Empirical, Logical, Testimonial: Which Kind of Evidence Fits Which Kind of Claim?

In discussions about the existence of non-physical causes or entities, the nature of evidence required to substantiate such claims often becomes a contentious issue. To assess these claims fairly, it is crucial to distinguish between different types of evidence and understand the appropriate context for each. This article explores various forms of evidence-empirical, logical, testimonial, documentary, physical, statistical, and anecdotal-and examines the standards each domain demands. It then considers the implications for claims about non-physical causes of the physical order and whether the demand for laboratory evidence is a well-formed requirement.

Types of Evidence

Evidence can be broadly categorized into several types, each with its own domain of applicability.

• Empirical evidence is derived from observation, measurement, or experiment. It is crucial in the natural sciences, but it is broader than laboratory repetition: astronomy, geology, evolutionary biology, and cosmology often test hypotheses by comparing their observable consequences with traces, patterns, and predictions.
• Logical evidence is based on the principles of logic and reasoning. Mathematical truths, for example, are established through logical deduction rather than empirical observation.
• Testimonial evidence comes from the accounts of individuals. It is significant in historical research and legal contexts, where direct observation of events is not possible.
• Documentary evidence consists of written records and documents. Like testimonial evidence, it is vital in historical and legal investigations.
• Physical evidence involves tangible objects or materials that can be examined directly, often used in forensic science.
• Statistical evidence involves data and statistical analysis. It is widely used in fields like economics, sociology, and medicine to establish patterns and correlations.
• Anecdotal evidence consists of personal stories or isolated examples. While it can be illustrative, it is generally considered weak due to its lack of systematic data collection and potential biases.

Appropriate Standards for Each Domain

Each type of evidence has a domain where it is most appropriate. For instance, mathematical claims are not established through empirical means; they rely on logical proofs. Historical claims, on the other hand, often rest on testimonial and documentary evidence, as the events in question cannot be directly observed or repeated.

In many common-law legal contexts, standards of evidence vary depending on the claim and the stakes. Criminal cases typically require proof beyond a reasonable doubt, while civil cases often use a lower standard, such as the preponderance of the evidence or the balance of probabilities. Some civil matters require intermediate standards. The broader point is that rational assessment does not use one evidential formula for every kind of claim.

Claims About Non-Physical Causes

When considering claims about non-physical causes of the physical order, such as the existence of deities or metaphysical entities, it is essential to ask what exactly is being claimed. A purely metaphysical thesis, such as the claim that the physical order depends on a non-physical ground or necessary cause, is not a laboratory hypothesis in the ordinary sense. But claims about events within the physical world, public miracles, answered prayers, or observable interventions do have empirical or historical dimensions. The evidence should match the content of the claim rather than be decided in advance by labeling the cause "non-physical."

Some arguments for non-physical causes are primarily philosophical: cosmological arguments, contingency arguments, fine-tuning arguments, moral arguments, and arguments from consciousness. These do not function exactly like mathematical proofs from undisputed axioms; they are usually metaphysical, abductive, or modal arguments whose premises and inferences can be challenged. Serious critics such as J. L. Mackie, Graham Oppy, and Paul Draper dispute particular premises, explanatory comparisons, or background assumptions rather than merely demanding a microscope slide. Testimonial and historical evidence may also be relevant when the claim concerns reported experiences or events, but then it must be assessed by the ordinary standards used for testimony and history.

Testimonial evidence should not be dismissed simply because it is testimonial. Much of history, law, and ordinary human knowledge depends on testimony. Its strength depends on criteria such as independence, proximity to the event, consistency, corroboration, absence of motive to deceive, and fit with background knowledge. Anecdotal evidence is weaker when it is isolated, uncheckable, selectively reported, or unsupported by other lines of evidence. For extraordinary or worldview-shaping claims, rational standards require especially careful checks, not automatic rejection.

Are Demands for Laboratory Evidence Well-Formed?

A blanket demand for laboratory evidence may be too narrow, because not every rational claim is a repeatable laboratory claim. Historical events, mathematical truths, moral principles, and many claims in astronomy or geology are assessed in other ways. Conversely, if a non-physical cause is said to produce regular or public effects in the physical world, those effects are open to empirical investigation. The fair principle is not "no laboratory evidence is required," but "use the kind of evidence the claim would reasonably be expected to leave behind." Philosophical scrutiny, logical consistency, explanatory power, and coherence with established knowledge remain critical criteria.

Critics of non-empirical claims often highlight the danger of using the non-physical nature of a claim to shield it from all forms of scrutiny. Professional philosophers of science generally treat Popper's falsifiability criterion as an important insight rather than a complete test of meaning or rationality. Some good science is not tested by single decisive falsifications, and some non-scientific claims can still be rationally assessable. The stronger point is that a claim should specify what would count for or against it; otherwise it risks becoming immune to evidence rather than merely non-laboratory.

Purely metaphysical claims may not be directly testable, but they can still be evaluated for logical consistency, explanatory scope, simplicity, fit with experience, and coherence with well-substantiated knowledge. Where a claim entails observable consequences, those consequences may be tested. Where it does not, the evaluation is philosophical rather than experimental. Claims that fail these tests should be treated with skepticism; claims that pass them are not thereby proven, but they may be rationally discussable.

Conclusion

Each type of evidence is suited to particular domains, and understanding these distinctions is crucial when evaluating claims. Some non-physical claims are not directly empirically testable; others have empirical or historical implications. In either case, they are not immune to rational critique. The appropriate standard is not automatically laboratory evidence, but the evidence reasonably suited to the claim: logical analysis where the claim is philosophical, historical assessment where it concerns reported events, and empirical testing where it predicts observable effects.

This discussion does not establish the truth or falsity of any particular claim about non-physical causes. Instead, it highlights the importance of applying the right kind of evidence and maintaining intellectual rigor in evaluating such claims. In doing so, we ensure that claims are held to a standard that respects both their nature and our commitment to rational inquiry.