1. Introduction

Suppose that I borrow Tolstoy’s book Anna Karenina from my friend Anne. The deadline falls on the 30th of July 2015. Then, other things being equal, I have the duty to return the book on this date. However, around the beginning of May I receive the information that my friend Anne has gone through a very painful love story. Married with an important member of the Supreme Court in Catalonia, she has fallen in love with a young, brilliant and ambitious university professor. For him she abandoned her husband and her little daughter and, after some months, the young professor decided to put an end to their romance. Anne is now very depressed, under psychiatric treatment, and has dangerous feelings oriented towards committing suicide. I know that she has no idea as to the content of Tolstoy’s novel, and I am sure that she will read the novel. Do I still have duty to return the book that I borrowed when the deadline comes?

This example evokes the famous passage from Plato in the Book I of The Republic, in which Socrates, arguing on justice, put the following question to Cephalus: “Suppose that a friend when in his right mind has deposited arms with me and he asks for them when he is not in his right mind, ought I to give them back to him? No one would say that I ought or that I should be right in doing so, any more than they would say that I ought always to speak the truth to one who is in his condition” (^{Plato 1888, 331c}).

A similar example is used by Alchourrón-Bulygin to illustrate their account of practical decisions: “Suppose that an agent A has been asked by his nephew B to give him a gun for his birthday” (^{1983, 23}). According to Alchourrón-Bulygin: “The regulative function of norms or deontic sentences consists in giving an answer to the practical question ‘What should I do’, that is, deontic sentences provide a guide for the decision an agent must take on a given occasion concerning his actions in some future moment” (^{1983, p. 19}).

This is an original way to put the question of the nature of norma- tive concepts. Recently, A. ^{Gibbard (2003)} has developed an account of normativity which is nicely illustrated by the following passage: “Thinking what I ought to do is thinking what to do” (2003, pp. ix- x). Normative concepts are plan-laden, they refer to the actions to-be-done.¹

In ^{Alchourrón-Bulygin 1983}, the authors display two theses on the realm of normativity:

All individuals in ideal conditions for any given set of histories would accept the same hierarchy. (1983, p. 33)

Acceptance of this principle gives rise to the absolutist branch of conceptualism. Its rejection yields the relativist branch. The authors also regard the objections to the absolutist branch as very difficult to overcome, and show a certain preference for the relativist account.

The paper contains an elegant and seminal development of the articulation of topological trees symbolizing the actual and future situations of the agent and the construction of several hierarchical orderings for the histories. In the case of Anne Karenina, once the book has been borrowed the situation before the deadline splits into two branches: in one, I return the book to Anne, in the other I do not.

In this contribution, I endorse this promising presentation of our practical problems. I shall provide some replies to the objections to the absolutist branch of conceptualism, thus making it more difficult to accept the relativist branch.

2. Alchourrón and Bulygin’s Objections

At the end of the paper (^{Alchourrón-Bulygin 1983}), the authors present three objections against the absolutist branch on behalf of the relativist. The first objection is formulated against ^{Frankena’s position (1963, pp. 33-34)}:

It is interesting to confront this point with Frankena’s remarks about relativism. He points out that in order to prove relativism it is not enough to show that the ethical and evaluative judgments of different people and societies are different, for such disagreements “might be due to differences and incompleteness in their factual beliefs”. What one must prove is that such judgments “would differ and conflict even if they were fully enlightened and shared all the same factual beliefs”. And as this proof is extremely difficult to produce, ^{Frankena (1963, pp. 93-94)} concludes that “metaethical relativism has not been proved”.

Frankena is certainly right, except in his characterization of relativism. A relativist needs not maintain that value judgments formulated in ideal conditions would still be different and incompatible. It is sufficient for him to state that one cannot rule out this possibility. In other words, all a relativist must do is to deny the principle of correlation, i.e. that the satisfaction of certain formal conditions leads necessarily to a consensus regarding the same hierarchy. And it is the absolutist who has the burden of proof. (^{Alchourrón-Bulygin 1983, pp. 33-35})

We shall call this argument the burden of proof argument.

The authors admit that the consensus which an absolutist needs is not an actual agreement of the members of a certain society: what is needed is instead “a hypothetical agreement of rational agents” (1983, p. 34). But, how can we prove the principle of correlation? If the disagreement about the hierarchy is always evidence that the ideal conditions have not been satisfied, then the principle of correlation is true by definition, although perhaps arbitrary. And if the disagreement about the hierarchy is only a sign or symptom of the violation of the ideal conditions, then we need an external criterion of correctness for the formal conditions (as in the realist account).

This is the second objection, which we can call the dilemma of Euthyphro’s argument.

The third objection consists in arguing against the possibility of obtaining the solution for all practical problems from the agreement on some very general moral principles: “It is for the absolutist to prove that it is always possible to infer from his general principles a concrete hierarchical order of histories we need for each particular practical problem” (^{Alchourrón-Bulygin 1983, p. 35}). This objection is the particularism argument.

3. The Burden of Proof Argument

The burden of proof argument is put forward by Alchourrón and Bulygin, against Frankena’s argument, in favour of the relativist branch of ethics. It seems that the authors think that rejection of the principle of correlation constituted the default position. It is on the affirmative part that the burden of proof lies. This is an argument with a long philosophical tradition, as, for instance, one can find it in George ^{Berkeley (1843, p. 180)}; here, while arguing for scepticism, the sceptic Philonus answers the dogmatic Hylas in the following way: “The proof should lie on him who has the affirmative.” The same idea is endorsed by Thomas Reid in the quotation which opens this contribution.

Perhaps appeal to this notion originates in the legal notion of onus probandi: agenti incumbit probatio. None the less, in legal procedures the principle owes its appeal to the fact that the burden of proof must be put on one of the parties in order for the decisions to be taken within a limited amount of time. Further, in this context it is not the case that the pursuit of truth constitutes the unique value at stake, as minimizing the punishment of innocent people -that is obtained with the presumption of innocence- also plays an important role. Something similar happens in formal debates, where every view should be rebutted in a certain procedural order. As Dare-Kingsbury said: “Formal debate and legal discourse, we have argued, are not solely or primarily truth directed. If truth was their main aim, we suggest, differential allocation of the burden of proof would be illegitimate, because it is not truth conducive” (2008, p. 508). This conclusion seems indeed to be widely accepted within the literature that deals with the value of the burden of proof in rational argumentation.³ In Keith Lehrer’s words:

generally arguments about where the burden of proof lies are unproductive. It is more reasonable to suppose that such questions are best left to courts of law where they have suitable application. In philosophy a different principle of agnoiology [the study of ignorance] is appropriate, to wit, that no hypothesis should be rejected as unjustified without argument against it. Consequently, if the sceptic puts forth a hypothesis inconsistent with the hypothesis of common sense, then there is no burden of proof on either side. (^{Lehrer 1971, p. 295})

In philosophical argument we can put forward almost any thesis and in the philosophical tradition there is room for almost any thesis: from the non-existence of the external world to the existence of God. What we need is arguments showing that the thesis which has been put forward fits better with other independently motivated theses, so as to achieve, to use a popular expression of ^{Rawls (1971, pp. 48-52)}, a reflective equilibrium among our considered judgments. And this was precisely the intention of Frankena. Frankena presented his ideas on relativism in the chapter of his book dedicated to meaning and justification, and he tried to show that there is a space for objectivity in moral reasoning.

Therefore, the burden of proof lies neither on the absolutist nor on the relativist, since both are obliged to justify and rationally argue for their claims. In fact, this objection does not seem to be very relevant to Alchourrón and Bulygin, given that they put forward other more powerful objections against the endorsement of the principle of correlation.

4. The Dilemma of Euthyphro’s Argument

As it is well-known, Euthyphro’s dilemma was put in the mouth of Socrates in ^{Plato (1892, 10a)}, where Socrates asked Euthyphro whether certain acts are pious because they are loved by the gods or they are loved by the gods because they are pious. Crispin ^{Wright (1992, pp. 79-80)} calls this opposition the Euthyphro Contrast and it can be applied in the following way to our question on absolutism and relativism. Let me assume the following biconditional which expresses the conceptualist account:

For any action x: x is right if, and only if, it would be chosen by ideal observers in normal conditions.⁴

If you opt, as Socrates does, for the left-to-right reading of the biconditional, you get moral realism. And if you prefer instead the reading from right to left, as Euthyphro does, then you have no guarantee of achieving acceptable moral judgments. Shafer-Landau put forward this version of the Euthyphro contrast in this way:

Either the initial conditions of choice or attitude formation are moralized or they are not. In other words, we are to envision the initial conditions as already incorporating moral constraints, or as operating free of such constraints. The problem with the latter option is that there is no reason to expect that the principles that emerge from such a construction process will capture our deepest ethical convictions, or respect the various platitudes that fix our understanding of ethical concepts. If fairness doesn’t characterize the contractors, or we can’t attribute benevolence or kindliness to an ideal observer, then we shouldn’t expect their responses to align neatly with our views about paradigmatic cases. Worse than that, we may be left with a set of views that are not recognizably moral ones. To justify abandoning our central ethical views for those that would emerge from such a process, we would need to have greater warranted confidence in the process than in the particular views themselves. That is highly unlikely.

Alternatively, if constructivists import moralized constraints, and so take the former option, then they effectively abandon constructivism, because this path acknowledges the existence of moral constraints that are conceptually and explanatorily prior to the edicts of the agents doing the construction. These constraints are not themselves products of construction, and so there would be moral facts or reasons that obtain independently of constructive functions. This is realism, not constructivism. (^{Shafer-Landau 2003, p. 42})

It is very difficult to overcome this objection and what I shall say is very tentative and framed in a rather speculative way. Maybe the competence that we rely on in speaking our native languages could be useful here. The comparison with linguistic practice follows ^{Rawls’s suggestion (1971)} when he deals with the sense of justice: “A useful comparison here is with the problem of describing the sense of grammaticalness that we have for the sentences of our native languages” (1971, pp. 46-47). This idea has been taken as a basis to defend the idea of a universal moral grammar (^{Mikhail 2011}), but it will not be this highly controversial idea what will be highlighted here. It is obvious that the notion of linguistic competence contains an idealization. And this is so even when it is argued that all adult and normal human beings are competent speakers of their native languages because they can understand its sentences and easily produce new sentences; our notion of correct sentence in a language L needs to meet certain requirements: they must be adults persons, they must not be drunk or excessively tired, for instance, and for complex sentences they must have certain level of education and culture. If, as it seems, there is room for error, then some level of expert knowledge figures among the ideal conditions. Therefore, it is only from linguistic practice that it is possible to reconstruct the minimal requirements of the ideal conditions.

Analogously, it must be from the salient features of moral practice, our folk morality, that we are able to reconstruct the requirements of our ideal conditions. They are the background of convictions, judgments and attitudes that shape our moral practice, the platitudes of our practice (the idea and its development follow Smith 1994, 39- 41).

First, we have the platitude of practicality. Our moral judgments are connected with what is (to be) done. If I say that a given action A is right, then ceteris paribus I have a pro-attitude to do A. It can obviously be that I do not do A, but it will be due to some shortcoming in my psychological mechanism: weakness of will, compulsion, depression and so on.

Secondly, if x thinks that A is the right thing to do, and y thinks that it is not, then x and y have a genuine disagreement. This is the platitude of claim of correctness.

Thirdly, human actions that share all the relevant natural properties must necessarily share their moral features. If it is morally wrong to annihilate Jews (as it was decided at the Wannsee Conference), then it is also wrong to annihilate any other ethnic group. This is the platitude of supervenience.

Fourthly, morality is entangled with our actions that affect in a significant way the well-being of human beings (some will prefer to extend this class to cover other living beings). The fact that I write this text in a Cambria typeface or in another typeface does not affect human well-being and, therefore, lacks moral relevance. Here we have the platitude of substance.

Fifthly, and finally, there is the platitude of procedure. When we disagree on moral issues, we often try to solve our disagreements by looking into those issues, which perhaps are more basic, where we already agree, and we are disposed to revise our convictions by taking the reasons of other people into account. Our moral reasoning has a coherentist form.

If this is a suitable presentation of the salient features that underlie our moral practice, then it will be reasonable that the ideal conditions of our moral observers incorporate these features. This incorporation will allow the ideal observers to avoid the biases that are always present in our judgments, for only by avoiding our biases can we achieve an element that is crucial to morality: reciprocity. Similarly, competent speakers of a language should be able to avoid the most idiolectal and dialectal uses, for only in this way they can claim reciprocity.

Obviously, a suitable account of the ideal conditions should justify the impartiality related to the idea of suppression of biases. It should also say what amount of factual information must be possessed by the observers, whether what is required from them is something like omniscience or rather only a degree of information that meets a sensible standard. It should also be explained in which sense the ideal observers have a unified and coherent set of desires. Moreover, one should articulate a notion of imaginative capacity and acquaintance that allows the observers to put themselves in the shoes of other people. In all these elements there is controversy and it is a major challenge in moral epistemology to endow the notion of ideal conditions with better conceptual foundations.

If the materials for the construction of morality are composed of the platitudes of our actual moral practices, and if these materials cannot be taken to be arbitrary -given that any arbitrary conclusion would violate the platitudes-, then it seems that we need no external element. And this is precisely our answer to the dispute that takes place in the Euthyphro: the ideal conditions are not empty since we attribute to the observers an optimal competence to represent the platitudes of the practice, but at the same time these conditions do not transcend the practice. Rightness is what the ideal observers choose, but these ideal agents are neither blind to morality nor do they make arbitrary choices, since they are idealized participants of an actual moral practice.

5. The Particularism Argument

Alchourrón-Bulygin argue:

Most absolutists have emphasized the possibility of agreement regarding some very general moral principles. Unfortunately this is not sufficient for the solution of all practical problems; what we need is the acceptance of a hierarchical order of histories that take into account all the contingencies of a specific situation. It may be the case that even having agreed about some general principles, disagreement persists regarding the ordering of histories for a given particular problem, and so we do not know which deontic sentences are true. (1983, p. 34)

And also: “Factual knowledge required for knowledge of deontic truth is an ideal hardly attainable in practice. It is empirically impossible for a human agent to know all past facts and to foresee all future contingencies. Hence a human agent cannot attain full knowledge of deontic truth” (1983, pp. 30-31). These considerations have a particularist flavour, since given the infinite number of possible combinations of features that we find in human life, there is no possibility to codify them in a set of principles of moral requirements.⁵

Let me go back to the example of borrowed books. Everyone accepts that, in normal conditions, when the deadline comes I ought to return the book. But, the particularist objects, what if I realize that the book has been stolen from the University Library, for instance? What if my friend disappears and I am not able to find her? Then, surely my duty is cancelled. According to particularism, the fact that a feature of a human action is relevant to its rightness in a certain circumstance does not mean that this feature will be relevant in all other circumstances.

I shall present here a feasible way to move from general principles to all things considered solutions. My proposal relies on the adoption of a specificationist account.⁶ Let me now go back to the example of Anne Karenina.

I set up the context, wherein universal standards are included, in a five stage process.

The first stage is devoted to delimiting the normative problem, that is, to select the human actions that constitute the universe of discourse.⁷ Ascertaining the universe of the normative problem allows us to make the problem tractable by avoiding having to take all human actions into consideration. In the Karenina case the universe of discourse can be: the human actions of returning borrowed books to friends.

The second stage is the identification of the standard and principles that are prima facie applicable to this universe of discourse. Here, obviously, are applicable the principles which establish the duty to return borrowed books when the deadline arrives and the duty to avoid irreparable harm to our friends.

The third stage consists in taking into account certain paradigmatic cases, actual or hypothetical, of the universe previously selected in the first stage. The paradigmatic cases have the function of constraining the scope of admissible reconstructions: those reconstructions are admissible which efficiently cover the paradigmatic cases. Paradigmatic cases lie in the background, often unarticulated, where practical reasoning is placed (^{Bayón 2001}). In our universe of actions, we could regard as paradigmatic cases circumstances such as the following: a) to return an exemplar of Anna Karenina in Russian language borrowed from a friend who needs it because she is preparing a doctoral dissertation on adultery in eighteenth century literature b) to return the book in spite of having discovered that the pages of the book have toxins of anthrax.

In the fourth stage the relevant properties of the universe of discourse are established. The relevant properties will lead to the normative solution. In our case the following properties are clearly relevant: the arrival of the deadline, and the avoidance of irreparable harm to our friends.

The fifth and last stage consists in the formulation of the universal standards for this context, rules that univocally solve all the cases in the universe of discourse. A rule, which seems to me indisputable, would be the following:

R1: When the deadline arrives and the borrowed book will not cause an irreparable harm to the owner, then you have the duty of returning the book.

And the following seems to me to be indisputable:

R2: If the returning of the borrowed book will produce an irreparable harm to the owner, then you have the duty of avoiding to do it.

Obviously, the last three stages are intimately intertwined. Ascertaining the validity of rules should involve testing in accordance with their potential to take paradigmatic cases into account. The selection of the relevant properties must be refined until this aim is achieved, and when starting anew with the selection process a new set of rules should also be formulated.

In this context, a context in which only these properties are relevant, the rules R1 and R2 are universally and univocally applicable. Obviously, if the circumstances change, for instance, because I discover that the book has been stolen from the University library or it turns out that the whereabouts of my friend remain unknown, then the normative solution should be different.

I shall leave for another occasion the examination of the extent to which this specificationist account is similar to moderate accounts of generalism in morality. ^{Holton (2002)} proposes a clause, that’s it, that guarantees by default the fact that the applicable principle is not superseded. ^{McKeever and Ridge (2006)} prefer the clause “and no other feature of the situation explains why. . . ”. ^{Väyrynen (2006)} opts for a way of hedging the moral principles. I choose to relativize to a context of use and a context of assessment the generality and the conclusive stringency of the principles.

Be that as it may, this strategy shows how to construct a channel between a set of general principles and conclusive deontic judgments.

6.Conclusions

In this contribution, I have tried to show that in an old paper of Alchourrón and Bulygin there are some ideas that are crucial in developing a theory of morality, understood as a theory of thinking how to live, evoking the title of ^{Gibbard’s book (2003)}.

I have also discussed their objections against the objectivist branch of conceptualism (constructivism). At the end the paper, they assert that “the expectations that lie behind the satisfaction of the ideal of systemic unity postulated by realists and absolutists are not ruled out by relativists” (1983, p. 35). Well, I hope that my arguments can help make a contribution to this ongoing debate. I cannot think of a better way of participating in this workshop in Washington dedicated to the philosophy of Eugenio Bulygin.⁸