All searches for truth can begin with a basic distinction: Is a statement analytic or synthetic?
Analytic statements are true by definition—no evidence is required.
Synthetic statements depend on observation or evidence.
Philosophers often disagree about which statements belong to which category. For example, Roger Scruton argues that mathematics is not analytic. Although “1 + 1 = 2” appears obviously true, its truth depends on definitions of “1,” “addition,” and “2.” Once those definitions are fixed, the statement becomes logically true, but the applicability of mathematics to the world still requires empirical confirmation. Pure logic may be analytic, but mathematics as practiced involves testing and verification.
By contrast, synthetic statements are grounded in evidence (a posteriori).
Example: “John needs an iron supplement.”
This cannot be known by definition; it requires medical testing.
David Hume initially treated mathematical truths as analytic, using examples like “the shortest distance between two points is a straight line.” The statement seems obviously true, but, as Kant later argued, determining whether it holds in physical space requires empirical checking. We draw lines, measure, and compare. Thus, many mathematical claims—especially those applied to the world—are synthetic, not analytic. Since science depends on mathematics, scientific claims are likewise synthetic.
Kant believed that objective moral duties (“Do the right thing”) are analytic. However, this conclusion is questionable. We often confuse simplicity with obviousness, and mistake statements requiring evidence for statements that are true by definition. Even if an objective moral order exists, that does not mean it is self-evident to us. A blind person cannot know a rock is present simply by definition; she requires evidence. Morality, if real, may be similar: it might exist independently of our capacity to perceive it clearly.
If God grounds morality, then one might ask: Why are moral truths not as obvious as analytic truths? The lack of clarity suggests either (a) morality is not analytic, or (b) God, if real, has not made moral truth immediately transparent. Either way, morality does not function as an analytic system.
Hume also argued that religious claims—such as “God exists”—are neither analytic nor supported by sufficient evidence to be synthetic. You cannot determine their truth by definition, and empirical verification is lacking. For Hume, such statements are therefore not meaningful truth-claims in the ordinary sense.
Critics note that Hume’s argument itself is not analytic and does not rely directly on empirical evidence, so it may appear self-undermining. But even if his conclusion is not absolute, it still raises a reasonable caution. If a claim cannot be verified analytically or empirically, then we have no obligation to treat it as a foundational truth. We need mathematics and empirical knowledge to survive; we do not need unverifiable religious doctrines. Religion may function socially as custom or opinion, but it often demands far more certainty than it can justify.
A final objection to Hume is sometimes raised:
If humans are too prone to error and deception to trust miracle reports,
then they are equally prone to error in their scientific observations.
But this misses a crucial difference. Science constantly re-checks, corrects, and tests its claims, precisely because humans can be mistaken. Miracle claims typically are not tested, and believers in miracles often do not apply rigorous methods of verification. The possibility of human error is taken seriously in science, but not always in miracle traditions.