Why do we generally place more value on knowledge than on belief, however true and justified?

Listen to audio

This lecture has been devoted to presenting the general framework of our thinking and a brief conceptual and historical overview. Although the project of defining knowledge goes back to Plato's Theaetetus, where it is presented as "true opinion endowed with reason ", it is in the contemporary philosophy of knowledge that the general framework has been systematically set for the undertaking to define our usual notion of knowledge, and to question the truth conditions of knowledge attributions of the form "X knows that P". Almost all analyses start from the traditional definition: the necessary conditions for a subject to know that P (have propositional knowledge) are :

a) that he believes that P (that he holds P to be true, the belief condition) ;

b) that P is true (condition of truth, or factivity), and

c) that he has a good reason, a good guarantee, or that he is fully justified in believing that P.

From then on, the epistemologist's work consisted in fulfilling the agenda set by these divisions, by constructing specific theories of belief, truth and the justification of knowledge. But this program of a general epistemology has been the object of two rather related kinds of objections, coming on the one hand from a cognitivist or "naturalist" turn and, on the other, from a "social turn" in epistemology, supposed to have awakened classical epistemology from its dogmatic slumber. The first objection is that epistemic notions such as belief, knowledge, evidence, reason or justification cannot be defined in exclusively conceptual terms, on the basis of thought experiments and counter-examples, pre-theoretical or semantic intuitions: thus, the intuitions solicited in epistemology on examples and counter-examples, via the method of "cases", would be highly relative and fluctuating (for example, Asians would not have the same reactions as Westerners), and the notion of knowledge would have to be analyzed instead on the basis of the causal and natural processes that produce it, in biological, psychological terms, and thus rather as a form of cognition. The second objection is that general epistemology is too individualistic, and fails to take sufficient account of the social dimension of knowledge. The subject who is supposed to "know" is an abstract individual, whose epistemic properties are defined without any reference to a social or historical context. Yet beliefs and knowledge are widely shared within a community, acquired and transmitted within a social universe. What both objections have in common is that they underline the relative and contextual nature of knowledge, and the need to define knowledge or justification not in strictly conceptual and normative terms, but in descriptive and natural terms. The task of description would be entrusted either to biology and the cognitive sciences, or to disciplines such as history, the social sciences, or any appropriate combination of these. From then on, the temptation is strong to turn to social epistemology, whose answer is: the value of knowledge lies in the fact that it is based on the common good. When it took an eliminativist turn, this epistemology developed a strong program that took the names of social constructivism, or the archaeology of knowledge-power. The aim is to replace epistemology by insisting on the social and historical relativity of knowledge. Traditional notions have no place here: "truth is a social institution" (Steven Shapin). The question "Who knows?" is more important than "What do we know?" or "How do we know? One of the aims of the lecture was to show that it is one thing, in extending classical epistemology, to take into account, for example, unquestionable sources of knowledge such as testimonies or forms of knowledge attribution that depend on varied conversational contexts (contextualism) (Reid, Coady), it is quite another to consider that this signs the death certificate of the philosophy of knowledge as such. If knowledge has value, and if this value has something intrinsic about it, then it is doubtful whether knowledge is not only entirely subservient to our practical and social interests, including the types and methods of reasoning that this most often implies, but also, contrary to what this or that pragmatist or sociological drift of knowledge advocates, that it is defined in these terms.

1. The philosophy of contemporary knowledge is a fast-growing field currently undergoing profound restructuring and expansion. A "value turn" or " axiological turn " has taken place. It has been realized - or rather, classic themes have been rediscovered - that questions such as the nature of epistemic justification are linked to problems such as justification in ethics (in both cases, evaluations and normative inquiry are involved), that there are analogies, and perhaps homologies, between the realms of theoretical and practical rationality, and that it is necessary to redraw the boundaries of epistemology. As a result, greater emphasis is being placed on the links between ethical and cognitive evaluation. Since the 1960s, knowledge had been examined primarily in terms of its Platonic tripartite definition as "justified true belief". An article by E. Gettier in 1963 (cf. Tiercelin, 2005; Dutant and Engel, 2005), which showed that the three conditions could not be used to define knowledge, gave rise to many competing analyses. If the subject - Gettier showed, for example - knows something only by pure chance, then he doesn't "know". To ward off this "epistemic fortune" (not unrelated to the "moral fortune" that Kant intended to ward off in ethics), philosophers proposed conditions that were either "internalist", reinforcing the conditions for the justification of beliefs accessible to the subject, or "externalist", reinforcing the conditions for the formation of beliefs independently of the subject's access, with no single analysis achieving unanimity. As modifications have been introduced to improve the definition of knowledge as justified true belief, a majority conception has developed around a fundamentally causal and reliable approach to justification (Goldman, 1986): justification refers to the possession of concepts of a certain type, and the evaluation of beliefs derives from the evaluation of belief-formation processes: for epistemology, it is therefore a question of evaluating the reliability of the latter, as well as that of the inference mechanisms that govern transitions between beliefs. As Goldman notes, "logic alone cannot determine which processes are admissible, since this depends on the properties of our basic cognitive apparatus", as well as on social organization. But such a fundamentally externalist approach, which we can see why it could fit in with a cognitive approach to knowledge, came up against a major tension: on the one hand, if it's a causal theory, it must only introduce reliable actual causal beliefs, describable in purely descriptive and objective psychological terms (i.e. independently of what the subject knows or believes about these beliefs), but on the other hand, this theory is not purely causal, since it must also include beliefs that the subject should form, i.e. "rules for the direction of his mind". This second condition is not only internalist ; it is also normative, in the strong sense that it presupposes that the subject must, in order to know, be aware of what he ought to know ("ought" here having a modal and deontic meaning: it relates to an obligation, not just a permission). In other words, reliability epistemology is intended to be descriptive, psychological and naturalistic, in line with the ideal of the cognitive sciences from which it draws its inspiration. But it must introduce modal, prescriptive and normative notions. Hence the dilemma: either it is impossible to explain these notions in terms that are themselves causal and naturalistic, and epistemology cannot therefore formulate its principles and rules of justification in a purely causal or psychological mode. Or it can, which is desirable, because if epistemology has to renounce the normative meaning of the notion of knowledge - which is the difficulty of any "naturalized" epistemology in Quine's sense - far from being a change of doctrine, it risks changing its subject altogether: we therefore need to explain how and why such normativity is compatible with a fundamentally naturalistic psychological analysis. The second major moment in epistemology in the last ten years has been marked by the emergence of an epistemology that focuses more on the agent, its capacities and virtues, than on the empirical evidence available to this agent to establish the necessary (though not sufficient) conditions of truth and justification for knowledge. Thus, Ernest Sosa (1991) proposed the following idea: a belief can only be guaranteed if it is not only the product of reliable causal processes, but the product of the subject's dispositions and capacities. These dispositions can be called virtues. Within an essentially post-Kantian and post-positivist epistemology, an Aristotelian theme of intellectual virtues has thus returned. According to some proponents of "virtue epistemology", dispositions are acquired aptitudes or abilities that are essentially defined in functional terms; according to others, they are voluntary habits involving a commitment on the part of the agent (Zagzebski, 1995). The general idea is that, if an agent has knowledge, it can be credited to him or her (Greco, 2003). Virtue epistemology therefore starts from the idea that epistemic properties (truth, justification) are not the only ones relevant to epistemic evaluation, but it also reintroduces the idea that a knowing subject is subject to an evaluation that may be, depending on the case, more or less close to ethical evaluation, or whose cognitive and metacognitive aspects we can try to understand. For theorists of epistemic virtue, "value monism", centred on truth and knowledge, must therefore give way to a broadening to include other values, such as understanding, or the flourishing of intellectual virtues associated with moral virtues (Kvanvig, 2000). This raises the question of whether or not notions such as voluntary control and commitment require a return to a form of internalism and a strong deontological conception of our epistemic obligations, or whether, including in such a vision of epistemology, there is reason to propose a naturalistic reading of control, consciousness, will and commitment.

2. We have first shown that the problem of the value of knowledge covers, as D. Pritchard(Knowledge and Value): 1) Should we follow our intuition that knowledge is more valuable than justified true belief? 2) The secondary problem: why is knowledge more valuable than any of its sub-parts (i.e. true belief, or justified true belief), when neither of these terms really allows us to arrive at a satisfactory definition of knowledge any more? 3) Should we consider that knowledge has a final value, which transcends all its intrinsic properties, whatever they may be? But before answering any of these questions, we had to ask ourselves whether or not we were right to consider that knowledge has more value than true and justified belief (with good reasons or guarantees).

3. We began by explaining whyknowledge cannot be reduced to true and justified belief alone, recalling Plato's classic tripartite definition of knowledge in the Theaetetus. in other words: S knows that p, if and only if 1) S believes that p ( condition of belief); 2) p is true (condition of truth or factivity); 3) S has good reasons (or guarantees) to believe that p (condition of justification). We returned to the problems raised by this definition according to Gettier (1963), and to the various failures of the many attempts made over the last fifty years to remedy them, to no avail.

4. We then saw that, although our intuitions seem to lead us to place a higher value on knowledge than on true belief, this is not necessarily self-evident, as illustrated by Plato's somewhat ambivalent attitude in a famous passage of the Meno, on the road to Larissa: a) knowledge is not superior to right opinion, when it comes to practical ends (but lacks "stability"); b) the value of knowledge is not reduced to its practical value. To demonstrate the ambiguity of our intuitions, we then presented L. Zagzebski's critique of reliabilism.

5. We then proposed some solutions to overcome these aporias. We presented the models most often proposed today: a) reliability: admitting that justified true belief is sufficient; b) the response of virtue epistemology: knowledge as credit given to the agent by success linked to the achievement of a cognitive ability (J. Greco); c) the response of virtue epistemology associated with an "anti-luck" condition (D. Pritchard).

We have concluded that these models offer only partially satisfactory answers; which is why, unless we give in to both the relativist and the skeptic, we must seek another conception of knowledge, if we are to be able to answer not only the question of the value of knowledge, but even more broadly the question of epistemic values (such as wisdom, being able to understand something).