Testing Worldviews Fairly

The Standard You Would Accept From Your Opponent

Lesson 4.4 ended with an assignment and a promise. The assignment was to write a page describing a rival worldview well enough that someone who holds it would sign it. The promise was that the criteria used on those rivals would be turned around and run on my own position without exemption. This lesson is where I pay that.

Start with why it matters, because the temptation is enormous and almost invisible. Every one of us grades on a curve. We ask an opposing view to explain everything and account for every hard case; we ask our own to be roughly plausible and emotionally satisfying. We call an unresolved problem in their system a fatal flaw and the same problem in ours a mystery. Nobody does this on purpose, which is what makes it dangerous. Motivated reasoning does not feel like motivated reasoning from the inside. It feels like being right.

So here is the framework, and it gets a name because you will need to recall it under pressure. The Four Tests: coherence, correspondence, explanatory power, and livability. And the rule that governs them, which matters more than the tests themselves: no home field. Whatever standard you apply to a rival, you apply to yourself in the same paragraph, at the same strength, in public.

Scripture is not neutral on this. 'But examine everything carefully; hold fast to that which is good' (1 Thessalonians 5:21, NASB 1995) is not addressed to skeptics. It is addressed to believers, about their own teaching. The Bereans in Acts 17 are commended — in the same chapter where Paul debates Athenian philosophers — for examining the Scriptures daily to see whether what an apostle told them was so. That is a community fact-checking its own authority figure and being praised for it. And Proverbs 18:13 (NASB 2020): 'One who gives an answer before he hears, It is foolishness and shame to him.' Scripture is the foundation here, above tradition and above denomination, and one of the things it founds is the obligation to test.

Now the executive case. Applying the same standard to your own convictions that you apply to your opponent's is the most reliable credibility engine available to a public leader, and in politics it is rare enough to function as a differentiator. Consider what it looks like. A mayor says: here is my position on charter schools, here is the strongest evidence against it, here is the study that most troubles me, and here is why I still hold it. The people who disagree now know two things — that he has read their case, and that he did not need to distort it. He has not converted them. He has done something more durable: he has made himself worth listening to next time.

Contrast that with standard operating procedure, which is to present your position as obvious and the alternative as either stupid or malicious. That has a short-run advantage — it consolidates people who already agree — and a long-run cost that arrives without warning. When you are eventually wrong about something, and you will be, you have built no reputation for candor to spend.

One more thing before the tests. Testing your worldview is not the same as holding it loosely. I hold mine tightly. The point is not to arrive at mild opinions but at opinions that have been through something, so that when they are attacked you know exactly which parts are load-bearing and which you inherited without checking. A conviction that has never been tested is not strong. It is merely untried, and those two states are indistinguishable right up until the moment they are not.

The Four Tests

Coherence comes first, because it is the cheapest and the most decisive. Does the view contradict itself? The sharpest version is self-refutation: a claim that fails by its own criterion. The classic case is a strong statement of relativism — the assertion that all truth is relative, offered as a non-relative truth. Note carefully what I said in Lesson 4.4: the careful postmodern thinkers largely avoided that formulation, so this test destroys the popular version and lands much more lightly on Foucault or Rorty. Coherence is powerful but its scope is narrow. It tells you a view cannot be true as stated. It never tells you a view is true, because a perfectly consistent system can be perfectly disconnected from reality. Every well-constructed fiction is internally coherent.

Correspondence comes second. Does the view fit what we independently know about the world? This is the test that does the most work in practice, and it is where worldviews collide with data. A worldview that entails a young earth collides with radiometric dating and ice cores. A worldview that entails human beings are infinitely malleable collides with behavioral genetics. A worldview that entails moral progress is automatic collides with the twentieth century. Every framework has to face facts it did not choose.

Explanatory power comes third, and it is the most interesting and the most easily abused. The question is not merely whether a view can accommodate the data — nearly any view can accommodate nearly anything if you are willing to add enough auxiliary hypotheses. The questions are how much it explains, how naturally, and at what cost. Ptolemaic astronomy could accommodate any planetary observation you brought it, provided you kept adding epicycles. It was the accumulating cost of the accommodation, not any single falsifying observation, that eventually broke it. So when you evaluate a worldview, watch for epicycles: the special exceptions, the after-the-fact adjustments, the this-case-is-different clauses. Then watch for them in your own.

Livability comes fourth, and I am going to be more careful with it than most people who teach this framework, because it is routinely oversold. The test asks whether the view can actually be inhabited — whether its adherents live as though it were true, or whether they systematically borrow from a rival to get through the day. This can be genuinely revealing. If a person insists that moral judgments are only expressions of preference and then reacts to a betrayal with unmistakable moral outrage rather than mere distaste, that gap is information about the view.

But here is the limitation, stated as forcefully as an opponent would state it. Reality is under no obligation to be livable. Steven Weinberg's line in The First Three Minutes — that the more the universe seems comprehensible, the more it also seems pointless — is not an argument he thought counted against his physics. A worldview that is difficult to live with might be difficult precisely because it is telling you something true that you would rather not hear. Comfort is not evidence. So livability raises a question worth pursuing; it does not settle one, and any teacher who deploys it as a knockout blow is inflating a soft criterion because it is the one his side wins. I am telling you that because I have watched people do it from my side of the table.

That ordering is deliberate: coherence and correspondence carry the most evidential weight, explanatory power carries substantial weight but requires judgment about what counts as an ad hoc addition, and livability carries the least. Anyone who reverses that ranking is doing rhetoric.

What Truth Is, and What Best Explanation Can Bear

The Four Tests assume something about truth, and honesty requires naming it, because the theories of truth are frequently misrepresented — including by people defending my position.

The correspondence theory holds that a statement is true when it matches how things actually are. Aristotle gave the classic formulation in the Metaphysics: to say of what is that it is, and of what is not that it is not, is true. Alfred Tarski developed a rigorous semantic version in the twentieth century. This is the theory ordinary people operate with and the one this academy holds. Its serious difficulty is not silliness; it is specifying what the relation of correspondence consists in, and how we could step outside our beliefs to compare them with unmediated reality. That difficulty drove some very intelligent people toward the alternatives, and pretending otherwise is unfair to them.

The coherence theory holds that truth consists in a belief's fitting into a maximally coherent system — associated with idealists like F.H. Bradley and later Brand Blanshard. Its strength is that coherence is something we can actually check; its standard problem is that more than one comprehensive system could be internally coherent.

The pragmatic theory, developed by Peirce, James, and Dewey, is the most caricatured. It does not say whatever makes you feel good is true. Peirce tied truth to the opinion fated to be agreed upon by all who investigate — an idealized limit of inquiry, which is demanding rather than permissive. James emphasized cash value in experience and was criticized for it by his contemporaries, some of it fairly. Deflationary theories, from Ramsey through Horwich, take a different route: to say a statement is true adds nothing beyond asserting it, so truth is not a substantive property needing analysis. That is a live position, not a dodge.

I hold the correspondence theory. Notice that I hold it as a philosophical position with known difficulties, not as an obvious fact only fools deny. That distinction is most of what this lesson is teaching.

Now the reasoning method. Worldview comparison is not deduction; nobody derives naturalism or theism from self-evident axioms. It is inference to the best explanation — abduction — the pattern Gilbert Harman named in a 1965 paper and Peter Lipton developed at book length: given a body of evidence, infer the hypothesis that would, if true, best explain it. Every serious argument in this module runs on it, in both directions. Draper's argument from the distribution of pain and pleasure is an inference to the best explanation. So is fine-tuning. So is the argument that the resurrection best explains the origin of the early church.

Which means you must know where the method is weak, and there are three places. First, van Fraassen's bad-lot objection, from Laws and Symmetry: the inference gives you the best of the hypotheses you happen to have thought of, and the true explanation may not be among them. Being the best available candidate is not the same as being probably true unless you have reason to think your list is reasonably complete. Second, best is doing enormous work — simplicity, scope, and elegance pull against each other and no algorithm weighs them. Third, underdetermination: the Duhem-Quine point that any body of evidence is compatible with multiple theories, and that we test hypotheses in bundles, so a failed prediction never tells you unambiguously which assumption to abandon.

None of that makes the method useless. It is how juries reach verdicts and how physicians diagnose, and we do not regard those enterprises as arbitrary. It does mean the honest output of a worldview comparison is this is the best account I have found and here is why, not this has been proven. If you hear a Christian claim mathematical certainty for a metaphysical conclusion, he has overstated it. If you hear an atheist claim the same, so has he.

Where the Boundary of Science Actually Runs

This is the section I most want you to get right, because it is the one most abused in public argument by both sides, and a leader who gets it wrong will be embarrassed by someone who did not.

Science adjudicates empirical claims. If a worldview entails that the earth is ten thousand years old, radiometric dating and ice cores bear directly on it. If it entails that prayer produces measurable physiological outcomes under controlled conditions, that is testable and has been tested. If it entails that human behavior is entirely a product of environment, behavioral genetics is relevant evidence. Where a worldview makes a claim about the observable world, science is the right instrument and there is no religious exemption from its verdict.

Science does not adjudicate whether the natural world is all there is, and the reason is technical rather than rhetorical. The sciences operate under methodological naturalism — a working rule that inquiry proceeds by seeking natural causes. That rule is enormously productive and I have no quarrel with it. But it is a rule of the practice, adopted at the outset, not a result the practice produced. A method that only looks for natural causes will only ever find natural causes; that is what it was built to do. Concluding from its success that only natural causes exist converts a finding about method into a claim about being, and that conversion is a philosophical move requiring a philosophical defense. Serious naturalists know this. Oppy does not argue for naturalism by pointing at a laboratory; he argues from the theoretical virtues of total worldviews, correctly conducted on philosophical ground.

The same boundary constrains my side, and here I want to be blunt, because this is where Christians most often overclaim. Fine-tuning arguments are real philosophical arguments and I take them seriously. They are not scientific proofs of God. The physical constants are physics; what their values imply is metaphysics, and the multiverse and anthropic-selection responses are live, held by physicists on their merits rather than as evasions. Likewise, the absence of a complete origin-of-life account is a genuine open scientific question, and building an argument for God on the gap is a bad investment — because an argument whose strength depends on what remains unknown gets weaker every time something becomes known. If your case for God is a mortgage on future ignorance, you have a bad mortgage.

A related boundary on the other side recurs constantly. Cognitive and evolutionary accounts of why humans form religious beliefs — agency detection, the cognitive byproducts hypothesis — are legitimate research programs. What they cannot do by themselves is settle whether those beliefs are true. Explaining the origin of a belief is not refuting it; the reasoning that says otherwise is the genetic fallacy. That cuts symmetrically: an evolutionary account of why humans form beliefs about mathematics does not refute mathematics.

The demarcation problem sits underneath all of this. Karl Popper proposed falsifiability as the line between science and non-science, and it remains a useful heuristic. Larry Laudan argued in a well-known 1983 essay that no proposed criterion has succeeded in drawing a sharp boundary, and the debate is unresolved. So be careful with the phrase that is not science as though it ended a conversation. It usually does not, and someone in the room will know it.

The practical rule for a public leader: when you make an empirical claim, defend it with empirical evidence and update when the evidence moves. When you make a metaphysical or moral claim, say plainly that it is one and defend it with reasons appropriate to its kind. Dressing a metaphysical commitment in scientific clothing works only until you meet someone who knows the difference — and in New York, at a press conference, you will meet that person on a Tuesday.

Running the Tests on My Own House

Now the promise comes due. Here is my position under the Four Tests, stated as an opponent would state the problems, with no softening.

On coherence, the recurring charge is that the classical attributes of God — omnipotence, omniscience, perfect goodness — generate internal tensions, most sharply between exhaustive divine foreknowledge and genuine human freedom. I find the theistic responses adequate. Adequate is not decisive, and honest people remain unpersuaded here.

On correspondence, the burden is concrete: my worldview makes historical claims that could in principle be falsified, and it stands or falls with them. Paul says as much — if Christ has not been raised, the faith is worthless. That is a remarkable thing for a religious founder-document to concede, and it puts the claim on testable ground rather than protecting it.

On explanatory power I think my position is strongest — a personal God accounts economically for a cluster of facts that otherwise require separate stories: that the universe is intelligible to mathematics, that conscious experience exists, that moral obligation feels binding rather than preferred, that human beings across every culture behave as though persons have a worth independent of their usefulness. I hold that as a considered judgment, not a proof, and Oppy would respond that a smaller theory explains all of it well enough. That is precisely the argument to have.

And here is where I carry the heaviest burden, stated without a bow on it. First, the problem of evil — not the abstract version, the concrete one. A child dies of bone cancer over eleven months. Six million people are murdered by an industrial process. If God is able to prevent this and good enough to want to, the fact that he does not is the hardest fact my worldview has to hold. There are responses I take seriously: free will defenses, soul-making accounts, skeptical theism's argument that our failure to see a justifying reason is weak evidence that none exists, and Marilyn McCord Adams' insistence that horrendous evils are answered not by a general theory but by a God who enters the suffering and is crucified inside it. I find that last one the most honest. I also find that none of them makes the fact stop hurting, and I refuse to teach you a version where it does.

Second, divine hiddenness. Schellenberg's argument is not about defiant unbelievers; it is about people genuinely open, who have looked, and who have not found. I know such people. Appeals to the value of a free response, or to a God discovered rather than imposed, are real answers and not fully satisfying ones. Third, religious diversity. Billions of sincere, morally serious people hold incompatible religious convictions, and the strongest form of the problem is not that they disagree but that religious belief correlates so heavily with birthplace. John Hick built a pluralist interpretation of religion on that observation. I do not accept his conclusion, because exclusive truth claims cannot all be true without emptying each of content. But the observation is real, and my position has to live with it rather than around it.

So why do I still hold it? Because when I run the Four Tests across every option I have examined, this is the account that explains the most with the fewest epicycles and has not been defeated by the objections I just made against it in my own words. That is not certainty. It is a considered judgment that has been through something, and it is the only kind worth having.

Take that into your own leadership, because this is not finally about apologetics. Every executive holds convictions about markets, about human nature, about what the organization is for. Name the strongest objection to one of them. Find the person who holds it sincerely and ask them to make it as hard as they can. Then tell your team what you found, including the part that did not go your way. That practice will cost you a few arguments and build something no argument can: a leader whose conclusions can be trusted because his method can be inspected. Lesson 4.6 takes that method into the public square, where the pressure is not to be honest but to be useful.