Justification and discovery – Methods in the context of discovery

Contexts of justification and discovery

Hans Reichenbach, who was close to logical positivism, distinguished between context of justification and context of discovery. The discovery context relates to the process that leads to proposing a theoretical result, while the justification context relates to the verification of the truth of a given theory or hypothesis, regardless of how it was obtained.

Reichenbach writes that “there are no logical rules in terms of which a ‘discovery machine’ could be constructed which takes on the creative function of genius”, thus meaning that only the context of justification can be adjudicated. a methodological analysis, while the context of discovery remains beyond the scope of such an investigation.

This question of the profound unity of the method, and therefore of science, is still the subject of discussion today. But everyone agrees, both among analysts and actors of science, that there is no general “recipe” that researchers would follow or should follow to produce new knowledge.

However, we can identify in scientific activity different methods applicable depending on the situation, both in the context of justification and in the context of discovery.

It should also be emphasized that the very distinction between the context of discovery and the context of justification is the object of criticism. However, it offers a conceptual framework for thinking about the scientific method.

Methods in the context of discovery

The main methods mobilized in the context of discovery are experimentation, observation, modeling and today numerical simulation, which are found to varying degrees in most scientific disciplines.

Observation

(Scientific observation goes through instruments, here stills for chemistry.)

Observation is the action of attentive monitoring of phenomena, without the desire to modify them, using appropriate means of investigation and study. Scientists resort to it mainly when they follow an empirical method. This is the case, for example, in astronomy or physics. It is a question of observing the phenomenon or the object without distorting it, or even interfering with its reality. Some sciences take into account observation as a separate explanatory paradigm because it influences the behavior of the observed object, such as quantum physics or psychology. Astronomy is one of the scientific disciplines where observation is central.

Karl Popper argues that observation can never be the first step in any knowledge building project, including in science. Science never begins with observation, but always with prejudice, hypothesis, theory.

There can be no observation that is “pure” of facts, Popper argues, and all observation is always, a priori, “tainted” by some theory, or even unconscious bias. He spoke at great length in his work on this problem of the primacy of theory, and this thesis constitutes one of the central elements allowing him, not only to demonstrate that the inductive method, or more generally, “induction” , is only a myth in the theory of knowledge, (no living creature, could learn or form any knowledge by the inductive method, and this, “from the amoeba to Einstein”; K. Popper, Objective knowledge), but also to assume the “murder of logical positivism”, as he writes in his book The Unfinished Quest.

Experimentation

In Leviathan and the Air-Pump: Hobbes, Boyle, and the Experimental Life (1985; 1989), Shapin and Schaffer analyze the birth of the experimental method.

Experimentation is also an instrument at the service of discovery. Certain types of experiments, called crucial, allow, according to Francis Bacon, to invalidate or confirm a hypothesis (Novum Organum, book II, aphorism 36). According to this experimental method, we imagine a hypothesis before the experiment itself, then we put it to the test, in order to verify or invalidate it.

Coming from physics, extended to chemistry and other experimental sciences, this method was the subject of an attempt to adapt it to medicine by Claude Bernard (1866). However, as far as the life sciences are concerned, in particular biology and medicine, these come up against the challenge of a multitude of parameters which are difficult to isolate, and whose, moreover, the isolation even takes us away from natural reality. In all the experimental sciences, the laboratory indeed plays a role of purifying experience: experimentation is thus distinguished from experience, in that if the latter is natural (thus giving rise, in terms of the philosophy of knowledge, to empiricist doctrines), that one is artificial, or constructed (see for example the experiments of Galileo on the fall of bodies). Experimentation requires a prior theory that can help to formulate it.

Thus, before the actual experiment, one looks for a hypothesis which could explain a determined phenomenon. We then develop the experimental protocol which makes it possible to carry out the scientific experiment which will be able to validate, or not, this hypothesis. Depending on the results of this experiment, we will validate, or not, the hypothesis.

This apparently simple scheme remained in force in Bacon’s experimental sciences until the 20th century, when it was questioned by some (Pierre Duhem in 1906). Indeed, according to Quine’s famous article, Two Dogmas of Empiricism, there is no “crucial experiment” that can confirm, or not, a scientific statement. Quine supports a holistic position, which does not deny any role to experience, but considers that it does not relate to a scientific statement, or hypothesis, in particular, but to the whole of scientific theory. Also, each time an experiment seems to give the lie to one of our hypotheses, we in fact always have the choice between abandoning this hypothesis, or keeping it, and modifying, instead, another of our scientific statements. The experiment does not thus make it possible to invalidate or confirm a determined hypothesis, but imposes a readjustment of the theory, as a whole; and we always have the choice to make the readjustment we prefer:

”Any statement can be held true come what may, if we make drastic enough adjustments elsewhere in the system. Even a statement very close to the periphery can be held true in the face of recalcitrant experience by pleading hallucination or by amending certain statements of the kind called logical laws. Conversely, by the same token, no statement is immune to revision. Revision even of the logical law of the excluded middle has been proposed as a means of simplifying quantum mechanics; and what difference is there in principle between such a shift and the shift whereby Kepler superseded Ptolemy, or Einstein Newton, or Darwin Aristotle?”

Modelization

Many sciences of nature (physics, chemistry), of the Earth and of the Universe (in particular: astrophysics, seismology, meteorology) are largely based on the use and/or on the development of models followed by their confrontation with observations of phenomena. The modeling activity consists in simplifying a complex reality to describe it and to be able to use the laws on the elements thus modeled. For example, the fall of an apple can be described by modeling the apple by a material point. This is an important simplification that eliminates the fact that the apple has a shape, a color, a chemical composition, etc., so much information that cannot be integrated into Newton’s laws.

In Earth science, physical modeling can consist of using another physical phenomenon than the one observed, but which would correspond to it sufficiently for the application of the laws on the model phenomenon to describe with sufficient relevance the phenomenon studied. For example, the fall of a meteorite on a planet can be modeled by the fall of a ball on a sand surface.

The model is not only what the scientist uses, it is also what the scientist produces. Models are not in nature, they have been constructed. They is a transition between observation and the elaboration of the theory.

Occam’s razor makes it possible to compare several models.

Numerical simulation

Simulation is the artificial reproduction of the operation of a device, machine, system, phenomenon, using a model or computer program, for the purpose of study, demonstration or explanation. Numerical simulation uses a specific program or possibly a more general software package, which generates more flexibility and computing power. Aircraft flight simulators, for example, are used to train pilots. In fundamental research, simulations, also called numerical modeling, make it possible to reproduce complex phenomena, often invisible or too tenuous, such as the collision of particles.

Analogy

Analogy must be understood in the strict sense of proportion (A is to B what C is to D) or in the broad sense of resemblance, similarity between beings or events, between properties or relations or laws. Some epistemologists (Hanson) argue that Kepler made his great astronomical discoveries by analogy. Mars has an ellipsoidal orbit, Mars is a typical planet (we can observe its retrogrades and movement from Earth), so all planets probably make ellipses. There would be here, not a generalization, but an analogy: the hypothesis linking A (Mars) and the Bs (planets) will be of the same type as that linking the Cs and the Ds because the Cs are D. (Hanson , “The logic of discovery”, in Pierre Jacob, From Vienna to Cambridge).

Indeed, an analogy between several physical quantities makes it possible to reuse the results of a scientific field already explored.

(Includes texts from Wikipedia translated and adapted by Nicolae Sfetcu)