What is science?

What exactly science is far from unanimous. If we bothered to ask the question “What is science?” to a group of scientists, is to bet that diverse and sometimes even irreconcilable answers will be collected. Such a situation should not move us as much as to draw our attention to the fact that it is the consequence of a reality that should be taken into account, that is, that science is a business with which it is not acute or well defined. .

This reality is manifested on at least two levels. On the one hand, “disciplinary”: at a certain point in its history, science is presented in a thousand and one varieties with different standards, methods and questions, from particle physics to sociology through the biology of the evolution or geology. On the other hand historical: for a given discipline, science is dynamic and changing. The current mechanics are certainly not that of Newton, nor that of Aristotle.

The need (however) for a definition

Given this situation, could we not immediately describe the project itself as defining science as useless? In a less radical way, we could not try to grasp it, if not through a precise definition, at least through a “family resemblance” by the point of reference of which we would consider as scientists all these activities that are ‘look a bit like, but without? can it ever be considered a common archetypal science?

Walking these “deflationary” paths is not without consequences, especially in the age of infodemias. Just as in a party without entry restrictions anyone can enter – or in any family photo an impostor can enter – refusing to put your finger on a concept of science allows you to count it as to business “sciences” that we would be reluctant to. consider as such. Refusing to define science is creating a hospital context with harmful versions of antiquity.

Find the least common denominator

In order to delimit the field of science and thus eliminate undesirable elements, a strategy would be to specify the least common denominator among those activities that we would intuitively consider scientific. In this sense, it would be acceptable to recognize that ancient astronomy, Darwinian theory of evolution, bioclimatology, and quantum gravity share, despite their obvious differences, a common ambition, that is, to generate knowledge.

Leaving aside the thorny issues related to the scientific method (considered here as everything that leads to the generation of knowledge), the purpose of science (which is reduced to everything that can serve the generation of knowledge, such as representation, l explanation or manipulation of facts), as well as finally the limits of scientific knowledge (here restricted to the “factual” dimension of the world, thus excluding from the outset other dimensions that would not be reduced to it), this minimalist approach is entirely empty. Far from leading to a true definition of science, it refers rather to the new question: “What is scientific knowledge?” »

A classic approach

To this question, one would be tempted to answer in the traditional way by specifying that scientific knowledge is a subvariety of true and justified beliefs, in particular those true beliefs that prove to be justified by a scientific-type method.

However, this approach is problematic. Literature is full of famous counterexamples in the form of situations that present true and justified beliefs that are not knowledge. Imagine having a raffle ticket out of a set of 10,000 of which only one will win. Once the draw has taken place without your knowledge, due to which your ticket has been recognized as a loser, can you claim to know that your ticket is a loser? Of course not, because even though you believe it, that it is true and that you have the justification to believe it (by a simple probabilistic judgment), you still have the possibility, albeit small, of having the right ticket. Now replicate this mental experiment by letting the number of tickets tend to infinity (which has the effect of making you more and more justified in believing that you have lost), you will never be able to know that you have lost. The moral of this reasoning is simple: for a given belief, there is not a degree of justification high enough to guarantee that it is true and therefore constitutes knowledge.

Beyond the fact that it has been a reason for many philosophers to deviate from the classical approach, this observation points to a problem more directly relevant to the enterprise of defining science (excluding anti-science). Because if the truth appears as a necessary condition for scientific knowledge, everything that scientists know is necessarily true, that is, always safe from proving that it is false. But isn’t this precisely the plot step that gives its very substance to the radical skepticism of “nothing can be known”? Because if scientific knowledge implies the impossibility of falsity, has not the history of science taught us — and more broadly not revealed to us our cognitive, instrumental, or computational limitations — that knowledge as it is designed is really inaccessible? To make the degree of justification of the scientific community tended to infinity, there will never be a state of science advanced enough to protect the community from the possibility, certainly tiny, perhaps even paranoid or conspiratorial, of to err, which the classical definition forbids.

A more appropriate approach

Therefore, in order not to give antiquity the stick with which to overcome it, it is up to us to abandon the idea that everything that generates true and justified beliefs of a certain kind is scientific.

To propose an alternative, let’s go back to the case of the draw. If it is not a higher degree of justification, what could turn the belief that your ticket is a loser into knowledge? Unequivocally: the elimination of the possibility that the ticket will be a winner based on certain conclusive data, for example that which would consist of observing the draw. While the absorption of this possibility is not a guarantee of truth – it is nothing, as evidenced by the fact that the draw could be fantasized or staged by potential conspirators – it is certainly a step in the right direction. correct. The challenge here is not so much to comfort yourself perfectly with the fact that your ticket is a loser (increasing the likelihood that it is indefinitely) as to comfort yourself imperfectly in the fact that it is not a winner.

In this perspective, we could argue that knowledge is about exploiting the conclusive data we have to minimize the chances of error. With a spirit close to that of the philosopher of science Karl Popper, therefore, it is not so much the responsibility of scientists to seek to touch the truth as to approach it by gradually moving away from falsehood. From this perspective, a definition of science would be:

Any business that generates knowledge (by a certain method, with a specific purpose and within certain limits), that is, any business that has as its vocation the identification and elimination of the possibilities of error.

Understood in this way, science is distinguished from other (claimed) discourses of knowledge, such as religion or politics, for which it is certainly not a constant concern to trace – and reduce – the chances of being false. . It also differs – if not in quality, at least in degree – from the generation of common sense knowledge for which the possibilities of serious error, given the context, can be legitimately ignored.

Let’s get back to the skeptical challenge

Is the proposed approach up to the challenge of keeping out of science these approaches that we would tend to call anti-science? Not quite, but actually pretty good. It is certainly useless to try to definitively suppress any form of radical skepticism regarding the possibility of authentic (scientific) knowledge, taking here the particular form of recognizing the impossibility of eliminating all and absolutely all possibilities of error.

But the definition manages to curb some rather crude forms of antiquity, and this can be a useful guide. Indeed, it authorizes not counting as science those companies that neglect to address the possibilities of error that the same scientific company has patiently updated and studied (cognitive, community or statistical biases, measurement or calibration errors, instrumentation errors, etc.). In this last aspect, the proposed definition also has the advantage of giving a privileged place to the dynamic and historical dimension of science, the evolution of which testifies to the continuous and progressive growth, over time, of the number of strategies put in place. by scientists to minimize the chances of them making a mistake.

Leave a Comment