looked for arguments supporting your predetermined conclusion, finding one is uninformative. The argument's internal structure is irrelevant to this epistemic point.
Question 4 True / False
Having a strong prior belief about a conclusion is equivalent to having 'written the bottom line' — in both cases, your subsequent reasoning is rationalization.
TTrue
FFalse
Answer: False
This is a critical distinction the topic explicitly addresses. A prior belief becomes the bottom line only when it is fixed — when no possible evidence could change it. Bayesian reasoning begins with priors and genuinely updates on evidence; the conclusion is not fixed in advance, even if the prior is strong. The test is counterfactual: could evidence change your mind? If yes, you are reasoning. If no possible evidence would move you, the conclusion was written before the reasoning started, and any subsequent 'evidence gathering' is rationalization.
Question 5 Short Answer
What practical test can you apply to determine whether you have written the bottom line on a question, and what does the test reveal?
Think about your answer, then reveal below.
Model answer: Specify in advance what evidence would change your conclusion. If you cannot name any such evidence — or if the evidence you name would have to be extraordinary while confirming evidence can be ordinary — you have likely written the bottom line. The test reveals whether your belief is held in a way that could, in principle, respond to the world.
The bottom-line problem is invisible from the inside — rationalization feels like reasoning. The test forces you to make your belief falsifiable in your own mind before evaluating evidence. If you find yourself unable to specify any conceivable evidence that would change your conclusion, that is a strong signal the conclusion is fixed and subsequent 'reasoning' is performance. A genuine reasoner can always describe, at least approximately, what the world would have to look like for them to be wrong.