One claim at a time
Before you answer, ask what the person is actually claiming and whether they want to examine it. A conversation becomes useful when both people can identify the question under discussion. If the speaker moves from a miracle to morality to the origin of the universe in one breath, choose one claim and return to it.
The examples below are constructed teaching scenarios. “Missionary” and “defender” name roles in the exercise, not a particular church or individual. They are not transcripts of particular missionaries or accounts of completed outreach. Use their reasoning as a starting point, then respond to what the person in front of you actually says.
“I felt it, so I know”
Missionary: “When I prayed, I felt an overwhelming peace. That's how I know the teaching is true.”
Response: “What helped you identify the source of that feeling? If someone in a different religion had the same experience, would it establish their teaching too?”
The question asks for a method that can distinguish competing explanations. It accepts that an experience occurred while examining the conclusion attached to it. The religious-experience study in our research notes provides one example of investigating the human processes involved. Research note 10
If the speaker answers that only their tradition produces the right feeling, ask how that difference can be checked independently. If the answer depends on first accepting the doctrine, identify the circular step. The next useful question is what evidence could be shared with someone who does not already believe.
“What if you're wrong and go to hell?”
Missionary: “Wouldn't it be safer to believe, just in case?”
Response: “Which warning should I accept, and how would I choose among incompatible ones? The size of a threatened consequence doesn't tell me which claim is true.”
Work through the logic. A threat can make a decision feel urgent while leaving its factual basis unestablished. Ask for the evidence for this particular account of punishment. If fear is the only support being offered, name that. “You've given me a consequence to fear. What establishes that the consequence is real?”
If a participant becomes distressed, pause the debate. They do not owe the room a public test of courage. An audience member who has lived with the threat may need privacy and time to examine it. The quality of an argument does not depend on forcing that examination in front of strangers.
“Without God, anything goes”
Missionary: “If there's no God, what stops people from doing whatever they want?”
Response: “Other people can be harmed. Their lives and choices matter. We can examine consequences, consent, fairness, and responsibility. How does calling an action a command settle whether it is good?”
Make the question ordinary. In this fictional case, you find a wallet on a bus with money and the owner's contact details inside. Keeping it would cost that person money and the trouble of replacing documents. Returning it respects their property and helps them avoid those losses. Ask the speaker which of those reasons they accept. This gives you consequences to discuss together; it does not pretend that one wallet story settles every question in moral philosophy.
If they insist that only authority makes an action good, ask how they would evaluate a claimed divine command to harm someone. Who verifies the command? What would allow a person to challenge it? The problem of evaluating authority remains even within the speaker's framework.
“But why do those things matter objectively?”
Response: “You’re asking what grounds a moral obligation, beyond what motivates someone to help. Our view starts with concern for suffering, agency, and fair treatment. I can defend those commitments, but I won’t pretend a list of good consequences solves every question about objectivity. How does your account establish its own starting point?”
If the reply is “God’s nature,” represent that position as discussed in lecture 7. Ask how it is known and how conflicting claimed commands are judged. A gap in a secular account would not, by itself, identify the correct religious account. Keep the question open long enough for both explanations to be examined.
“Science changes its mind”
Missionary: “Scientists keep changing their answers, so why trust them?”
Response: “Which answer changed, and what evidence caused the change? A method that can correct an error gives us something to inspect. How does your method correct its own mistakes?”
Choose a real claim instead of defending every scientific statement ever made. For an argument about Earth's age, explain the specific dating method and ask for a specific criticism. USGS describes the evidence behind the approximately 4.54-billion-year estimate. If someone objects, identify what part of that evidence their objection addresses. Research note 14
A legitimate methodological criticism should be examined. A refusal to permit any result that conflicts with doctrine deserves a different response: ask whether the test was ever allowed to produce a contrary conclusion.
“The text proves the text”
Missionary: “We know this book is divinely inspired because the book says so.”
Response: “What evidence supports that claim independently of accepting the book's authority? Would the same argument work for another sacred text?”
Historical questions need historical methods. Ask about authorship, dating, transmission, independence of accounts, and how close the evidence is to the event. Evidence that a place existed does not by itself establish every supernatural story set there. A source can preserve genuine history and still contain claims that require further investigation.
If somebody presents a particular manuscript or archaeological finding, examine it specifically. Do not invent a historical rebuttal because you expect the overall doctrine to be false. Precision is part of earning the confidence you ask other people to reconsider.
“The universe must have a first cause”
Defender: “Things that begin require a cause. The universe began. Its cause must transcend the universe.”
Response: “Let’s examine those premises, then the attributes you think follow. What establishes that this cause has intentions, moral authority, or the identity described by your religion?”
The kalam argument concerns things that begin; replying “then who caused God?” alone misses its proposed distinction between a beginning universe and an uncaused cause. Craig adds further reasoning about the cause’s properties. Those steps need to be examined rather than silently granted. Other cosmological arguments concern contingency rather than a temporal beginning, so ask which argument is actually being made. Research note 23
Next round: “A timeless cause could choose to create.” Ask how agency is established and which rival explanations are excluded. If neither of you can evaluate the relevant physics or metaphysics, identify the exact premise and return with sources. An unknown origin gives you a question; it does not identify a particular deity.
“The universe is too finely tuned to be accidental”
Defender: “Life depends on highly specific conditions. Design explains those conditions better than chance or necessity.”
Response: “Which conditions and physical model are you using? What makes design more probable, and what assumptions about a designer’s intentions enter that comparison?”
Craig frames fine-tuning as an inference among necessity, chance, and design. A serious response examines that comparison. Ask how the probability distribution over possible conditions is justified and what observations each explanation leads us to expect. Avoid declaring a multiverse established, or treating the fact that observers can only observe compatible conditions as a complete explanation. Research note 24
Next round: “You have no better explanation.” We can acknowledge a live explanatory problem while asking whether the proposed answer earns its added claims. A possible designer would still leave questions about identity, intentions, revelation, and obligations.
A historical claim: the resurrection
Defender: “Early testimony, appearances, and the rise of Christianity are best explained by Jesus rising from the dead.”
Response: “Which sources establish each part, and which are independent? Let’s start with one passage rather than treating an entire historical argument as one fact.”
Craig’s case appeals to historical testimony and explanatory comparison. It deserves that response, rather than being reduced to a book endorsing itself. Research note 25
Work through 1 Corinthians 15:3–8. Paul passes on a tradition about death, burial, resurrection, and appearances, and includes a claim about his own experience. The passage reports an appearance to more than 500 people; it does not supply 500 separately preserved statements. A fair reading distinguishes Paul’s account, material he says he received, and reports about others. Primary passage, research note 26
That distinction matters without proving the event false. Establishing that a belief was expressed early and explaining what caused the reported experiences are different tasks. Ask about dating, authorship, transmission, independent attestation, and competing explanations. Manuscript preservation can help establish what a text said; it does not automatically verify the event described.
Next round: “What would change your mind?” Independent, well-documented evidence that meaningfully favors the resurrection over alternatives should change our assessment. We should not invent a fraud, hallucination, or conspiracy to fill every unknown. State which alternative is supported and where the evidence leaves the question unresolved. This example teaches source evaluation; it is not an exhaustive historical adjudication.
When the standard keeps moving
After you identify a rule, watch whether it changes when the answer is inconvenient. You can say: “A moment ago, personal testimony was sufficient. Now you're asking for independent records because the testimony supports a different belief. Which standard should we use for both?”
Motivated reasoning provides a framework for understanding selective scrutiny, but the conversation should focus on the actual inconsistency you can show. Research note 4
If no possible evidence would change the position, say so plainly. “Then we have reached the limit of what evidence can do in this conversation. The claim is being protected from testing.” You can end an unproductive exchange without pretending it was settled.
Exercise and discussion
In pairs, choose one scenario. One person presents the claim; the other asks three questions before offering a rebuttal. Then switch roles. The observer, if available, checks whether the response represented the claim accurately, kept one evidence standard, and gave the speaker a chance to answer.
Discuss where the conversation became clearer and where it turned into a contest. A good result is a more precise question, a visible inconsistency, or an agreed next source to examine. The next lecture considers who carries the risk when that conversation becomes public.