Reality January 31, 2022 Updated: September 16, 2026 12 min read THY-REALITY

Materialism: scientific method, metaphysics, or a prison of worldview?

Science methodologically seeks natural, testable explanations. But does its success prove physicalism as an ontology? A distinction between method, data, model, and metaphysics.

In the previous article, the word materialism referred to a value orientation: the extent to which a person ties happiness, success, and self-worth to money and possessions. Here the same word means something entirely different. In philosophy, metaphysical materialism—or, more commonly today, physicalism—is a thesis about what fundamentally exists: that everything real is physical or wholly dependent on the physical.

This must be distinguished from scientific method. When explaining observable phenomena, science seeks models connected to measurement, prediction, repeatable testing, and public criticism. In practice it therefore works with natural causes and processes that can enter such testing. But a methodological rule about how we investigate is not automatically identical to a metaphysical claim about everything that exists.

The distinction looks small, but it changes a great deal. A person can consistently practice neuroscience, physics, or biology without inferring from every successful experiment that the philosophical questions about the nature of consciousness, causation, or reality are finally settled. Conversely, openness to metaphysical possibilities does not license filling gaps in knowledge with arbitrary explanations unsupported by evidence.

This article therefore examines three separate issues: why methodological naturalism is so effective in science, why physicalism is a serious and influential metaphysical candidate, and when a useful working orientation can turn into a dogma that decides in advance what reality is allowed to contain.

One word, three different meanings

The word materialism is dangerous mainly because people often use it for three different things. The first is the psychological materialism discussed in the previous article: a value orientation toward acquisition and possessions. The second is historical philosophical materialism: the view that matter is fundamental reality. The third is modern physicalism, a broader position usually claiming that everything actual is physical or depends on the physical in the relevant sense.

Why move from “matter” to “the physical”? Because modern physics no longer describes the world through the naive image of solid particles as tiny balls. It deals with fields, quantum states, symmetries, energy, spacetime, and structures that fit poorly with the everyday concept of matter. Physicalism is therefore the better term for most contemporary versions of the view.

Yet even the term “physical” is not philosophically self-explanatory. If we define it by present physics, we risk being too narrow because we know our current theories are incomplete. If we define it by a complete future physics, we do not yet know what that theory will contain. This is the core of the so-called Hempel dilemma and the first sign that physicalism is a philosophical thesis requiring further interpretation, not merely an appeal to the word “science.”

Title page of La Mettrie’s L’Homme Machine from 1748. — The title page of La Mettrie’s L’Homme Machine (1748) documents a historical mechanistic-materialist tradition in which the human being was described as a machine. It is not a statement of contemporary physicalism and does not show that scientific method itself entails materialist metaphysics; it does show that the ontological debate over the human being as a wholly physical system long predates modern neuroscience. Image: Wellcome Collection / Wikimedia Commons CC BY 4.0

What scientific method actually requires

Science needs a constraint: an explanation must be determinate enough to be compared with observation. An experiment must at least in principle distinguish expected from unexpected outcomes; methods must be describable; data must be open to criticism; and errors must have the power to force a model to change.

In this sense science works methodologically naturalistically: in explaining the natural world it seeks natural processes and causal mechanisms that can be treated empirically. The National Academies, in discussing the nature of science, emphasize testing explanations against observation and experiment and note that a supernatural explanation defined so as to escape empirical testing cannot play the same scientific role.

An important qualification follows: science does not test verbal labels; it tests the consequences of claims. If someone claims that an unknown or even “nonphysical” factor reliably moves instruments, changes biological outcomes, or produces a repeatable signal, those effects are empirically investigable. If a claim has no possible observable consequence, scientific method simply lacks leverage with which to distinguish it from its negation.

Methodological naturalism is not metaphysical physicalism

This is the central distinction of the article. Methodological naturalism is a rule of inquiry: do not stop explanation at a cause that cannot be operationalized and tested. Metaphysical physicalism is an ontological claim: ultimately there is nothing over and above the physical, or everything else is wholly grounded in it.

The first does not necessarily require the second. A physician can investigate neurological mechanisms in an epileptic seizure because those mechanisms are diagnostically and therapeutically testable, without thereby proving a complete metaphysics of the universe. Likewise, a religious believer, an agnostic, and an atheist can use the same laboratory protocol if they accept the same rules of measurement and criticism.

The National Academies have explicitly marked this asymmetry when discussing the nature of science: because science uses natural processes in its explanations, it does not follow that science teaching must advocate the metaphysical thesis that supernatural influence is impossible. Method sets the boundary of scientific explanation; by itself it does not provide a complete inventory of everything that exists.

Why physicalism is nevertheless a strong candidate

If methodological naturalism does not prove physicalism, it does not follow that all metaphysical views are equally supported. Physicalism has a powerful historical advantage: more and more phenomena once explained by separate vital, mental, or mysterious forces have received successful physical, chemical, and biological explanations without the need for an additional substance.

A particularly influential case is the argument from causal closure or completeness of the physical. In rough form: physical events have sufficient physical causes; mental states apparently influence physical behavior; and if we want to avoid systematic double causation, it is attractive to conclude that the mental is not ontologically separate from the physical. This is one of the central contemporary arguments for physicalism—but it remains a philosophical argument whose premises and exact formulation are debated.

A second reason is practical success. Physics, chemistry, biology, and neuroscience form an increasingly connected network of explanations. Physicalism promises a unified ontology in which we do not introduce a new kind of substance whenever a new phenomenon appears. That is a genuine theoretical virtue: fewer basic categories, greater integration, and fewer unexplained interactions.

Brain and consciousness: strong dependence, not yet complete identity

The strongest pressure in this debate comes from the study of mind. Brain injuries can alter memory, personality, perception, and decision-making; anaesthetics can reversibly abolish conscious responsiveness; and research on neural correlates links patterns of brain activity to particular conscious states. The empirical dependence of conscious life on the brain is therefore extremely strong.

But dependence does not by itself entail identity without an additional argument. If changing A reliably changes B, we have established an important relation; we have not necessarily established that A and B are, under every ontological description, one and the same thing. This is where the philosophical dispute about consciousness begins: can a neural description ultimately explain why subjective experience exists, or will an explanatory gap remain between third-person brain descriptions and first-person experience?

Joseph Levine popularized the term explanatory gap for this problem. Importantly, the existence of such a gap does not by itself prove a nonphysical soul. Some regard it as evidence against reductive physicalism, others as a temporary limitation of our concepts, and still others think it is compatible with physicalism. The problem is real; which metaphysics ultimately solves it is not established.

Where data end and ontology begins

A scientific result can be extremely strong while a metaphysical conclusion still requires an additional premise. If scanning shows that a particular experience reliably tracks a particular neural dynamic, that is data. The statement “consciousness is therefore nothing but that neural dynamic” is an interpretation that may be reasonable, but it is ontologically stronger than the correlation itself.

Philosophy of science adds the idea of underdetermination of theory by evidence. A body of data can sometimes permit multiple theoretical descriptions, at least until new measurements or discriminating methods become available. This does not imply relativism, and it does not mean all theories are equally good. It means we should separate what was directly measured, the model organizing the measurement, and the broader metaphysical interpretation of that model.

This distinction matters most around large questions—consciousness, the fundamental interpretation of quantum theory, the beginning of the cosmos, time, and mathematics. Science may advance rapidly in predictive and experimental power while more than one metaphysical interpretation remains open for some time.

Alternatives to physicalism are not automatically scientific theories

Once we recognize that physicalism is not logically identical to scientific method, a second mistake often appears: as though this automatically confirmed dualism, idealism, or panpsychism. It does not. Showing that candidate A is not established by the definition of a method is not evidence for candidate B.

Dualism holds that the mental and physical are fundamentally different kinds of reality; it immediately inherits a difficult problem of interaction and integration into physical causation. Idealism treats the mental as fundamental, but must explain the stability, intersubjective agreement, and mathematical structure of the world described by physics. Panpsychism attempts to place consciousness or mentality into the fundamental nature of the world, but faces the combination problem: how putative elementary forms of experience could compose a unified consciousness.

There are also intermediate programs such as Russellian monism, which notes that physics may describe structure and dynamics extraordinarily well without exhausting the question of the intrinsic nature of whatever realizes those structures. These are serious philosophical research programs. Their existence shows that the conceptual space is wider than “materialism or miracle”; it does not show that any alternative is empirically confirmed merely because physicalism remains philosophically debated.

When materialism becomes a prison of worldview

Physicalism becomes a problem only when a reasonably supported thesis turns into an untouchable definition of the possible. If every anomalous datum is answered in advance with “that cannot exist because it is not material,” evidence has been replaced by an ontological filter. The history of science contains many real phenomena observed before their mechanisms were understood. An unknown mechanism is not permission for any explanation, but neither is it proof that the phenomenon is unreal.

The opposite dogma is equally dangerous. If every gap in a physical explanation is declared evidence for a soul, consciousness as a cosmic force, or some preferred metaphysics, the same epistemic leap has simply been made in the other direction. We do not yet know is not the same as therefore my explanation is correct.

The healthiest stance is therefore asymmetrically open: use the most successful testable models without apologizing for their success, while labeling their metaphysical extensions as extensions. What is strongly confirmed should carry more weight. What is possible but weakly tested should remain a possibility—not an equal fact.

A practical epistemic test: what level is the claim on?

For large claims about reality, a simple five-step audit helps. First: what was actually observed or measured? Second: what model organizes those data and what does it predict? Third: what additional premises are required for the ontological conclusion? Fourth: what alternative explanations would generate the same or similar data? Fifth: what possible result would force us to revise the metaphysical conclusion?

This test works in both directions. A physicalist should be able to say what would count as a serious problem for the view, rather than simply relabeling every future discovery “physical.” An anti-physicalist should show what the alternative explains better, what novel predictions it yields, and how it could be distinguished from psychological preference, metaphor, or a gap in present knowledge.

This does not remove philosophy from inquiry. On the contrary, it makes philosophy more disciplined. Metaphysical questions are legitimate, but their evidential standard should not become lower merely because the questions are large.

Conclusion: science is stronger when it does not hide its metaphysics

Materialism or physicalism is not the same thing as scientific method. Science is extraordinarily successful because it demands public, testable connections between claims and data and because it does not stop investigating natural phenomena at an explanation that generates no further tests. That method strongly supports the search for physical mechanisms and has historically displaced many unnecessary additional entities.

Physicalism is therefore a rationally strong candidate, especially given the success of integrated natural-scientific explanation, the brain dependence of mental states, and arguments concerning physical causal completeness. But methodological success does not, without further philosophical premises, entail that ontology is finished. Consciousness, the meaning of “physical,” and the relation between models and reality remain serious philosophical problems.

The most open worldview is neither one that rejects materialism nor one that turns it into an unfalsifiable definition of reason. A better stance is stricter: use the method without compromise, give evidence priority over preference, and always label metaphysics as metaphysics—even when it is currently the most persuasive option.