Navigation – Plan du site

AccueilNumérosHS-19V. Considérations méthodologiquesEpistemology and Responsibility

V. Considérations méthodologiques

Epistemology and Responsibility

Fabio Bacchini

Résumés

Cet article a pour objectif de mettre en évidence l’unité qui sous-tend les travaux de Pierre-Yves Raccah relevant des champs divers tels que théorie de l’argumentation, éthique, épistémologies générale et morale, philosophie des sciences et sciences cognitives. Dans un premier temps, appuyé sur sa conception argumentative de la justification, je montrerai que son antifondationnalisme moral est une conséquence directe de la prise en compte de la reponsabilité en éthique et tout particulièrement de l’exigence relative à la méta-éthique : les êtres humains sont totalement responsables de leurs positions morales aussi bien que des actions exécutées par rapport à ces positions morales. Dans un deuxième temps, mes interrogations seront portées à son anticonventionnalisme aussi bien qu’à ses considérations concernant les métapréférences, qui sont mis en œuvre pour pouvoir maintenir la responsabilité dans le domaine des prescriptions. Pour terminer, je mettrai en lumière que son épistémologie générale est fondée sur la thèse selon laquelle nous sommes moralement responsables des théories scientifiques aussi bien que des constats empiriques acceptés comme vrais, la responsabilité étant pertinente dans le domaine des descriptions fondées sur l’empirisme.

Haut de page

Texte intégral

Introduction

1In this paper, I will show that there is a deep unity underlying Pierre-Yves Raccah’s views in the theory of argumentation, ethics, moral epistemology, general epistemology, philosophy of science and cognitive psychology. A corollary of this position is that Raccah has an inclusive philosophical system to offer, whose numerous specific standpoints are consistently and harmonically intertwined. I will focus on the notion of “responsibility”, and argue that the key role it plays in Raccah’s ethics and moral epistemology parallels the crucial function it has in his general epistemology. I claim that the central thesis is that people are fully morally responsible of the moral positions they hold, as well as of the actions they perform in observance to these moral positions. The whole set of the other stances he takes – his argumentative conception of justifications, his moral antifoundationalism and nonconventionalism, his view of human freedom – is required to be consistent with that central thesis. When we consider Raccah’s general epistemology and philosophy of science, however, we discover that the situation is very similar: just as Raccah’s moral epistemology is aimed at rejecting the claim that we may not be morally responsible of embracing the particular moral position we embrace, his epistemology and philosophy of science are aimed at rejecting the claim that we may not be morally responsible of embracing the particular description of the world we embrace. Mutatis mutandis, the same stances are taken in the descriptive as in the prescriptive realm under the same rational pressures and dependently to the same philosophical reasons. I explain in what sense Raccah’s antifoundationalist position about descriptive argumentation corresponds to his antifoundationalist position about normative reasoning, and I conclude by showing that there is no incoherence – no idealist drift, and even no detachment from empiricism – in embracing Raccah’s idea that we are fully morally responsible of all the empirical statements we accept as true.

1. Rejecting moral foundationalism

2I first met Pierre-Yves Raccah in 1994, and I believe that one of the cornerstones of his philosophy is also one of my first indelible memories of him. We immediately started using to spend hours discussing any claim just for fun – what we regularly keep doing. One night, some of us made a moral claim. Someone else attacked it. The claim was repeatedly charged and defended, and the burden of proof shifted on several claims that we happened to use as premises, until he said: « You cannot go further. At the end of any discussion of this sort you will always arrive to “J’aime” or “Je n’aime pas”». His thesis deeply impressed me:

  • 1 He still holds this view to date (personal conversation 2014). See Raccah (2004) for an attempt to (...)

The J’Aime / N’Aime Pas Thesis: at the very beginning of every prescriptive argumentation, what we ultimately find is “J’aime” (“I love it”) or “Je n’aime pas” (“I do not love it”)1.

  • 2 See Kant (1790). Also see, e.g., Zangwill (2013).

3This is perhaps the most radical way of formulating an antifoundationalist position about normative reasoning – unless you think that you can found a moral claim on a judgement of taste that is commonly exhibited as the typical subjective non-universal judgement. In fact even in Kant, while “X is beautiful” is a subjective, universal judgement of taste, “I like X” or “I love X” are just subjective, non-universal ones: they do not claim any validity; we cannot be said to think of them being universal; they feature no normative constraint2. And one may presume that we cannot found normative reasoning on subjective non-universal judgements of taste.

  • 3 Since Hume set the problem, many philosophers have written in support or against his thesis. For a (...)

4But how is the whole story supposed to go? We may reconstruct it by saying that, in our effort to argumentatively justify a prescriptive (e.g., a moral) statement, we must of course make use of further prescriptive statements as premises. As Hume famously claimed, in fact, we cannot go from an “is” to an “ought” – that is, there is no argument that has only descriptive premises, has a prescriptive conclusion, and is valid3. Hence the problem of argumentatively supporting one normative statement – the original conclusion – turns into the problem of argumentatively supporting some further normative statements – the premises. What Raccah was saying is that in the normative field the justificative procedures have no end, and our quest for a stable foundation of a prescriptive statement is doomed to frustration. The only way of stopping the infinite regress is renouncing the need for justification somewhere by allowing an unjustified premise.

5Of course this situation is the same we are in when we look for an inferential justification of descriptive premises in empirical science. We need to consider as true at least some empirical premises in order to inferentially justify any statement, and this need starts an infinite regress as well. Jacob Friedrich Fries (1807) famously established that there are only three options here, and none seems acceptable: infinite regress, dogmatism, and psychologism. 150 years later, Albert (1968) reaffirmed Fries’s trilemma by slightly modifying it – he named it “Münchhausen’s trilemma”, unified dogmatism and psychologism, and added circularity as the third horn. So we find this tri- or maybe quadri-lemma again in Raccah’s account of prescriptive argumentation, in one pack with Raccah’s solution: we must drop our concern for foundational justifications and accept that our justifications are not founded at all – since they necessarily finally depend on our “J’aime”s and “Je n’aime pas”s. Obviously Raccah believes that “J’aime” and “Je n’aime pas”, when used as the final premises of a justificative argumentation, should not be taken as dogmas or psychological foundations. He rather thinks that, by making explicit that our justifications of our prescriptive positions are hung on our unjustifiable, subjective, non-universal and maybe non-rational judgements of taste, we make it clear that our prescriptive positions are not something we are not responsible of.

6I claim that the notion of responsibility is one of the most crucial – if not the most crucial – in Raccah’s overall philosophy. Thus it is important to clarify what is the relation between his moral antifoundationalism and the importance he assigns to responsibility in ethics and even in epistemology.

7Suppose for a moment that moral positions could be argumentatively justified moving from ultimately non-argumentatively justified prescriptive premises. We do not have to specify here how it is that these premises happen to be non-argumentatively justified: it might be that they are self-evidently correct (psychologism), or we are accepting them as non-debatable (dogmatism), or they are acknowledged as having any altogether different status. What happens is that we are now morally justified in embracing our moral positions as a consequence of our being morally justified in embracing whatever is argumentatively justified moving from whatever non-argumentatively justified prescriptive premises. However, since we are evidently not morally responsible for what is completely independent from us, we are not morally responsible for the moral positions we are morally justified to hold. In fact, we are not morally responsible of the fact that our moral positions can be argumentatively justified moving from certain premises – nor are we morally responsible of the fact that these premises are non-argumentatively justified premises. It turns out that we can be perfectly morally justified in embracing certain moral positions without being morally responsible of embracing them.

8This conclusion would seem absurd to many; I am sure it has appeared intolerable to Raccah, and that his moral antifoundationalism should be viewed as a move aimed at avoiding it. Raccah wants to conclude that we are fully responsible of what moral positions we hold – as well as of what we intentionally do as a consequence of our embracing certain moral positions rather than others. Moral foundationalism seems, on the contrary, a theoretical view entailing that we are not morally responsible for what moral positions turn out to be correct. Moral foundationalism makes people acting accordingly to well-founded moral statements not morally responsible of how they act.

  • 4 By ‘the most robustly founded moral position’ I want here to refer to either the moral position tha (...)

9In a Kantian spirit one could object that, no matter that a moral position is well-founded, we are still morally responsible of embracing it when we embrace it, in as much as we are free not to embrace it. The answer is that, if moral positions admit foundations, it could be easily argued that we are not really free to embrace the most robustly founded moral position we actually embrace4. Let us consider two separate steps in which our embracing the most robustly founded moral position could reasonably take place:

Step 1. We commit to embrace whatever moral position will turn out to be the most robustly founded one;
Step 2. We actually embrace moral position M as a consequence of our identifying M as the most robustly founded moral position.

10The point is that both steps can be said to lack the kind of freedom that is necessary for moral responsibility to obtain. In fact, taking Step 1 seems required by reason, in the sense that we appear not free to embrace an ill-founded (or, a non-founded) moral position rather than the most robustly founded one – just like as we are not free to believe a descriptive proposition which is contradicted by evidence rather than one which is supported by evidence. In other words, as long as moral positions admit foundations consisting in argumentative justifications moving from non-argumentatively justified prescriptive premises, as sketched above, we simply seem to lack freedom to resist the rational pressures of these foundations with regards to the choice of the moral positions we will explicitly hold. The right kind of freedom seems also absent in Step 2, where we appear as forced to embrace the particular moral position M that satisfies the description of being the most robustly founded one. Thus, if M turned out to entail that – say – gratuitously causing the sufferance of animals is morally preferable to omitting to do so, we would be obliged to embrace M as a consequence of Step 1 and of M’s satisfying the relevant description, no matter that many of us happen to dislike causing the suffering of living creatures. As a result, sadistic individuals could start procuring sufferance to animals without being morally responsible of it. They could just defend themselves saying that:

1. It is M that asks us to cause the suffering of animals.
So,
2. we only would be morally responsible of causing the suffering of animals if we were morally responsible of embracing M.
But
3. we are not morally responsible of what we are not free to avoid;
and since (if moral foundationalism holds)
4. we are not free to escape from Step 1,
and
5. we are not free to escape from Step 2,
then
6. we are not free to abstain from embracing M.
As a result of 3. and 6.,
7. we are not morally responsible of embracing M.
And, as a result of 2. and 7.,
8. we are not morally responsible of causing the suffering of animals.

11On my view, Raccah’s philosophy is importantly characterized by the need to reject 8. He would admit 1., 2. and 3. He also would admit that 4. and 5. entail 6.; 3. and 6. entail 7.; and 2. and 7. entail 8. His strategy to reject 8. consists in rejecting both 4. and 5. by rejecting moral foundationalism.

2. How free are we?

12I think that the J’Aime / N’Aime Pas Thesis is sufficient for guaranteeing the rejection of moral foundationalism. However, the rejection of moral foundationalism is only necessary in Raccah’s philosophy in order to reject 8. and assuring that we are morally responsible of the moral positions we hold, as well as of the actions we perform in observance to these moral positions. So the real question is whether the J’Aime / N’Aime Pas Thesis is sufficient for assuring these important conclusions. I am afraid that we cannot take it for granted that it is. And the main reason is this: what we really must provide is a philosophical basis for rejecting 6. What the J’Aime / N’Aime Pas Thesis offers is just the discard of a proper subset of all circumstances in which 6. can be true – that is, the subset of those circumstances in which 6. is true because moral foundationalism holds. However, 6. can also be true if moral foundationalism does not hold. We have no prima facie reason to think that the J’Aime / N’Aime Pas Thesis allows us to reject 6. also under all circumstances in which moral foundationalism does not hold.

13Moreover, we have some prima facie reasons to suspect that the J’Aime / N’Aime Pas Thesis actually confirms 6. The argument goes like this:

Suppose that moral foundationalism is false. Suppose that the J’Aime / N’Aime Pas Thesis holds, and that 9. we just happen to embrace M because we have a good argumentative justification of M moving from our strongest, least dispensable “J’aime” and “Je n’aime pas”.
In other words, we embrace M rather than any other moral position because the “J’aime”s and “Je n’aime pas”s serving as premises of our M-supporting argumentations appear to us as stronger and less dispensable than any other “J’aime” and “Je n’aime pas” supporting any other moral position different from M.
Yet it is true that
10. we are not free to avoid liking what we intensely like and disliking what we intensely dislike,what in turn entails that
11. we are not free to avoid having the particular strongest and least dispensable “J’aime” and “Je n’aime pas” that we have.
As a consequence of 9. and 11.,
6. we are not free to abstain from embracing M.
What, as we know, also entails 7. and 8 (with the help of, respectively, 3. and 2., whose truth is not under discussion).

14What is wrong in this argument? I think that this is a very compelling one, and that it puts at risk the usefulness of the J’Aime / N’Aime Pas Thesis. No doubt that a conventionalist alternative to the J’Aime / N’Aime Pas Thesis would be a good solution to solve the problem. What we need is some guarantee that we are morally responsible of using certain prescriptive statements as the premises of the justifications of our moral positions. By making the individuation of the premises a free act, conventionalism easily achieves this result.

15Yet conventionalism has the negative aspect of making moral positions the result of an arbitrary, “coldly” deliberative act. First of all, if we freely choose our prescriptive premises without being pressed by any freedom-threatening reason, it seems that we literally have no reason for choosing the particular premises we choose rather than any other. Having no reason supporting one’s choice means choosing non-reasonably: therefore, we could be said to non-reasonably hold each moral position we happen to hold. Secondly, our moral positions are not just prescriptive statements we “coldly” subscribe to. Each moral statement expresses a moral sentiment, where a moral sentiment is a disposition whose occurrent manifestations are moral emotions (Prinz 2007). This entails that a moral statement actually expresses both a standing moral sentiment and its occurrent moral emotional manifestation. For example, moral indignation is a moral emotion, and “Killing is wrong” may express the moral emotion also expressible as: “Killing: what a moral indignation I feel!”. If I am not feeling this emotion at the time I am uttering the sentence, however, the statement still expresses the moral sentiment consisting, among other things, in my disposition to feel moral indignation if someone kills. Now, conventionalism seems to neglect the fact that moral emotions and moral sentiments are very “hot” mental items, and that the very fact that we happen to feel them enters non-conventionally in our choice of certain moral premises rather than others. This is also part of the story why embracing a moral position has a constitutive – though not insurmountable – motivational aspect. No doubt that, however, moral emotions and moral sentiments are once again something that we do not seem to be free of avoiding. Thus we may suspect to be caught in a dilemma: whether we account for our freedom to choose our moral premises – what rescues moral responsibility at the price of neglecting the emotional, sentimental and motivational aspects of morality and maybe also making all our moral positions non-reasonable – or we account for the non-arbitrariness of the moral premises by connecting them to our emotional life and finally to our personality – what saves a large part of our intuitions about morality at the price of threatening moral responsibility.

16My solution to escape this dilemma is adopting a form of quasi-conventionalism consisting in choosing to use as prescriptive premises those prescriptive statements that express moral sentiments which we experience with an intensity high enough for temporarily considering them universal and non-debatable. In substance, the moral sentiment expressed by a basic prescriptive statement must appear so undeniable as to render temporarily superfluous any line of argument supporting that statement. I have called this approach “Popperian” because it develops for prescriptive premises the same strategy that Popper (1934) proposed for individuating basic statements in science (Bacchini 2015). Yet my attention is mainly focused onto making it rational – in the sense of intersubjectively approvable – to stop at some unjustified prescriptive statements and use them as premises. This is why I propose that we take the intensity of a moral sentiment, and of a moral emotion, as a temporary clue to its universality and non-debatableness.

17I think that Raccah’s approach is different. His J’Aime / N’Aime Pas Thesis is not meant to be consistent with my idea that we choose those prescriptive premises that we presume express universal and non-debatable moral sentiments. Of course I take the intensity at which we personally experience the moral sentiment as the only evidence we must use to infer its universality and non-debatableness – so my position is not fully conventionalist and cannot be easily accused to collapse into a form of tactical opportunism. Yet my proposal takes final argumentative moral agreement as an important goal. Neither the J’Aime / N’Aime Pas Thesis nor Raccah’s general conception of argumentation do so.

  • 5 From this point on, whenever I say “action” and “acting”, what I intend is “action or omission” and (...)
  • 6 He uncompromisingly expressed this position in many scientific talks and other public occasions. On (...)
  • 7 E.g., under both the counterfactual and the actual sequence view: it is true that I could have done (...)
  • 8 Raccah (personal communication 2014).

18So what is Raccah’s solution to the dilemma? His orientation is towards rejecting – or at least importantly weakening – 10. and therefore 11. In a certain measure, according to him we can choose what we like and dislike. Those who know Raccah can say how many times he insists that you can stop liking or disliking something – or equivalently, you can start doing so – if you just want to. This attitude is part of his most general conception of human freedom as a feature that is much more present in our actions (and omissions5) than what is usually acknowledged. Raccah claims that we are literally free also in situations in which we would routinely say that we are not. Take, for instance, the case in which a robber threatens me to kill my son if I do not give him the combination of the strongbox. Obviously I give him the combination; and later I describe what happened by saying that “I had no choice”, or, that “I wasn’t free not to give him the combination”. Raccah thinks that these reconstructions are manifestly false6. In fact I was completely free not to obey the robber. I was free to opt for the possible worlds in which I do not obey the robber and risk that my son be killed by him. But I did not want these possible worlds to become actual. And I freely chose the alternative possible worlds in which I obey the robber. Under all the main philosophical accounts of human freedom7, I was fully free. The only situations in which I am really unfree are those in which I undergo physical coercion: but since these are also situations in which I don’t act – and rather I am acted – Raccah can argue that whenever we act, we are free to act8.

  • 9 Note: we need to be able to know that the effect is caused, among other factors, by our acting as w (...)

19This general thesis, which I call the Extra-Freedom Thesis, entails that we are morally responsible in much more occasions than we admit we are – I call the latest claim the Extra-Responsibility Thesis. It is simple to notice that the Extra-Responsibility Thesis just follows from the conjunction of the Extra-Freedom Thesis and the claim that we are morally responsible of whatever (i) is partly caused by our freely acting in a certain way, and (ii) we are able to know is partly caused by our acting in that way9 (or, some other adequate epistemic requirement). I take the Extra-Responsibility Thesis to have a central role in Raccah’s philosophy – impacting not just his ethical but also his epistemological positions, as I will show in the next section.

20As a consequence of the Extra-Responsibility Thesis, and contrary to popular opinion, we may be morally responsible of some tastes and preferences we have. And this would be only possible if, as a consequence of the Extra-Freedom Thesis, and contrary to popular opinion, we may be free to have those tastes and preferences. But how can Raccah maintain that we may be free to choose our tastes and preferences? At a first approximation, his answer is that we evidently are not forced to have all the preferences we happen to have – and, conversely, we are not forced to have not all the preferences we happen to have not. Suppose, for example, that you have to hire a new worker in your company. You discover to have a preference for hiring a white person. As soon you discover to have this preference, however, you also start wanting to stop having it. And what you do is precisely stopping having the preference: you successfully choose the new employee without being influenced by this preference because you stopped having it.

21Of course it is also possible that you continue having the preference and you successfully try to limit its influence on your decision. But what is claimed here is that it is also possible that you successfully altogether eliminate the preference from the set of the preferences you have. In as much as this is possible, it is true that we can at least partially reject 10. and 11.

  • 10 “Siempre queda la posibilidad de resistir a la tendencia a connectar el grafo” (Raccah 2004: 8).

22Raccah’s view of human psychology is indeed highly hierarchical. He thinks that it is a very remarkable fact about our psychological architecture that we can recursively build up representations about representations. What seems crucial to reject 10. and 11 is our mental capacity to develop higher-order preferences about lower-order preferences, together with the fact that higher-order preferences are sometimes effective in modifying the set of lower-order preferences that we have. Consider, for example, Raccah (2004), where the personality of a particular individual is represented as a specific network of positive and negative commitments to ordinary judgments of taste. He maintains that we happen to have higher-order preferences about how our personality may change – for example, we typically may desire to have a personality whose representation is a more compact and connected network of lower-order preferences. He also believes that we are free to modify the network according to these higher-order preferences. In particular, we are free to drop a lower-order preference we no longer like to have because of its being scarcely integrated with the others. Or, we are free to create new lower-order preferences in order to connect an isolated lower-order preference to a more integrated group of them10.

23What Raccah needs for rejecting 10. and 11. is nothing but claiming that:

a. Our mental life includes not only representations but also metarepresentations;
b. Some metarepresentations we have are higher-order preferences that a lower-order preference we have is removed from the set of our preferences – or, that a lower-order preference we do not have is added;
c. A number of metarepresentations described in b. are effective.

24Claims a. and b. seem not problematic, and I will assume them to be true. On the other hand, c. can be contested. Yet I think that it is not difficult to bring psychological evidence supporting it. We do not have to forget that a metarepresentation of the kind described in b. can result to be effective also in virtue of its successfully generating some effective volitions to change one’s behaviour in such ways that facilitate the content of the original metarepresentation to become true. For example, my higher-order preference for stopping disliking dogs may turn out to be successful in virtue of its successfully generating my effective volition to start spending my time with the lovely dog of my best friend.

25This solution to the dilemma sketched above can be said to be Frankfurtian, of course, because Harry Frankfurt (1971) first famously distinguished between our freedom to act accordingly to whatever are our first-order desires (which he called ‘freedom of action’) and our freedom to act accordingly to those first-order desires determined accordingly to our second-order desires (which he called ‘freedom of the will’). Frankfurt said that:

A person who is free to do what he wants to do may yet not be in a position to have the will he wants. Suppose, however, that he enjoys both freedom of action and freedom of the will. Then he is not only free to do what he wants to do; he is also free to want what he wants to want. It seems to me that he has, in that case, all the freedom it is possible to desire or conceive. There are other good things in life, and he may not possess some of them. But there is nothing in the way of freedom that he lacks. (Frankfurt 1971: 17)

26The problem, for our Frankfurtian Raccah, is that there is a possibly lethal objection to his move. The objection is this. Let us assume that we can choose our “J’aime”s and “Je n’aime pas”s accordingly to our higher-order preferences about our “J’aime”s and “Je n’aime pas”s. These higher-order preferences can be also described as higher-order “J’aime”s and “Je n’aime pas”s serving as the real final premises in the justifications of our moral positions. Now, do we choose to have them? If we do not choose to have them, as it seems, both 10. and 11. turn out to be true, and 6., 7. and 8. also turn out to be true. In other words, there is no real progress from a situation in which our first-order “J’aime”s and “Je n’aime pas”s are not freely chosen, to a situation in which our higher-order “J’aime”s and “Je n’aime pas”s are not freely chosen – because in both cases we are not free to abstain from embracing M. The same objection, of course, can be directed to Frankfurt: if my second-order volitions are not the second-order volitions I want to have, I am not free to have them; this means that I am choosing freely neither my first-order volitions (which I determine accordingly to my second-order volitions) nor my actions (which I determine accordingly to my first-order volitions).

27I think that Raccah has two good answers to give here. The first answer is that it is not necessary that the “J’aime”s and “Je n’aime pas”s we use as unjustified premises in the justifications of our moral positions be the highest-order metapreferences we have ascended to. Suppose the highest-order metapreference you have arrived to is the preference that you prefer to prefer to avoid make animals suffer. Yet one of the final unjustified prescriptive premises of the argumentative justifications of our moral positions can be “Je n’aime pas make animals suffer”, or, “J’aime preferring to avoid make animals suffer”. Clearly these “J’aime”s and “Je n’aime pas”s are lower-order with regards to the highest-order metapreference you have arrived to – what means that they have freely been chosen.

28In some sense, this answer parallels an analogous defence that one could supply to Frankfurt’s position. In fact, in the same spirit one could defend Frankfurt by saying that an action is free just as long as the person acts accordingly to her will – no matter that the will is free; and, a person’s will is free just as long as it is determined accordingly to her second-order desires – no matter that these second-order desires are determined accordingly to higher-order desires. On this view, lack of freedom is not inherited down through representational levels.

29The second answer is that there is in principle no limit to the recursive human psychological capacity to form higher-order metarepresentations. We can form metarepresentations of the third order, the fourth order, and so on. This means that we always have the possibility to choose whatever level of desires, preferences, tastes and volitions we want. Although we practically must stop at some level, it is never true that we have to stop at that level. And, contrary to Frankfurt, this truly seems “all the freedom it is possible to desire or conceive”. On this line of reasoning, the J’Aime / N’Aime Pas Thesis still holds. Yet we can affirm that we are never forced to adopt any particular set of prescriptive preferences – what is sufficient for making us morally responsible of whatever are the “J’aime”s and “Je n’aime pas”s we use as unjustified premises in the justifications of our moral positions. At least, we are morally responsible of having accepted them without bringing our control up to the higher level – what we always are free to do.

30As I see it, Raccah can successfully argue that the “J’aime”s and “Je n’aime pas”s we use as unjustified premises in our justifications are constitutive of that part of our personality we have the power to shape and we bring the moral responsibility of. If I use a certain “J’aime” as a premise in my argumentative justifications, not only I have it: I also freely decide to acknowledge it the status of a premise in my argumentative justifications – what is a completely different thing, and what seems to be a free act of mine which I am fully responsible of. I can easily abstain from acknowledging that preference such status, no matter that I have the preference. This is clearly a respect under which Raccah’s position is more easily defendable than Frankfurt’s.

31This concludes my analysis of how, in Raccah’s philosophy, an argumentative conception of justifications, the J’Aime / N’Aime Pas Thesis, moral antifoundationalism and moral nonconventionalism meet both the requirement that our moral premises be non-arbitrary and connected to our emotional life and personality, and the requirement that we can be said fully morally responsible of the moral positions we hold.

3. Epistemology

32Just as Raccah’s moral epistemology is aimed at rejecting the thesis that we may not be morally responsible of embracing the particular moral position we embrace, his general epistemology is aimed at rejecting the thesis that we may not be morally responsible of embracing the particular description of the world we embrace.

  • 11 Raccah (2005b).

33Raccah’s general epistemology parallels his view of moral epistemology. We are fully morally responsible of the descriptive statements we embrace, just as we are fully morally responsible of the prescriptive statements we embrace. Not only are we not forced to accept any scientific theory as true, but also we are not forced to accept any empirical statement as true. The Extra-Freedom Thesis and the Extra-Responsibility Thesis apply here. As a consequence of this account of human knowledge, “scientists and divulgators have a central responsibility in the way the world we live in is”11: they freely choose what is the true description of the world, and cannot be exempted from their responsibility on the grounds that what is true is independent of them.

34Let us examine how these claims find a specific support in Raccah’s philosophy. First of all, he maintains – like Duhem, Quine and many other philosophers of science – that scientific theories are constitutively underdetermined by empirical evidence. This is to say that there is no valid inference from empirical statements to scientific theories: many different and incompatible theories are consistent with the set of empirical statements accepted as true, whatever their number and accuracy.

35But what about the idea of founding scientific theories in the way we (try to) found moral positions – that is, by argumentatively justifying them moving from those justified descriptive premises that empirical statements accepted as true after careful consideration can be supposed to be? If we could be successful in this attempt, we would be able to say that we are not responsible of the scientific theories we embrace (the equivalent of claim 7.), no matter that an “argumentative justification” is an invalid inference. Note, however, that this is exactly the situation described by Fries (1807) and Albert (1968): the only options seem to be infinite regress, dogmatism, psychologism or circularity.

36Mutatis mutandis, in fact, Raccah’s moves are the same as in the moral field. He asks what could make any descriptive statement a justified descriptive premise. Arguably the answer is that an empirical statement that is accepted as true after careful consideration is the true description of one or more facts: an obtaining state of affairs, apparently independent of us. But what if someone questions our justification by saying that one of our descriptive premises is not the true description of a fact? We presumably should try to justify the statement that “the premise called into question actually is the true description of a fact” moving from more robustly founded premises. These premises, however, would be at best other empirical statements about which the same question could be raised. « At the end of any discussion of this sort – Raccah can repeat – you will always arrive to “J’aime” or “Je n’aime pas”.» What kind of “J’aime” or “Je n’aime pas”, this time? Well, this kind: “I love / want to consider descriptive statement d as a true description of a fact” and “I do not love / want to consider descriptive statement d as a true description of one or more facts”. So there is an equivalent to The J’Aime / N’Aime Pas Thesis for the descriptive domain after all:

  • 12 By “descriptive argumentation” I mean any argumentation whose conclusion is a descriptive propositi (...)

The equivalent to the J’Aime / N’Aime Pas Thesis for the descriptive domain: at the very beginning of every descriptive argumentation12, what we ultimately find is “J’aime” (“I love / want to consider descriptive statement d as a true description of one or more facts”) or “Je n’aime pas” (“I do not love / want to consider descriptive statement d as a true description of one or more facts”).

37And – like for the moral domain – Raccah adds that we are basically free to choose our “J’aime”s or “Je n’aime pas”s. His conclusion is that we are fully responsible of the descriptive statements we embrace as true. We are morally responsible of both the scientific theories, and of the empirical statements we accept as true.

38Of course the two latter claims are not equivalent at all. In fact many philosophers, while admitting that accepted scientific theories must somehow be “chosen” as an effect of their being underdetermined by evidence, would maintain that facts are imposed to us by the world, and that empirical statements accepted as true after careful consideration – since they (hopefully) just are true descriptions of facts – are also imposed to us by the world. The argument seems to go like this:

12. An empirical statement that is accepted as true after careful consideration is a descriptive statement that is a true description of one or more facts. But
13. facts are imposed to us by the world, so we should also concede that
14. empirical statements accepted as true after careful consideration are imposed to us by the world. Therefore
15. we are not free to choose the empirical statements we accept as true after careful consideration; and
16. we are not morally responsible of the empirical statements we accept as true after careful consideration.

39What is Raccah’s strategy to rebut this argument? According to Raccah, there are two very different concepts of a fact that we may have in mind. Let us call the things they are a concept of, respectively, “fact1” and “fact2”. A fact1 is an obtaining state of affairs independent of us and distinct from any description of it. 13. is true of facts1, if any. But we cannot know anything about facts1, not even whether they exist or not, because we have no cognitive access to them. As Raccah clearly asserts:

We cannot know what the “essential features” of the world are, principally because we could not be in the position to exhibit “essential features” of the world which our cognitive system does not handle: the picture our cognitive apparatus can give us of the world is only a picture of what we can grasp about it... (Raccah 2005a: 9)

40Since we have no cognitive access to facts1, we cannot ever know whether a specific descriptive statement is a true description of a fact1 or not. In other words, if we assume facts=facts1, we can never know whether a specific descriptive statement satisfies requirement 12. or not. Thus, we should accept as true no empirical statement, and no accepted-as-true empirical statement should exist. Since we accept as true a lot of empirical statements, however, we must conclude that they are not the kind of empirical statements satisfying requirement 12. under the condition facts=facts1. Therefore, if facts=facts1, we cannot conclude 14., 15. and 16.

  • 13 See also Raccah (2006: 3-4), Raccah (2008: 62) and Raccah (2010: 125) on this point.

41Let us now turn to the concept of “fact2”. Facts2 are part of “the picture our sensorial and cognitive apparatus can give us of the world”, and this is why we do have a cognitive access to them. But since they are part of that picture, they have a representational nature, and this is why they are not identical to facts1, which are characterised as being “distinct from any representation of them”. According to Raccah’s empiricist position, the picture of the world we can produce is only made out of our “raw perceptions” (Raccah 2005b: 8) (the empirical raw materials) via unconscious and conscious interpretations of them. Facts2 are nothing but the products of our interpreting in certain particular ways our sensorial and perceptual experiences. Arguably we produce facts2 by interpreting our sensorial and perceptual experiences as if they were giving us cognitive access to facts1 – which is not actually the case, as previously explained13. Facts2 are constructed by the subject; when Raccah talks about “the construction of observable facts”, he is referring to what I am calling “facts2”:

The role of the observer in the construction of the observable facts cannot be considered as a mere ‘deformation’ of a reality which would be external to her / him, but rather, as a construction of an accessible reality. The externalisation of that accessible reality constructed by the observer can only be posterior to its construction. This externalisation is the result of a social and cognitive process. (Raccah 2007: 1088)

42Although we may sometimes speak of “directly observable facts”, we must remember that no fact is directly observable, and that we gain cognitive access to facts2 just because we construct them moving from the only empirical raw materials we have available, i.e. our sensorial and perceptual experiences:

Technically, nothing is really directly observable, since we must interpret what our senses grasp; however, some entities are more indirectly observable than others: for instance, the mass of a block is observable only through a device which one must trust in order to assess it, while the existence of the block need not such an external device in order to be assessed. In the second case (when no external artificial device is needed), it is a normal language use to speak of ‘direct observation’, though, we insist, it is only a less indirect observation. (Raccah 2007: 1087)

  • 14 In such cases we may have no idea of what the new facts2 (the new “observable facts”) could be like (...)

43Facts2 are the result of a certain number of interpretative acts. Being the result of a certain number of interpretative acts is constitutive of facts2, since facts2 would not be what they are if they were not the result of that certain number of interpretative acts. Our interpretative acts, however, are not imposed by the world. No doubt that we are free to interpret our sensorial and perceptual experiences in many different ways, and some of them do not end in the construction of facts2 (for example, we are free to interpret our sensorial and perceptual experiences as if they were experienced in a dream, or as if they were caused by a Cartesian demon, and so on14). Moreover, also in case we choose to interpret our sensorial and perceptual experiences as to produce facts2, there seem to be many different and incompatible ways to do it. We can express this idea by saying that evidence is underdetermined by sensorial and perceptual experiences, just as scientific theories are underdetermined by evidence. Thus, if facts=facts2, 13. must be rejected, and we cannot conclude 14., 15. and 16.

44Since both under the case in which facts=facts1 and under the case in which facts=facts2 we cannot conclude 14., 15. and 16., it seems that the argument can be rejected. So Raccah has no difficulty in maintaining that we are morally responsible of the empirical statements we accept as true. In short, his position can be summed up like this: empirical statements would be imposed by the world if their truth-maker were facts1; but no empirical statement can be said to be true of facts1; empirical statements can be at best said to be true of facts2, and facts2 are not imposed by the world. Since we are free to make choices about what facts2 we construct, we necessarily are also free to make choices about what empirical statements we accept as true. Therefore we are morally responsible both of the facts2 we construct, and of the empirical statements we accept as true of them.

45Two specifications are in order. Firstly, once we have fixed facts2, naturally we are no longer free to choose the empirical statements we accept as true of them, since what empirical statements are true of facts2 is the consequence of facts2 and semantic rules. Yet Raccah’s point is that, in our being free to interpret our sensorial and perceptual experiences as to produce different sets of facts2, we are ipso facto free of making true different sets of empirical statements. Secondly, to say that facts2 – and empirical statements true of them – are not imposed by the world, does not amount to say that the world has no role in determining them. No doubt that the world has a direct causal role in producing our sensorial and perceptual experiences. Yet these experiences are the consequences of the world and of our sensorial and perceptual apparatus; and, it is impossible for us to distinguish the one causal role from the other (what is equivalent to say that it is impossible for us to identify facts1). Moreover, facts2 – and empirical statements true of them – are underdetermined by our sensorial and perceptual experiences, and we can just get to facts2 – and to empirical statements true of them – through our interpreting the sensorial and perceptual experiences we have. Thus facts2 are not arbitrary nor independent of the world (in fact they would be different were the world different from what it is) while being, at the same time, constructed by us. This is what Raccah remarks when he writes:

  • 15 See also, for example, Raccah (2011: 314): “Croire que la science n’est qu’idéologie est aussi ennu (...)

Acknowledging that our beliefs about the existence of what we perceive cannot be invoked as a proof of its existence is something different from believing that those beliefs are false.
Acknowledging that the way we perceive the world is influenced by our theoretical biases is something different from believing that the world plays no role in the way we perceive it.
Understanding that there is no way for us to know the distance between how we represent the world and how it really is is something different from believing that there is no real fact in the world, and does not prevent from believing that science does help shortening the distance. (Raccah 2005b: 12-13)15

46Supposing facts1 to be such-and-such, and even supposing them to exist, is irrational. Raccah confesses to have himself this belief, but he demands that “no scientific method or conception can be based on it” (Raccah 2007: 1086). On the contrary, interpreting our sensorial and perceptual experiences as to provide certain facts2, and accordingly accepting certain empirical statements as true of these facts2, is not irrational. This is the only thing we can do in order to increase our knowledge of the world. However, we should not confuse facts2 and facts1. Facts2 are (partly) chosen by us; we are morally responsible of accepting them (while we would not be responsible of accepting facts1 if we could have cognitive access to them).

47On my view of Raccah’s overall philosophy, the concept of moral responsibility plays a major role in his epistemology, theory of knowledge and philosophy of science, just as it played one in his moral epistemology and ethics. Scientists cannot ever say that “it is not their fault” that a certain fact holds, or that a certain empirical statement is true. We are morally responsible of the empirical statements we hold as true, and appeal to empirical truth is not sufficient to exempt us from this responsibility – since even observational statements are not imposed by the world. Thus Raccah can claim that “scientific neutrality is irresponsible and false” (Raccah 2005b: 27), and that no scientific theory can be argumentatively justified moving from non-argumentatively justified descriptive premises. In fact, there is no non-argumentatively justified descriptive statement. Science, like morality, cannot be founded – and scientists are morally responsible of the particular picture of the world they suggest us to adopt, as well as of the facts, entities and properties they encourage us to put into it. Even in science, justification cannot be but argumentative justification in the end, and the equivalent to the J’Aime / N’Aime Pas Thesis for the descriptive domain holds: the real final premises necessarily are of the kind “J’aime” (“I love / want to consider descriptive statement d as a true description of one or more facts”) or “Je n’aime pas” (“I do not love / want to consider descriptive statement d as a true description of one or more facts”).

48Since both in ethics and in science all justification cannot be but argumentative justification in the end, I take it only logical that Raccah decided to successfully become one of the major philosophers of argumentation of our time.

Haut de page

Bibliographie

Albert Hans, Traktat über kritische Vernunft, Tubingen, Mohr Siebeck, 1968. English Edition, Treatise on Critical Reason, Princeton, Princeton University Press, 1985.

Bacchini Fabio. 2015. “A Popperian Approach to Rational Argumentation in Applied Ethics”, Journal of Argumentation in Context, 4, 3 : 243-285.

Frankfurt Harry Gordon, “Freedom of Will and the Concept of Person”, Journal of Philosophy, 68, no 1, 1971, pp. 5-20.

Fries Jacob Friedrich, Neue Oder Anthropologische Kritik der Vernunft; 2nd edition 1828-1831, Heidelberg, Winter, 1807.

Hill Scott, “Good News for the Logical Autonomy of Ethics”, Argumentation, no 23, 2009. pp. 277-283.

Kant Immanuel, Critique of Judgment, 1790. Eng. transl., Oxford, Oxford University Press, 1928.

Karmo Toomas, “Some Valid (but not Sound) Arguments Trivially Span the “Is”–“Ought” Gap”, Mind, no 97, 1988, pp. 252-257.

Nelson, Mark T., “Is It Always Fallacious to Derive Values from Facts?”, Argumentation, no 9, 1995, pp. 553-562.

Nelson, Mark T., “More Bad News for the Logical Autonomy of Ethics”, Canadian Journal of Philosophy, no 37, 2007, pp. 203-216.

Popper Karl Raimund, Logik der Forschung, Wien, Julius Springler Verlag, 1934. English edition, The Logic of Scientific Discovery, London, Hutchinson, 1959.

Prinz Jesse J., The Emotional Construction of Morals, Oxford, Oxford University Press, 2007.

Raccah Pierre-Yves, “Puntos de Vista y Persona. Cuestiones de Semántica”, Acta Poetica, 25, no 2, 2004, pp. 203-254.

Raccah Pierre-Yves, “What Is an Empirical Theory of Linguistic Meaning a Theory of?”, in Zygmunt Frajzyngier, David Rood and Adam Hodges (eds.), Linguistic Diversity and Language Theories, Amsterdam, Benjamins, 2005a, pp. 51-80.

Raccah Pierre-Yves, “Scientific Neutrality Is Irresponsible and False”, Invited lecture [manuscript], Pécs University (Hungary), 2005b.

Raccah Pierre-Yves, « Polyphonie et argumentation : des discours à la langue (et retour…) », in Simonffy Zsuzsa (éd.), L’un et le multiple, Budapest, Tinta Könyvkiadó, 2006, pp. 120-152.

Raccah Pierre-Yves, “Points of View and Argumentation: Semantics as an Empirical Science”, Proceedings of the 6th Conference of the International Society for the Study of Argumentation, SicSat, 2007, pp. 1085-1093.

Raccah Pierre-Yves, « Contraintes Linguistiques et Compréhension des Énoncés : La Langue comme Outil de Manipulation », Entretiens d’Orthophonie, Paris, Expansion Formation et Éditions, 2008, pp. 61-90.

Raccah Pierre-Yves, « Racines Lexicales de l’Argumentation: La Cristallisation des Points de Vue dans les Mots », Verbum, 32, no 1, 2011, pp. 119-141.

Raccah Pierre-Yves, “Linguistique Critique: Une Exploration Cognitive…”, Intellectica, no 56, 2011, pp. 305-314.

Zangwill, Nick, "Aesthetic Judgment", in Edward N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Summer 2013 Edition). Disponible sur Internet : <http://plato.stanford.edu/archives/sum2013/entries/aesthetic-judgment/>.

Haut de page

Notes

1 He still holds this view to date (personal conversation 2014). See Raccah (2004) for an attempt to develop the idea that the personality of an individual can be described by drawing a map of the individual’s positive and negative commitments to ordinary points of view mainly consisting in judgements of taste.

2 See Kant (1790). Also see, e.g., Zangwill (2013).

3 Since Hume set the problem, many philosophers have written in support or against his thesis. For a recent debate concerning the possibility to violate Hume’s law, see e.g. Karmo (1988), Nelson (1995, 2007), Hill (2009).

4 By ‘the most robustly founded moral position’ I want here to refer to either the moral position that turns out to be founded - in case only one moral position turns out to be founded - or, to the moral position whose foundation appears to us as altogether most compelling - in case that more than one moral position turns out to be founded.

5 From this point on, whenever I say “action” and “acting”, what I intend is “action or omission” and “acting or omitting to act”.

6 He uncompromisingly expressed this position in many scientific talks and other public occasions. On this point see also Raccah (2007). A corollary of this claim is the thesis that “no human being can justify a reprehensible action arguing that they have been told or even ordered to accomplish them. If a war criminal tries to do so, he or she will give the justified impression that he or she is not behaving like a human being, but rather like a kind of animal or robot.” (Raccah 2007: 1088). The latter thesis parallels the “anti-matter” hypothesis (“The linguistic effects of an utterance are not due to material causes”) and the “anti-magic” hypothesis (“The linguistic effects of an utterance are not directly caused by them”) introduced in linguistics as limitations to the general hypothesis that “utterances can cause behaviours” (ibidem).

7 E.g., under both the counterfactual and the actual sequence view: it is true that I could have done otherwise if I had wanted to, and it is equally true that the course of events was at least partially caused by a mental state of mine consisting in an intention to act as I actually acted.

8 Raccah (personal communication 2014).

9 Note: we need to be able to know that the effect is caused, among other factors, by our acting as we act; and the act must be free; but it is not necessary that we are able to know that the effect is caused, among other factors, by our freely acting as we act. We are morally responsible also if we (even necessarily) falsely believe that we are not free in our acting, provided that we have potential access to knowing that we, among other factors, cause the effect by acting as we act.

10 “Siempre queda la posibilidad de resistir a la tendencia a connectar el grafo” (Raccah 2004: 8).

11 Raccah (2005b).

12 By “descriptive argumentation” I mean any argumentation whose conclusion is a descriptive proposition. In particular any argumentation concluding that a certain scientific theory or a certain empirical statement is true, is a descriptive argumentation.

13 See also Raccah (2006: 3-4), Raccah (2008: 62) and Raccah (2010: 125) on this point.

14 In such cases we may have no idea of what the new facts2 (the new “observable facts”) could be like. This is why I take these interpretations as prima facie not necessarily ending in the construction of any kind of facts2.

15 See also, for example, Raccah (2011: 314): “Croire que la science n’est qu’idéologie est aussi ennuyeux, stérile et dangereux que croire que la science est neutre.”

Haut de page

Pour citer cet article

Référence électronique

Fabio Bacchini, « Epistemology and Responsibility »Corela [En ligne], HS-19 | 2016, mis en ligne le 12 septembre 2016, consulté le 19 avril 2024. URL : http://journals.openedition.org/corela/4509 ; DOI : https://doi.org/10.4000/corela.4509

Haut de page

Auteur

Fabio Bacchini

Laboratory of Applied Epistemology, Dadu, University of Sassari & Department of Philosophy, Stanford University

Haut de page

Droits d’auteur

CC-BY-NC-SA-4.0

Le texte seul est utilisable sous licence CC BY-NC-SA 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search