On the Ontological Status of Scientific Entities: Empirical Reality without Metaphysical Truth

On the Ontological Status of Scientific Entities: Empirical Reality without Metaphysical Truth

The status of scientific entities, such as atoms, species, and minds, has been the subject of a perennial debate in the philosophy of science; on one side are the scientific realists, and opposing them, the antirealists. Scientific realism contends that how we casually interact with observable and unobservable entities provides evidence to justify their genuine existence. Particularly, realism asserts that the entities produced by successful scientific theories are approximately true. Conversely, antirealists are skeptical about the reality of unobservable entities, especially considering the long history of accepted scientific theories that have eventually been discredited. While realists assert that both observables and unobservables are real, antirealists only believe in the reality of observables, which can be empirically verified. The difference between realism and antirealism is their belief in the ability of science to ascertain metaphysical truth; that is, whether they believe that scientific entities can approximate any amount of truth about a metaphysical realm beyond our experiences.

Scientific entities possess a form of reality, but it is a reality that exists within the constraints of what we can empirically observe. In other words, scientific entities are real to the extent that we can experience them; however, we have no reason to believe they exist independently of how we conceptualize them. Here, ‘real’ indicates that entities logically interact with observable conditions within a scientific system of beliefs. However, while there is an empirical reality to scientific entities, metaphysical truth can never be extrapolated from these entities. The following points will be discussed to illustrate why scientific entities can only claim empirical reality and not metaphysical reality: (1) highly semantical language concerning “truth” and the public assertibility conditions that accompany this language has no bearing on metaphysical truth, (2) no amount of empirical evidence can ever prove that something is metaphysically true, and (3) despite what the No-Miracles Argument suggests about the link between success and truth, we cannot discern the reality of entities outside of the successful theories which have articulated them.

            To understand the status of scientific entities, one should examine how scientists employ their words to produce scientific truths. Namely, we must consider what aspects of the language used in scientific discourse can provide evidence of a correspondence between scientific entities and their purported metaphysical realities. For these entities to be metaphysically real, we must show that how we have theorized them corresponds to the metaphysical realm beyond what we can experience. Naturally, one may be inclined to reason that scientific entities are real because they correspond with the facts that support specific scientific theories. For example, one can reason that species are real because the system of classification they propose is consistent with the facts that support the Theory of Evolution. However, in instances like this, caution is warranted when using highly semantical terms such as “fact,” “real,” “true,” and “knowledge” to justify the existence of scientific entities. These words lack inherent meaning beyond what we have collectively assigned to them through public or scientific assent. Ian Hacking recognizes this when he says one “should not explain why some people believe p by saying that p is true or corresponds to a fact, or the facts” (Hacking, pg. 81).

For example, the cosmologist who asserts that “the entropy of the Big Bang was remarkably low” is making a true statement only if the entropy of the Big Bang was indeed remarkably low. This cosmologist may provide various reasons for believing this hypothesis. However, something they should not offer as evidence of their claim is that “it is true that the entropy was remarkably low because of X facts.” Mainly because our determination of what is “true,” “fact,” or even “remarkably low” is entirely based on collective decisions that have defined the meanings of these terms. Extending this rationale to scientific entities, just because science can construct a coherent theory based on what we have deemed empirical fact does not automatically confer truth or reality upon that entity. When asserting that an entity is approximately true, one is not comparing the entity to a metaphysical reality we cannot otherwise detect. Instead, one is saying that the entity is true to the extent that it aligns with the agreed-upon conditions for asserting that it is empirically true. Metaphysical truth does not factor in what we consider fact, truth, or knowledge because it is necessarily removed from our world experience. To this point, Thomas Kuhn writes that to assert the reality of an entity merely requires that one relate the entity, by resemblance and by modeling, to “one or another part of the scientific corpus which the community in question already recognizes as among its established achievements” (Kuhn, pg. 81). So, to claim that a scientific entity like species is real only implies that its reality is contingent upon the conditions required to produce empirical evidence for evolution within the scientific community. The language used in scientific discourse and the assertibility conditions that underpin scientific facts are constructed with scientists, their community, and their empirical observations in mind—not metaphysics. Thus, what we hold as fact and knowledge of our material world ultimately provides no evidence of a direct link between our words, how we theoretically represent them, and metaphysical reality.

While we can empirically demonstrate the reality of entities by observing their behavior and noting their reliable consequences in the material world, it is imperative to recognize that no empirical evidence can prove the metaphysical reality of what we label an entity. Empirical evidence cannot determine these entities’ reality or degree of truth despite being regarded as vital to scientific inquiry. Sir Karl Popper is one challenger of the inductive method for making empirical observations. Popper postulates that no amount of empirical evidence can definitively establish the truth of a scientific theory. He says, “no scientific theory can ever be deduced from observation statements,” nor can it be described as a “truth-function of observation statements” (Popper, pg. 40). He describes truth-functions as atomic propositions or ‘atomic facts’—simple statements that outline the possible states of affairs that could, in principle, be established or rejected by empirical observation (Popper, pg. 39). While Popper aligns himself with metaphysical realism by advocating for actively seeking truth in science, indeed ‘absolute truth’ in science, he paradoxically rejects the idea that scientific theories can ever undergo verification as true.

Like Popper’s ideas, the underdetermination thesis complicates science’s dependence on empirical evidence to establish metaphysical truth. The thesis proposes that empirical evidence cannot exclusively ascertain the reality of a scientific theory, arguing that multiple theories can be used to explain a single set of empirical data. The underdetermination thesis highlights the challenges posed by theories that are evidently indistinguishable and empirically equivalent. To be empirically equivalent refers to a situation in which two complete theories explaining a singular phenomenon are empirically valid yet logically incompatible. Because both theories function logically, we cannot know which of the two (or possibly infinite) theories is correct. Empirically equivalent theories pose a challenge for affirming the existence of scientific entities because they reveal that, at their core, theories are devoid of approximations of truth. These theories can be equivalent because, rather than express approximations of truth, they express consequences through purely empirical language. Thus, evidence gathered empirically cannot provide certainty about reality beyond pure observation, showing that scientific entities cannot be real outside these observable limits.

The No-Miracles Argument (NMA), frequently employed by scientific realists, posits that the consistent success of science implies the metaphysical truth of accepted scientific theories and their associated entities. The NMA suggests a link between the success of scientific theories and their degree of truth in the metaphysical realm. However, the NMA falters in making this link for several reasons. For one, history has shown us numerous examples of scientific theories that, while initially considered highly successful, were eventually proven incorrect. For instance, despite the success of Newtonian physics in describing physical phenomena, we now know that his laws do not work in all circumstances. As a result, Newton’s model was ultimately surpassed by Einstein’s more successful quantum mechanics and relativity theories.

Beyond historical examples of false theories, the NMA does not acknowledge that plenty of intentional falsehoods are littered throughout successful theories in contemporary science. Scientists routinely incorporate elements that are known to be incorrect in their models as a way of simplifying them. In astronomy, for example, when modeling the interiors of stars, scientists make simplifying choices in their equations regarding how they represent factors like density gradients, gas ratios, the effects of radiation pressure, and the ideal behavior of gases. Although how these factors are described in models and equations differs from what they might observe in experiments, scientists still use them for their convenience in model-making and computation. The sticking point here is that the success of any given scientific model does not necessarily imply its truth or metaphysical reality. Many successful scientific models and entities contain false claims. Therefore, the NMA’s attempt to establish a direct link between success and truth is implausible. If scientists can produce successful scientific theories using premises that contain inaccuracies, it is because the indicator for success in science is not truth but that a theory works. Likewise, the indicator for error is not falsehood but a failure to establish and maintain a reliable link between a theory and a successful prediction. Another way to put this is that there is one sort of correspondence that scientists do indeed check. “This is not the correspondence of the theory with reality, but the correspondence of the theory with itself” (Bloor, pg. 38). Once a theory is established, what matters most to scientists is not that the theory can correspond to reality but that it can reliably produce results consistent with itself. Thus, the empirical reality of scientific entities depends solely upon the success of a theory that can reliably predict the entity’s behaviors.

            If truth cannot explain the success of scientific theories that deal with atoms, species, and minds, then what does? One explanation provided by Van Fraassen says, “Any scientific theory is born into a life of fierce competition, a jungle red in tooth and claw. Only the successful theories survive…” (van Fraassen, pg. 40). However, the idea that theories are successful because they survive against all other rejected hypotheses does not explain why a particular theory succeeds above the rest. Consider the statement, “Crows are good at flying.” Suppose we want to explain why crows are so successful at flying. One way we might explain the success of crow flight is: (1) we specify a wing structure X and show that this structure exemplifies aerodynamic properties; (2) we show how X evolved in crows through natural selection; (3) we conclude that because crows have wings of structure X, then X explains their success at flying. Applying this analogy tracks the realist’s explanation for the success of scientific theories and their entities: (1) they specify a property of scientific theory X, namely truth, which leads to success; (2) they show how X survived through the selective process of science; (3) they conclude that selected theory X is true. The explanation for the success of the crow’s flight, or the success of a theory, is only acceptable if we can provide irrefutable evidence that crows have wing structure X. This is the most basic premise upon which we can soundly explain the success of crow flight. However, the realist’s reliance on truth to explain theory success does not meet this minimum requisite. Rather than offering a direct link between truth and success, the realist treats truth as evidence for success. Thinking back to Hacking’s point, he says, “We cannot explain some people believe p by saying that they do so because p is true...” (Hacking, pg. 82). It is not sufficient to say that a theory is successful because it is true or that it is true because it is successful. Success alone cannot serve as an argument for the truth or reality of scientific theories and their associated entities. The complex and nonlinear relationship between success and truth implies the need for an agnostic stance regarding the ability of science to make metaphysical approximations of truth. We must confine our talk about truth to what we can empirically observe and prioritize what is most helpful in advancing scientific pursuits.

Science grapples with an unattainable goal in the quest for absolute metaphysical truth. To reject the notion that scientific entities can ever approximate absolute truth does not come without consequences. Suppose we accept that science can never determine the reality of an entity beyond what we can empirically observe. In that case, questions about the meaning of endeavors scientists have taken up throughout history rise to the forefront of our minds, threatening to render their discoveries futile in understanding any absolute external truth about the world. Suppose science can never discern the reality of entities outside of the means of language, collective corroboration from scientists, empirical evidence, or successful theories. What has the work of science accomplished? Popper argues that science should aim to continuously attempt to grasp significant truths using “powerful and improbable theories” (Popper, pg. 58). The most potent theories offer high corroboration but low probability of truth. However, despite the allure of improbable theories, one should remain unconvinced that the primary goal of science is to uncover a singular, convergent truth. Instead, we should recognize that science excels at discovering consistent patterns in what it can empirically observe, suggesting that it should aim for a more practical purpose than absolute truth. The goal of science should be to find pragmatic solutions to the world’s problems. If scientific entities can contribute to tangible progress in public health, climate change, and resource distribution, then there is no reason to deny their reality in concrete, non-abstract terms.

Citations

Hacking, Ian. The Social Construction of What? Harvard University Press.

Kuhn, Thomas. The Structure of Scientific Revolutions. 3rd ed., University of Chicago Press, 1996.

Popper, Karl. “Science: Conjectures and Refutations.” CONJECTURES AND REFUTATIONS: The Growth of Scientific Knowledge, Basic Books, 1962.

Van Fraassen, Bas C. The Scientific Image. Clarendon Press, 1980, http://epistemh.pbworks.com/f/2.+Oxford.University.Press.USA.The.Scientific.Image.Okt.1980.pdf.