Reproducibility Crisis, Part 2: Scientific Realism and Antirealism

Antirealism, Scientific Antirealism


In his epitaph for Sir Isaac Newton, Alexander Pope wrote: “Nature and Nature’s Laws lay hid in Night: God said, ‘Let Newton be!’ and all was light.” The scientific pursuit was once (and for some still is) religious in nature. Examining everything as miniscule as a bacterium or as vast as the cosmos was a means of understanding the universe, and in turn God. Findings were presented as ‘truths’, as they were found to be exact products of creation and expressions of the divine will. Science was effectively a lens into the mind of god. As the philosopher of biology Phillip Kitcher explains: “To seek the laws of nature [was] thus to reveal the divine rulebook, and to rejoice in the wisdom and beneficence of God.”

The Scientific Method 

Important in this conception of science is the necessity of the ‘scientific truth’. The scientific method, when correctly applied, was thought to produce truths about the nature of the universe as if they were straight out of the Book of Genesis. These truths were exact. They were explanatory in nature. They were divine.

Over the past hundred years (though definitely for even longer), this line of thinking has lost much of its popularity among the scientific establishment. Skepticism has seemingly overtaken belief as the fundamental prerequisite for scientific research. Though a notion of scientific truth could exist in a skeptical worldview, it loses much of its explanatory appeal.

The Fall of the Newtonian Physics and Antirealism

Moreover, several philosophers of science have been distancing themselves from a strong notion of scientific realism—our experiences and observations, according to them, do not necessarily represent ‘reality’, but rather just our personal or societal interpretation of what ‘is.’ This newly popular conception of our relationship with reality, antirealism, exists in several forms that all highlight the inherent epistemic limitations of scientific knowledge.

The historical antirealism advocated for by Laudan is rooted in the fall of the Newtonian physics (Lakatos notes scientists before Einstein “believed that Newton had deciphered God’s ultimate laws”). The systematic undoing of Newtonian physics in the first half of the 20th century by Einstein’s relativity and the general advances in quantum theory led to the unravelling of the most confirmed scientific theory in human history, and necessitated a radical reconsideration of the nature of scientific knowledge.

Religious Rhetoric

These new doubts in the epistemic nature of science coupled with a move from a philosophy traditionally rooted in religious rhetoric have led to the decline in popularity of notions of scientific truth within the philosophical and scientific establishments (some religious people though have integrated antirealism into their worldview).  An interesting note is that one of realism’s biggest advocates in recent years, Hilary Putnam, used anti-miracle rhetoric to establish the necessity for scientific realism—realism, he claims, “is the only philosophy that doesn’t make the success of science a miracle.”

We are now left with a science that may or may not (depending on your particular point of view) correspond with reality. Science may be a vehicle for establishing what actual ‘is’ (or, according to convergent realist, gradually approaching the knowledge of what ‘is’), or is a human representation that in no way necessarily corresponds with what ‘is.’

The antirealism movement has changed what we mean when we describe the ‘exactness’ of science. There is a limit to both the exactness and explanatory power of ideas that are entirely human constructs, as they exist in a different plane from what actually ‘is.’

The Intersection

So what does this all have to do with reproducibility?

I argue that the outrage over the reproducibility crisis should be tempered by philosophical skepticism in the absolute explanatory power of science. That is to say, we have to start expecting less from science. The ‘we’ may not be the scientific establishment, as this philosophical skepticism has somewhat been woven into the fabric of scientific discourse in recent years. The ‘we’ are the general public. The ‘we’ are the people reading articles entitled Nearly All of Our Medical Research is Wrong orEverything is Crumblingand wondering what suddenly happened to the scientific process (I apologize if you came to that conclusion from my previous post).

It is important to note, as I did in Part 1, that there any many important, external factors to reproducibility. Systematic changes and advances in technology will surely lead to major improvements if applied correctly.

What we must acknowledge is that, from the antirealist perspective, there are absolute limits to the explanatory power of our scientific knowledge. That is not to say that we should stop exploring or stop composing complex mathematical models, but rather that we should understand the theoretical limits of our scientific pursuits. Layers of complexity are lost when viewing a system from the perspective of a model as opposed to as it ‘is’. Variables can easily be neglected while others are given inappropriate weight.

Inexactness does not imply incorrectness, but rather that it does not wholly represent reality. 

Irreproducibility bothers us because the system appears to be broken. We lose the ability to trust scientific disciplines when time after time we see highly regarded studies unable to hold up to the scrutiny of reproduction.

Some degree of irreproducibility, however, is inherent in the scientific process. According to antirealists, we study physical, biological and psychological phenomena indirectly through a set of human constructs. The statistical analysis of a psychological study may allow for the assigning of significance, but they do not allow for the absolute determination of causation. Our understanding of relativity and quantum mechanics has undermined the simplicity of many physical truths held dear since the time of Newton. We construct models of increasing complexity, but nonetheless cannot map reality to any degree of exactness.

So What Now?

The science writer Christie Aschwande wrote an article last summer titled: Science Isn’t Broken, It’s Just a Hell of a Lot Harder Than We Give It Credit for. Though her frequent use of the word “truth” would make any antirealist cringe, the basic premise of the article lends to a very difficult conclusion—we need to change how we talk about science.

We frequently still talk about science with the confidence of a religious realist. We need more uncertainty in our rhetoric. The possibly indirect way in which we relate with the Universe must not only be acknowledged but wholeheartedly confronted. When we see that a psychological study is shown to be irreproducible, we may despair, but we also must realize the inexactness of the study even if it had been reproduced. Inexactness does not imply incorrectness, but rather that it does not wholly represent reality.

A big problem is that a change in rhetoric won’t make us feel any better. We were once outraged that the scientific system flawed, and now we will just be outraged that we are being told that our expectations for the explanatory power of science are unrealistic. It’s going to be hard. This change in rhetoric is basically taking the word ‘science’ and stripping it of many of the connotations we have all intuitively associated with it.

Nevertheless, we have to give it a shot.

Part3 of the ‘Reproducibility Crisis’ Series

  • author's avatar

    By: Wade Miller

    Wade Miller is a senior chemistry major at the University of Pennsylvania. In addition to his interest in drug development and pharmacology, Wade has studied both the philosophy and history of science. Wade is also member of the Philadelphia branch of Neuwrite, a collaborative writing group for scientists and writers.

  • author's avatar