Version classiqueVersion mobile

L’épistémologie historique

Jean-François Braunstein
Iván Moya Diez
Matteo Vagelli

Historicités, objectivités, rationalités

Can the History of an Epistemic Norm Bear Normative Value?

Some Reflections on the Status and Tasks of Historical Epistemology

Eugenio Petrovich

Texte intégral

1This chapter focuses on the status and the tasks of historical epistemology. In particular, the main question that will be investigated is the following: can the history of an epistemic norm bear normative value in turn? This question emerges quite naturally when we consider inquiries into historical epistemology dealing with normative objects, such as the so-called epistemic norms. The classic of this kind of research is the masterpiece by Lorraine Daston and Peter Galison Objectivity (written in 2007). This book will be used as a starting point to reach the kind of theoretical problem that will be discussed in the present study.

2The chapter is structured as follows. In the first part, Daston and Galison’s Objectivity will be briefly summed up, and the theoretical problem of normativity will be pointed out. In particular, three key meanings of normativity will be discerned. In the second part, I will draw the attention to two historical moments in which the notion of normativity was at the very heart of philosophy of science, i.e. the age of logical empiricists (with a special focus on Hans Reichenbach) on one side, and the Neo-positivistic reaction to Kuhn’s Structure of Scientific Revolutions during the Sixties, on the other. In the third part, it will be discussed how philosophical arguments centered on the notion of normativity can be used as theoretical tools to provide an answer to the question raised in the title of this chapter. In doing so, a line of reasoning borrowed from Kant’s Critique of Practical Reason will be employed as a useful theoretical source. Finally, in the last part of the study, I will argue that the proper task of historical epistemology does not consist in normativity but in critique.

An Example of Historical Epistemology: Objectivity by Lorraine Daston and Peter Galison

3Daston and Galison in their work focus on the historical evolution of different epistemic norms, such as Truth-to-nature, Objectivity and Trained judgment. They describe an epistemic norm as a complex of prescriptions that tell scientists how to conduct parts of their research. Specifically, the focus of Objectivity is a very concrete part of scientific work, i.e. the making of scientific atlases. An epistemic norm advises the atlas maker on how to produce the images in the atlases by specifying the ideal towards which the images should point. The ideals that are upheld by the epistemic norms may change (and did change) throughout the history of science, but their aim is always the same: providing regulative ideals to guide scientific practice.

  • 1 L. Daston, P. Galison, Objectivity, New York, Zone Books, 2007, p. 195.

4It is important to underline the fact that epistemic norms are not mere speculations good for “pious prefaces and after-dinner addresses at scientific meeting1”. Rather, they entail very concrete consequences for the practice of research:

  • 2 Ibid.

To embrace [Truth or Objectivity] could translate into the choice between an exquisitely colored, sharply outlined drawing and a blurred black-and-white photograph, or between the image of an idealized type sketched freehand and that of a particular individual meticulously traced from a projected image2.

  • 3 Ibid., p. 321.
  • 4 Why then at a certain moment do scientists choose to follow new epistemic norms? Even if this quest (...)

5Still, one could wonder what the difference between epistemic norms and simple protocols of research could be. The difference lies in the fact that epistemic norms, as opposed to simple protocols of research, are ideal in their very essence, namely they can never be fully satisfied by real practice. As Daston and Galison point out, it is very important to underline that their proponents are perfectly aware of this. When, for instance, the advocates of Mechanical Objectivity in the Mid-Nineteenth century struggled for images untouched by the human subject, advocating an ideal of mechanical production of the images, they were very aware of the fact that the intervention of the experimenter could not be totally avoided, even in the case of photography. Nevertheless, they did not give up the norm of Mechanical Objectivity: the reason is that Mechanical Objectivity was considered an ideal aim, a sort of horizon to point to, worth of pursuing even if it could not be totally expressed in practice. As Daston and Galison claim, “for these scientists Mechanical Objectivity was a regulative ideal, a shaping ambition that conditioned whatever and when practitioners sought to improve what they did on the page, in the field, and at the laboratory bench3”. Consequently, scientists do not simply abandon an epistemic norm in favor of another just because they realize that the epistemic norm cannot be fulfilled in practice: they have always been aware of ideal nature of it4.

Meanings of Normativity

6In this part, I will focus on the concept of normativity in historical epistemology. The concept of normativity plays a key role in Daston and Galison’s argument, because it grounds theoretically the comparison between epistemology and ethics the two historians stress throughout the book. They compare the concept of epistemic norm to that of ethical virtue, pointing out that both epistemic norms and ethical virtues function as regulative ideals. This brings into the picture three consequences, which in turn shed some light on the meaning of normativity in general.

7First, both epistemic norms and ethical virtues serve to guide practice, providing the ideal fulfillment of a certain type of actions. Second, both remain valid even if the actual practices did not fully realize the ideal. Thirdly, both concepts are perfectly suited for evaluating actual practices (scientific work or ethical actions), because they provide an ideal to which the real can be compared and assessed. I will call these three key features or meanings of normativity (guide, ideality and evaluation) the regulative functions, and everything that is endowed with them will be said bearing a normative value.

8This brief analytical clarification allows to frame better the question posed at the beginning of the paper: can the history of an epistemic norm bear a normative value in turn? As now it should be clear, this problem can be rephrased as follows: given the historical account of an epistemic norm, is it possible to extract the previously mentioned three regulative functions, namely the guiding, the ideal and the evaluating functions, from it? In other terms, can the historical account provide valid ideals to guide and evaluate actual practices?

9It is worth noting that questions like this do no emerge only within historical epistemology. In fact, they are typical of every field dealing in some way with normativity-based concepts: think about classic normative disciplines such as ethics and political philosophy. In political philosophy, for instance, the three regulative functions are at the core of every normative approach since Plato’s De Republica. When a political thinker or philosopher suggests an ideal form of government or state, she is perfectly aware that the ideal form does not exist in reality and maybe will never exist. But her aim is not to describe something—in the present or in the future does not matter—but to present an ideal model both to guide the reform of actual political practices and to evaluate them. The ideal form can guide reformation because it provides a specific aim for it, and it can evaluate it too because it provides an ideal benchmark. Therefore, in political philosophy it is a standard move to distinguish between normative approaches (sharing the regulative features) on one side, and descriptive approaches (aiming simply at describing a political system without evaluating it) on the other side.

Normativity in the History of Philosophy of Science

10Ethics and political philosophy are not the only fields where normativity plays such a pivotal role. Indeed, it has played a significant role also in philosophy of science, at least in two moments of its recent history. For two times in the Twentieth century, the concept of normativity stood at the heart of the very definition of the object and tasks of philosophy of science. These two moments will now be described, analyzing the theoretical structures in which the concept of normativity was embedded. Such an analysis will provide us with conceptual tools useful to answer the question concerning the normative value of historical epistemology.

11The first historical moment that I shall recall is the dawn of modern philosophy of science in Europe, namely the times of logical empiricists. In particular, I will consider the masterpiece Experience and Prediction, published in 1938 by Hans Reichenbach, one of the leading figures of the so-called Berlin Circle (the German counterpart of the more famous Vienna Circle).

12In the first chapter of this book, Reichenbach outlines the tasks and objects of epistemology or philosophy of science. What is interesting is that he uses normativity precisely to distinguish epistemology from other disciplines that study scientific knowledge, such as psychology and sociology. According to Reichenbach, epistemology has a normative dimension that lacks in merely descriptive sciences. Even when the epistemologist pursues a descriptive task (which is the first task assigned to epistemology by Reichenbach) her description of a phenomenon of knowledge has nothing to do with a psychologist’s account of the same phenomenon. Indeed, Reichenbach claims that the epistemologist, unlike the psychologist, is not interested in a report of the actual processes of thinking that really take place in the mind or brain of the scientist. On the contrary, she aims to describe what Reichenbach calls a rational reconstruction of the thinking process. Now, the notion of rational reconstruction is defined with an explicit reference to a normative dimension. Rational reconstruction consists in displaying thought-processes in the way they ought to happen, not in the way they actually happen. As Reichenbach puts it:

  • 5 H. Reichenbach, Experience and Prediction: An Analysis of the Foundation and the Structure of Knowl (...)

Epistemology does not regard the processes of thinking in their actual occurrence; this task is entirely left to psychology. What epistemology intends is to construct thinking processes in a way in which they ought to occur if they are to be ranged in a consistent system […] Epistemology thus considers a logical substitute rather than real processes5.

13In the rest of the chapter, two other dimensions of normativity can be recognized, namely the ideality of rational reconstruction (consisting in its difference from reality) and its evaluative potential (i.e. its being an effective benchmark for actual practices). The evaluating function, in particular, is fully articulated in the second task Reichenbach assign to epistemology, i.e. the critical task. If logical reconstruction is a better way of thinking, then, Reichenbach argues, it can be used by the epistemologist as a point of reference against which real scientific practices can be compared, assessed and eventually criticized. Finally, the guiding function is presented by Reichenbach in the third task he assigns to epistemology, the advisory task. By the means of logic, conceived as a normative discipline teaching the correct way of thinking, the Reichenbachian epistemologist can advise the scientist by suggesting her the best course of thinking. In sum, within Experience and Prediction the concept of normativity, in all its three meanings (ideal, guide, evaluation), plays a central role in the definition of the object of epistemology (logical reconstruction) and in the delineation of the primary tasks of the discipline (description, evaluation and advise).

14Moreover, normativity is crucial for Reichenbach to forge another conceptual tool, that will turn out to be pivotal in the Neo-positivist conception of epistemology: the distinction between the context of discovery and the context of justification. In Reichenbach’s system this dichotomy is fundamental in restating the difference between psychology and epistemology. While the context of discovery covers the psychological reasoning of the scientist, the context of justification covers the logical foundation of scientific theories, which have an intrinsic normative value. Reichenbach explicitly claims that epistemology must focus only on the second context (the context of justification), leaving aside the context of discovery (which is matter of psychology).

15Interestingly, this same dichotomy was rediscovered and brought to new life in the Sixties by the defenders of what was once known as the Standard View in philosophy of science. This is the second episode in the history of philosophy of science I will focus on in this study. The advocates of the Standard View used the dichotomy and the difference between normative versus descriptive as powerful argumentative weapons to counter the challenges posed by the historians of science, and by Thomas Kuhn in particular.

16The publication of The Structure of Scientific Revolutions in 1962 produced a mobilization of intellectual powers within the community of Neo-positivist philosophers of science to counterattack Kuhn’s attack on Neo-positivism. One of the main strategies of this reaction consisted in the reestablishment of a normative-descriptive boundary between philosophy of science and the new enemy, i.e. history of science, via the discovery versus justification dichotomy.

  • 6 See C. R. Kordig, The Justification of Scientific Change, London, Springer, 1971 and I. Scheffer, S (...)

17For instance, the philosophers Carl Kordig and Israel Scheffer, blamed Kuhn for confusing the very basic distinction between the two contexts6. In the conclusion of the first chapter of The Structure, Kuhn famously claims that historical arguments should be relevant for philosophy of science. Kordig and Scheffer, however, replied that this could not be case, because, they argued, philosophy of science only deals with the context of justification, whereas the history of science deals with the context of discovery—and there is no possible interaction between the two contexts. Kordig and Scheffer argued that history of science is a mere description of what science is and has been in the past, whereas philosophy of science is an inquiry into what science ought to be. Since there is no logical connection between what science is and what science ought to be, it follows that nothing borrowed from history can concern philosophy of science.

18It is interesting to notice that Neo-positivists were not the only ones to raise this kind of criticism to Kuhn’s work. Even Paul Feyerabend, who was, as it is well known, very critical of Neo-positivist orthodoxy, stated his objection to Kuhn in the light of the same contrast between normativity and descriptivity, prescription and description:

  • 7 P. K. Feyerabend, Philosophical Papers 2. Problems of Empiricism, Cambridge, Cambridge University P (...)

19Whenever I read Kuhn, I am troubled by the following question: are we here presented with methodological prescriptions which tell the scientist how to proceed, or are we given a description, void of any evaluative element, of those activities which are generally called “scientific”7?

20Then, according to Scheffler, Kordig and even Feyerabend, Kuhn committed a serious philosophical mistake: he confused description with prescription, ignoring the normativity intrinsic to philosophical discourse on science. Kuhn would have supposedly broken Hume’s Law, that states that one cannot derive an ought from an is, i.e. one cannot derive a prescription from a mere description. Insisting on doing so will result in a variant of the naturalistic fallacy.

21Now, I will not address in this study whether Kuhn did commit or not this fallacy. Rather, my aim is to see how the previous conceptual tools forged by Reichenbach and the Neo-positivists of the Sixties can be applied to the problem of epistemic norms and the relevance of historical epistemology.

Kant and the Difference between the Genesis and Validity of a Norm

22The distinction between the context of discovery and the context of justification is very useful to capture a fundamental distinction concerning norms in general, i.e. that between the genesis and the validity of a norm. We can find this line of reasoning in Kant’s Critique of Practical Reason, where he deals with the deduction, that is the justification, of the moral law. Kant suggests that the moral law cannot be derived from experience like natural law such as the law of Newton. The reason is that moral law contradicts experience in most cases. For instance, the norm Do not steal is often contradicted by experience, since there are thieves that actually steal. However, we do not consider the existence of thieves as a falsification of the law. On the contrary, the existence of thieves is exactly the reason why the norm holds. This Kantian suggestion can be easily linked to the Reichenbachian and Neo-positivistic considerations stated above.

23Consider again a simple moral norm like “Do not steal”. It is possible to ask two very different kinds of questions about it. The first kind includes questions like: When was this norm born? In which societies does it hold? Where and when people have considered it valid? The second kind of question instead comprises questions like: Is it a right or a wrong norm? Should we follow it or not? The difference between the two kinds of questions lies in the fact that the former kind of questions can be answered independently of whether we think stealing is in itself right or wrong. On the contrary, the second kind of questions can be answered exactly only if we decide about this topic. Indeed, the first type can be answered even if we remain neutral about the rightness or wrongness of the norm, whereas the second type question require us precisely to leave a neutral stance and take a decision about the rightness or wrongness of stealing. In Reichenbachian terms, the first kind of answer concerns the context of discovery of the norm, while the second concerns the context of justification of the norm.

24It should be clear now that two types of questions belong to two different fields, because they can be answered independently. In general, one can write the history of a norm, while completely avoiding the issue concerning its rightness or wrongness. Conversely, one can write a treatise on its rightness or wrongness without knowing a line of its history. We have to do with two independent enterprises. Moreover, confusing one with the other results in a naturalistic fallacy. For instance, to judge stealing wrong just because it is considered wrong in some societies leads to erroneous (and even dangerous) conclusions, such as concluding that slavery is right just because it was considered right in some societies of the past. It is here that Kant’s point turns out to be crucial, for it teaches us that the validity of a moral norm has nothing to do with experience, even with the historical evidence concerning its genesis.

No Prescription from History

25Thanks to the conceptual tools gathered in the previous sections, it is now possible to answer the question stated at the beginning: can the history of an epistemic norm bear a normative value in turn? In the light of the previous considerations, the answer must be negative. The history of an epistemic norm cannot bear any normative value. The mains reasons for this conclusion can be summed up as follows.

26Firstly, the history of an epistemic norm cannot bear normative value because it only concerns its context of discovery, not the context of its justification. Nothing concerning the second may be derived from the first, because of Hume’s Law “no ought from an is”.

27Secondly, the second feature of normativity (ideality) implies that norms are not made of states of affairs. Rather, they are made of idealities, that is pieces of how things ought to be. Now, since the history of a norm is, by its very nature of historical description, made up of facts, it follows that it cannot be used to build norms, which are made of idealities.

28Thirdly, history cannot provide the guiding and evaluating functions typical of norms. This is a consequence of the difference between norms’ context of discovery and context of justification. As previously argued, it is possible to find the answers to the second type of questions concerning validity only in the norm’s context of justification, and not in the context of discovery that regards its genesis. Therefore, if we want to have valid norms, we have to deal with justification, not with genetic considerations, namely history. Doing otherwise will lead us to a justificatory conception of history, that is a conception that considers a norm valid just because it existed—an erroneous and even dangerous approach to history.

Critique, instead of Normativity

29However, if we accept at least one of the previous arguments against the normative value of historical account of norms, are we allowed to dismiss the historical epistemology of epistemic norms as a mere historical enterprise that has nothing interesting to say to contemporary science and philosophy? Indeed, if historical epistemology cannot provide us with valid epistemic norms, but merely with the historical reconstruction of past epistemic norms, should we conclude that historical epistemology cannot say anything about the norms of contemporary science?

30Although it is true that historical epistemology has no normative value, I think that normativity is not what we should expect from the historical epistemology of epistemic norms. My suggestion is that what historical epistemology has to offer is not normativity, but critique. Remember the three regulative functions defining normativity (see Section 3): they lead to conceive normativity as a positive enterprise, namely as a discourse providing positive ideals for guiding and evaluating research. Critique, on the contrary, is a negative task. Its aim is not to provide an ideal horizon to pursue, but to show how the horizons within which scientists move nowadays—that is the epistemic norms that guide contemporary science—are not absolute and a-historical. Critique allows the recognition of the historicity of epistemic norms, not in order to dismiss one norm for another or epistemic norms in general, but to shed light on the norms’ historical contingency.

31Hence, if historical epistemology cannot fulfill the normative task, it can still provide useful critique to the present science, because it can show us the complex historical web of social practices and intellectual concepts in which present epistemic norms are intertwined. Note that unfolding this web displays at the same time the reasons why contemporary epistemic norms—and not others—hold. As a result, the historical epistemology of norms is not a nihilistic or “debunking” kind of critique that uses historical contingency to deny any value to contemporary norms; on the contrary, it provides the historical pattern in which contemporary norms find their horizon of meaning. Historical epistemology conceived as critique is not a way to deny any ground for norms, but a way to highlight the historical nature of the ground.

32In conclusion, I believe that critique is what historical epistemology can bring to the scientific enterprise. In Section 4 it was showed that there are reasons of principle that prevent historical epistemology from having normative power. Still, it must be noticed that, even if historical epistemology could have normative vale, it would not be very useful for science. As historical epistemology itself shows, normativity is not something that comes to the sciences from the outside, but it is a normal product of the scientific practice. Therefore, believing that an external theory, be it historical epistemology or philosophy of science, can establish epistemic norms in science from the outside with some success, is simply ignoring the self-sufficiency of science under this respect.

33Conversely, what science very often lacks, but what it is in very need of, is an authentic awareness of the historical contingency of its practices, concepts and even epistemic norms. My belief is that only historical epistemology can provide it—and this is, after all, a quite strong reason to pursue it.


1 L. Daston, P. Galison, Objectivity, New York, Zone Books, 2007, p. 195.

2 Ibid.

3 Ibid., p. 321.

4 Why then at a certain moment do scientists choose to follow new epistemic norms? Even if this question is interesting and challenging for the historian of science, it falls out of the scope of the present chapter.

5 H. Reichenbach, Experience and Prediction: An Analysis of the Foundation and the Structure of Knowledge, Chicago, University of Chicago Press, 1938, p. 5.

6 See C. R. Kordig, The Justification of Scientific Change, London, Springer, 1971 and I. Scheffer, Science and Subjectivity, Indianapolis, Bobbs-Merrill, 1967.

7 P. K. Feyerabend, Philosophical Papers 2. Problems of Empiricism, Cambridge, Cambridge University Press, p. 132


università degli studi di Milano

Le texte et les autres éléments (illustrations, fichiers annexes importés) sont sous Licence OpenEdition Books, sauf mention contraire.

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search