What Makes Scientific Knowledge True

This paper illustrates a series of debates among realism, anti-realism and structural realism to decide whether one should believe that it is possible at all to discover any truth through science. First, it defines realism and anti-realism and shows pure realism can be “surely untenable” (149 Worrall). Second, it presents structural realism, a fusion of both realism and anti-realism, and how using it to justify realism is inadequate. Third, the paper offers its own objection to structural realism based on the Kantian notion of things-in-itself. Finally, it defends structural realism from this objection and offers a tentative answer to the original question: One can be structural realist about current scientific theories because human observation affects reality and thus our “biased” scientific theoretical structures must contain some truth because these theoretical structures affect human perception (this paper will use the term “truth” to refer to knowledge that reveals anything truthful regarding the actual nature of the universe).

Before this paper begins, it would like to discuss what it means for a scientific theory to be truthful because a clear definition of “truthful” is essential to this paper’s understanding of realism. Theoretical entities and logical forms that describe theoretical entities’ relationships amongst themselves form a scientific theory that explains phenomena in the universe. By the universe, this paper means not the observable universe, but the complete universe free of biases in human perception. If any/all of these theoretical entities genuinely exist in the universe and any/all parts of the logical form describe these theoretical entities error-freely in accordance to the laws of the universe, the scientific theory is partly/completely truthful. In other words, the scientific theory contains some/all truths (i.e. error-free knowledge about the universe).

Scientific realism, which I will refer to as realism for the rest of this paper, is the belief that “the enormous empirical success of” certain scientific theories “legitimizes the assumption that these descriptions of an underlying reality are accurate…or ‘approximately’ accurate” (139 Worrall). The entities (e.g. atoms) and their relationships postulated by such theories that outline observed phenomena should be considered real and part of the true nature of the universe, so scientific theories contain truths. On the other hand, anti-realists believe that “there is no reason to assume that even our best theories are true nor even ‘approximately’ true, nor even that the aim of science is to produce true theories” (140 Worrall). One should not believe the entities and their relationships described by scientific theories are real, so scientific theories do not contain truths.

Some realists justify realism through the “no miracles” argument (140 Worrall). It asserts that it would be miraculous if highly empirically predictive and phenomenally explanatory theories such as Newton’s Theory of Gravitation do not contain any truth about the content and structure of the universe. In contrast, anti-realists believe this argument is untenable. Although highly empirically predictive theories exist, historically, other successor theories with strong empirical evidence falsified them. A famous example is how Maxwell’s theoretical entities (electromagnetic field) replaced Fresnel’s theoretical entities (“an all-pervading, mechanical medium”) in the theory of light, even though Fresnel’s theory was highly predictive and explanatory (155-156 Worrall). This historical phenomenon in the development of scientific theories is the basis of “pessimistic induction”, the anti-realist counterargument to “no miracles” (149 Worrall). “Pessimistic induction” states that there are insufficient grounds to think that science has found any truthful theories and theoretical entities yet because many theories originally regarded as truthful were falsified. If we agree on that the aforementioned historical phenomenon is accurate, believing in pure realism and assuming current theories are true is incautious because the majority of the scientific community will reject the truthfulness of past, present and future theories and the entities they posit.

Structural realism resolves realism and anti-realism. Worrall proposes that instead of focusing on debating about the truthfulness and accuracy of the content of scientific theories, one should place one’s belief in the truthfulness of structures of scientific theories. These structures relate observable phenomena with each other, such as Newton who expresses relationships among phenomena “in the mathematical equations of his theory” (410 Ladyman). It is more rational to consider theoretical structures as candidates that contain truths than the content they encapsulate because while theoretical entities kept changing, “there was continuity or accumulation in the shift…. of form or structure, not of content” in scientific theories throughout history. Many believe structural realism has the best of both worlds because it disregards controversial assumptions in both anti-realism and realism. It leaves room for optimism about the ability of science to acquire some truth in the universe and avoids two realist assumptions: 1) we should treat existing theories as if they have identified entities and described phenomena we have observed perfectly correctly, and 2) science progresses “miraculously” and revolutionarily (409 Ladyman). One example among many Worrall uses to support the legitimacy of structural realism is the structural “continuity between Fresnel’s and Maxwell’s theories” about light because Maxwell’s theory uses Fresnel’s mathematical equations but simply changes the theoretical entity represented in the equation, but Maxwell has caused a dramatic change in people’s understanding of the nature of light (158 Worrall). Such cases demonstrate that throughout history, there is a continuity between old and new theoretical structures and a constant replacement of theoretical entities. Hence, it is arguable that there is a “cumulative growth at the structural level” of theories “combined with radical replacement of the previous ontological ideas” and theoretical entities (160 Worrall). In exchange for this viewpoint, structural realists have given up the ideal that humans can ever grasp “the nature of the basic furniture of the universe” (162 Worrall). One will never find true entities in the universe but improve theoretical structures for capturing and presenting new theories to describe relationships among observable phenomena in the universe. These theoretical structures, considering their steadfast rather than revolutionary evolution like theoretical entities, are more likely to be partly truthful.

Structural realism is a defensible child of realism. It is resistant to criticisms because “there is retention of mathematical structure across theory change which is in other respects radical” (414 Ladyman). However, it does not support pure realism. Many theoretical shifts in science may preserve relations but not their relata and logical type. For example, even though Ehrenfest’s theorem (quantum mechanics) inherited its form from the Newtonian mathematical formula for calculating force (classical mechanics), they use drastically different operators and variables because these theorems posit different theoretical entities. The content of a new theory can change so dramatically from the content of the theory’s predecessor that there is no “continuity of ontology” (415 Ladyman). Structural realism does not guarantee that humans will achieve perfect knowledge because it only places faith in the possibility of theoretical structures being truthful but not theoretical entities. But realism believes theoretical structures and entities can both contain truths. Therefore, structural realism only supports half of realism and is an inadequate argument for us to believe in realism (163 Worrall). 

To further explore the debate, this paper will move onto presenting an argument against structural realism, which still supports realism half-heartedly regardless, to attempt to eliminate any possibility in believing that science achieves true knowledge. Then, this paper will conclude with a defense for structural realism against this objection.

Is it possible that theoretical structures can never be truthful because they are created by the human mind? Immanuel Kant argues that because human perception is arbitrary and subjective, one can never understand any thing-in-itself. By thing-in-itself, Kant means what the object actually is free from human perception (e.g. humans experience time, so our conception of time affects how we perceive objects) (B52 Critique of Pure Reason). External objects excite our senses, and our mind interprets such sensations to perceive these objects as how they appear to our senses and mind, so we only perceive objects “either as appearances, or as objects of thinking in general” (B391 Critique of Pure Reason). The appearances of objects are “mere modifications or foundations of our sensible intuition” (B63 Critique of Pure Reason). If there is no human perception, “the represented object with the properties that sensible intuition attributed to it is nowhere to be encountered” because only our subjective perception “determines its form as appearance” (B63). One example of subjective human perception is our a priori (i.e. not acquired through empirical experience) representation of space, which is the foundation of how we perceive any external physical object. The first reason is that the mind can imagine a space with objects but not no space. The second reason is that without an a priori notion of space, one cannot distinguish sensations within and outside of one’s body. Because our perception has such a priori conceptions for representing reality, an object in itself “remains unknown to us” (B63 Critique of Pure Reason). 

To apply this argument to structural realism, when scientists develop logical forms for explaining relations among theoretical entities, they are not developing logical forms for objects in themselves but appearances of these objects, which are unknowable according to Kant’s argument. Because how these objects appear to us are subject to a priori notions rooted in how one perceives, 

One famous example is quantum entanglement, during which if one particle in a pair of particles is measured, the measurement of the other particle is affected instantly regardless of how far these particles are from each other. Many experiments have confirmed that this theory is likely to be true (New Scientist). Now let’s consider what implication quantum entanglement has for structural realism. The theoretical structures science has developed about most observable phenomena are based on the assumption that these phenomena occur during time A to time B where A < B, not where time A = time B and time A > time B. Quantum entanglement indicates that there are many phenomena that are yet to be observed during their occurrences during time A = B by humans directly. We physically are not equipped to do so because to observe external phenomena, our body and mind take time to experience and process our sensations. If quantum entanglements happen all the time, there exist countless instantaneous processes that humans cannot observe during their unfolding but only their consequences. Human perception is flawed and limited because of how our senses and mind are structured, so we can never experience the world fully and objectively.

It follows that as predictive and descriptive as the knowledge we obtain through science can get, it is always a distorted and limited version compared with the knowledge a more advanced being who can also perceive instantaneous events has about the world independent of human observation. Scientists observe through their biased senses and minds to create the theoretical structures structural realists hold so dear, which will always be inevitably biased and limited. Thus, we should be neither structural realists nor realists at all about scientific theories.

A structural realist defense would be that if the world perceived by human senses forms part of the universe independent of human observation, theoretical structures we have do contain some truth about the universe because the world we observe is part of the greater reality. This statement is a contradiction because a biased perception of the universe independent of human observation cannot be truthful. However, there may be no universe independent of human observation. Quantum mechanics and quantum entanglement do suggest that human observation affects reality at the quantum level (ScienceDaily). If our observation affects the actual universe, to understand the universe fully and truthfully, one must also understand the biased human perception. Since theoretical structures affect how humans perceive the world and what humans perceive, theoretical structures affect the ever-changing universe. One must understand the developing structures of human scientific theories on a constant basis to understand how these structures affect human perception and the universe. Therefore, because our scientific theoretical structures affect the universe at the quantum level and the universe’s constitution by affecting human perception, what the universe is is dependent on scientific theoretical structures. Creating these theoretical structures is also creating new content in and truths regarding the universe, so these theoretical structures must contain some truthful knowledge about the universe.

To summarize, if one agrees with “pessimistic induction”, being a pure realist is irrational because history has shown that theories regarded as true today may be falsified tomorrow. Structural realism is a more defensible position than realism because it rests on the historical phenomenon that theoretical structures develop cumulatively. Structural realism becomes an irrational position if one assumes the universe is independent of human observation because our fundamentally biased perception will always make our scientific theories deviate from true knowledge about the universe. Nonetheless, some empirical evidence suggests the universe may not be independent of human perception, so it is acceptable to be a structural realist and possible to create and discover truths through science.

Citation

“Quantum Theory Demonstrated: Observation Affects Reality.” ScienceDaily, ScienceDaily, 27 Feb. 1998, https://www.sciencedaily.com/releases/1998/02/980227055013.htm.

Aron, Jacob. “Quantum Weirdness Proved Real in First Loophole-Free Experiment.” New Scientist, 28 Aug. 2015, https://www.newscientist.com/article/dn28112-quantum-weirdness-proved-real-in-first-loophole-free-experiment/.

Kant, Immanuel, and Zöller Günter. Prolegomena to Any Future Metaphysics That Will Be Able to Present Itself as Science: with Two Early Reviews of the Critique of Pure Reason. Oxford Univ. Press, 2005.

Worrall, John. Structural Realism

Ladyman, James. What is Structural Realism

Leave a Comment

Your email address will not be published. Required fields are marked *